Evaluating the Efficiency of Capital Markets: An Empirical
Analysis
Introduction
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
Efficient markets theory postulates that securities prices fully reflect all
available information at any given time to make them unbiased and difficult
to consistently outperform without accepting additional risk. This implies
market participants incorporate new publicly available information rapidly
and rationally, making it impossible to realize above-market gains through
strategies based on historical price patterns or other anomalies. However,
behavioral biases and information asymmetries create inefficiencies that can
potentially be exploited. This paper aims to empirically evaluate capital
markets efficiency through analyzing historical data and academic studies on
anomalies like momentum, value, and post-earnings announcement drift.
Patterns uncovered would contradict strong-form efficiency assumptions
while persistence over time challenges weak-form. Overall, the intent is to
assess evidence for and against efficiency to determine the degree markets
may indeed be semi-strong in practice.
Definitions of Efficiency
Different forms of the efficient market hypothesis (EMH) have been proposed
based on type and timeliness of information reflected in asset prices:
- Weak-Form EMH: Prices reflect all historical price and volume data. No
technical trading rules can achieve excess returns.
- Semi-Strong Form EMH: Prices rapidly reflect all publicly available
information like earnings announcements. Fundamental analysis cannot
consistently outperform passive strategies.
- Strong-Form EMH: Prices reflect all public and private information. Insider
trading cannot beat the market since information spreads perfectly.
Only the semi-strong form remains plausible as even EMH proponents
acknowledge limitations to strong-form assumptions about universally
shared knowledge. However, anomalies appear contradicting full semi-strong
efficiency as well. This paper evaluates various anomalies through academic
studies to gauge capital markets true information efficiency level.
Analyzing Momentum Anomaly
One of the most robust anomalies documented relates to price momentum
where past winners continue winning, losers continuing losing in the short-
run (6-12 months). Studies include:
- Jegadeesh and Titman (1993) found buying past 3-12 month winners,
shorting losers generated 1% monthly abnormal returns during 1965-1989.
- Momentum persisted profitably into the 1990s with strategies
outperforming market index by 3-4% annually.
- Momentum effect exists across asset classes like currencies, commodities
weakening EMH claims.
- Daniel and Moskowitz (2016) backtested 150 years of stock data,
momentum profits remained intact even when transaction costs included
after adjusting for risk.
- Hong and Stein (1999) rationalize momentum persisting if information
diffuses gradually rather than being instantly impounded into prices.
- Risk-based models fail adequately explaining momentum profits. Behavioral
interpretations involving overreaction, representativeness seem more
compelling.
Momentum contradicts EMH assumptions of prices efficiently adjusting
immediately to all public information. That patterns endure over decades
casts doubt on full market efficiency even in its semi-strong form.
Value Anomaly Analysis
Another regularly documented anomaly relates to value stocks with low
price-to-book, earnings yields outperforming growth stocks on risk-adjusted
basis over the long-term (3-5 years):
- Fama and French (1992) documented the value premium in US stocks
between 1963-1990 with returns 6-7.5% greater annually versus growth
counterparts.
- Value effect continued globally during 1990s across over 45 countries
studied by Fama and French.
- Lakonishok et al. (1994) found value outperformance of nearly 2% per year
from 1975-1990 among NYSE stocks held for 10 years.
- Asness et al. (2013) still observed value investing worldwide beating the
market by 3.8% average annually during 1990-2012 with profits resilient to
transaction costs.
- Behavioral arguments involving overreaction to recent growth, anchoring
biases better explain value effect than risk-based models.
The persistence of value across markets/eras lends credence to behavioral
drivers as rational risk premia arguments fall short. While value could
represent a risk factor, long-run alpha suggests relative mispricing at times
contra EMH.
Post-Earnings Announcement Drift
Another well-known anomaly is positive abnormal returns following quarterly
earnings surprises that persist for up to a year:
- Ball and Brown (1968) first reported positive drift for both good and bad
news companies lasting up to 12 months post-announcement.
- Bernard and Thomas (1989, 1990) verified drift of 4-5% over following year
concentrated in small stocks.
- Mendenhall (1991) still observed drift among NYSE/AMEX stocks during
1963-1988 periods.
- Chan et al. (1996) found drift concentrated after large surprises, diminished
but persisted after controlling for risk.
- Daniel et al. (1998) rationalize drift arising if analysts underreact to
information or news takes time spreading to market.
That surprise news continues influencing price trajectory months after
release when information has presumably been digested violates EMH
premises of prices instantly impounding all public facts. Investing strategies
appear able to exploit drift contrary to semi-strong efficiency.
Analyzing Size and Book-to-Market Effects
Models incorporating firm size and book-to-market equity ratios help
summarize cross-sectional differences in average returns beyond simple
market indexes:
- Fama and French (1992, 1993) three-factor model links average returns to
market, firm size and book-to-market factors. Captures value and size premia
not explained by CAPM.
- Some argue capturing risk factors rather than true anomalies. But Fama
and French (2015) still observed premia even controlling for other known risk
explanations.
- Momentum factor model of Carhart (1997) added momentum term,
strengthening explanation of anomalies versus market model alone.
- Factors models built on identified empirical effects rather than pure theory
hold up well out-of-sample, further validate anomalies found not fully priced
by markets.
Overall, multi-factor models empirically substantiate role of firm attributes
like growth, earnings quality etc. exerting persistent influence on returns
inconsistent with single-factor market efficiency alone. This supports semi-
strong efficiency having only limited explanatory power.
International Evidence
Anomalies also appear cross-nationally questioning global efficiency claims:
- Hou et al. (2014) found value, momentum and accruals anomalies
profitable across 42 countries between 1981-2013.
- Value premium persisted outside US during 1990s-2010s according to Fama
and French (2012).
- Makar (2016) confirmed momentum, value, quality factor premia existed in
19 developed nations 1985-2013.
- Jegadeesh et al. (2004) showed momentum persisted outside US markets
during 1990s at magnitudes similar to domestic findings.
That such patterns endure internationally where different conditions pertain
implies potential limits to EMH beyond any single jurisdiction or era, placing
doubt on perfectly informationally efficient global capital markets.
Other Anomalies
Further anomalies emerging from academic study undermine EMH
interpretations:
- Accruals anomaly - High accrual firms underperform cash flow stocks,
suggesting markets misprice earnings quality. Sloan (1996)
- IPO underperformance - Initial public offerings underperform benchmarks
following issuance. Ritter (1991), Loughran and Ritter (1995)
- Operating leverage effect - Firms with high operating leverage outperform.
Novy-Marx (2013)
- Net stock issues effect - Firms performing secondary equity issues
underperform versus repurchasing stocks. Spiess and Affleck-Graves (1995)
- Discounts on closed-end funds - Funds trade at persistent discounts,
defying net asset value rationales. Lee et al. (1991)
- Contrarian profitability - Out-of-favor/distressed stocks outperform. De
Bondt and Thaler (1985, 1987)
That many different types of anomalies have separately held up to scrutiny
over years casts doubt any single theoretical critique fully accounting for
empirical observations.
Evidence Synthesis
In aggregate, extensive research finds numerous persistent empirical
patterns in stock data worldwide inconsistent with prices fully impounding all
public information as EMH assumes. The diversity, longevity, international
scope and resilience of identified effects even after accounting for trading
friction argue against simple risk-based rationalizations:
- Anomalies cannot convincingly be explained as chance findings or data
mining tricks given decades of verification by academics using different
methodologies.
- Neither CAPM, multifactor models nor more recent anomaly-based factors
satisfactorily account for extent and persistence of anomalies suggesting
incomplete information efficiency.
- Behavioral arguments involving representativeness bias, overconfidence
and underreaction to information have superior explanatory power over
rational models to date.
While semi-strong EMH posits markets are difficult rather than impossible to
beat, anomalies imply potential for alpha generation in practice via
exploiting predictable relative mispricings even if such strategies do not
guarantee outperformance in every period. Overall, capital markets appear
“informationally semi-efficient” at best based on robust empirical
contradictions to pure EMH propositions.
Conclusion
In conclusion, a comprehensive empirical analysis of long-term academic
research provides substantial evidence questioning prevailing efficient
market theory assumptions, particularly the semi-strong form. Numerous
anomalies documented globally and over extended periods contravene
claims of prices fully reflecting public information immediately. Persistence
internationally and resilience to trading costs further weaken theoretical
critiques.
While rational risk-based stories offer logical aspirations, behavioral
justifications have superior empirical validity so far in accounting
parsimoniously for observed cross-sectional and time-series patterns. Active
investment strategies therefore hold prospect of capturing abnormal returns
by perceptively exploiting predictable relative mispricings, even if market
timing proves difficult. Overall, capital markets seem informationally
inefficient to a significant degree in practice, though not readily susceptible
to arbitrage either through permanent misalignments or transactional
barriers slowing adjustments. Future research incorporating new datasets
and methodologies may refine the degree and persistence of anomalies
further, offering ongoing opportunities for alpha-generation albeit requiring
active risk management in imperfectly efficient environments.
