Factor-Based Investing Strategies: Implementing Advanced Factor Models for
Investment Portfolio Construction
Introduction
Factor investing has become an increasingly important consideration for investors seeking to
drive excess returns. Rather than solely relying on stock returns or market indices, investors are
turning to factor-based strategies that target specific risk premia. This paper will delve into some
of the leading factor models that have emerged in recent years and how they can be
implemented to construct investment portfolios. It will begin by outlining the rationale behind
factor investing and some of the key factors that are commonly targeted. From there it will
explore advanced factor models and how they can enhance portfolio construction beyond
traditional single or multi-factor approaches. The implementation of such models will also be
discussed, including practical considerations for investors. Overall, this paper aims to provide an
in-depth look at state-of-the-art factor investing techniques and how they can empower investing
decisions.
Rational and Historical Basis for Factor Investing
At its core, factor investing is based on the principle that certain risk factors like size, value, and
momentum have historically generated excess returns or premiums over the broader market.
Over long periods of time, empirical research has shown that investing in stocks exhibiting these
"risk premia" characteristics has led to outperformance versus a simple market-capitalization
weighted index. Some of the seminal research establishing the importance of these factors
includes:
- Fama and French (1992,1993,1996): Their three-factor model demonstrated that size (market
capitalization) and value (book-to-market ratio) Premiums beyond market returns help explain
stock returns and should be considered systematic risk factors that compensate investors for
bearing additional risk.
- Carhart (1997): Extended the Fama-French model to include the momentum factor, showing
that stocks with positive past performance tend to continue outperforming for a limited period
(the "momentum premium").
- Asness et al. (2013): Comprehensively analyzed size, value, momentum, quality and low risk
factors across 26 developed and emerging markets over 40+ years, providing compelling
evidence of systematic risk premia globally.
Due to this extensive empirical foundation, factors like size, value, and momentum have
become key inputs for portfolio construction and risk modeling. By targeting stocks exhibiting
the characteristics associated with these premiums, investors seek to generate higher returns
over the long run. However, a limitation of early factor models is they typically involved only a
handful of factors and did not account for complex interactions between them.
Advanced Factor Models
More recently, quantitative researchers have developed progressively more sophisticated multi-
factor models that aim to better capture factor dynamics and generate more robust investment
signals. Several of the most advanced approaches include:
- Barra Models: Developed by MSCI Barra, these are multi-factor risk models that employ up to
30 individual factors including exposure to regions, industries, size, value, and quality
characteristics. The interaction effects between factors are explicitly modeled.
- Axioma Models: Axioma introduced a multi-factor model with over 100 covariates designed to
pick up nuanced industry and country factor exposures. It can distinguish factors like
sustainability versus cyclicality that impact stock returns.
- Northfield Models: Northfield factors represent underlying macroeconomic conditions rather
than just stock attributes. Their multi-region International model identifies over 60 factors
including terms of trade, real rates, and market liquidity dynamics.
- Anthic Factor Zoo Models: This AI-driven platform identifies adaptive factors in real-time using
unsupervised machine learning on vast financial datasets. Novel factors well beyond the classic
ones continually emerge without prescribed definitions.
By incorporating a much broader array of explicitly modeled factors and interactions, these
advanced approaches aim to provide a more complete picture of the true drivers of expected
returns. Their factor signals are also intended to be more robust to changing market
environments over time compared to staticfactor models.
Implementing Factor Investing Strategies
Given the depth of research on factor premia and availability of advanced factor models, the
next area of focus is implementing factor-based strategies effectively. Several crucial
considerations include:
1) Factor Tilt vs. Targeted Exposure
Portfolios can take a modest multi-factor "tilt" approach focusing on a handful of factors like
value, quality and momentum. Or pursue targeted factor "budgeting" constraining exposures to
specific factors and neutralizing others. The appropriate strategy depends on objectives,
constraints and risk tolerance.
2) Factor Combination & Interaction
Factors are not independent and their combination matters. Adding factors may improve
diversification, but too many can dilute signals. Interactions must also be modeled, such as
value stocks exhibiting momentum. Engaging a model provider is important for nuanced factor
blending.
3) Rebalancing Frequency
Rebalancing maintains targeted factor exposures and limits deviations from intended style.
However, frequent trading amplifies transaction costs. Balancing expected benefits against
friction, a quarterly or semi-annual rebalancing regime is often employed.
4) Implementation Vehicle
Strategies can utilize individual stocks and ETFs, factor mutual funds or model portfolios, smart
beta/strategic beta ETFs, or segregated accounts/separately managed accounts. Focus of the
mandate and scale/costs informs vehicle selection.
5) Fees & Liquidity
Transaction costs and management fees must be diligently assessed against gross alpha
potential of the strategy. Strategies with frequent rebalancing or narrow holdings increase
trading expenses. Liquidity screening also prevents bid-ask spread issues.
6) Risk Controls and Performance Tracking
Ongoing monitoring of portfolio factor exposures, risk budgets, tracking error and style drift
relative to intended targets is crucial. Performance attribution analysis decomposes returns by
factor, region and stock-level drivers to evaluate strategy success.
In practice, assembling the optimal mix of individual stock positions or investing vehicles, choice
of a suitable factor model, and setting rebalancing policies requires careful planning. But this
groundwork enables systematically capturing factor premiums over the long-term.
Case Study: International Small Cap Value Factor Portfolio
As an example, consider implementing an international small cap value factor tilt strategy as
follows:
- Universe: MSCI World ex-USA Small Cap Index (~1,500 stocks across 23 developed markets
ex-US)
- Factors: Small Cap exposure boost and Value factor screen via book-to-price ratio targeting
stocks in bottom 30%
- Model: Northfield International Multi-Factor risk model identifies exposures to 60+ macro and
company-specific factors
- Constraints: Target +1.5% small cap exposure and +2% value exposure vs benchmark,
neutralize all other factors
- Holdings: Equally weight 250-300 stocks passing factor criteria with liquidity screen
- Rebalancing: Semi-annually in June and December each year
- Vehicle: Separately managed account with low-cost ETF trades
- Performance: Track information ratio vs benchmark and attributes to factors annually
By leveraging a sophisticated multi-factor model, implementing tight style constraints via factor
budgeting, and minimizing costs through a concentrated portfolio and infrequent trading, this
international small value strategy aims to profitably capture premia in a risk-controlled manner
over market cycles. Ongoing governance ensures it adheres to the intended objective function.
Regulatory and Reporting Considerations
As factor investing grows in prominence, regulatory issues also deserve attention. SEC
guidelines including those from Rule 2a-7 on Money Market Funds and Rule 3c-5 on Diversified
Funds define criteria for concentrated, non-diversified or narrowly-focused investment products.
Strategies pursuing highly concentrated factor tilts must carefully consider these rules.
Additionally, various reporting standards exist for communicating factor portfolio compositions
and performance attribution in a transparent manner. The Global Industry Classification
Standard (GICS) is integral to classifying holdings by industry groups. The United Nations
Principles for Responsible Investment (PRI) provide a framework for assessing Environmental,
Social and Corporate Governance (ESG) integration.
Quantitative regulatory filings involving 13F holdings disclosures and Form N-PORT for US-
registered funds demand factor exposures align well-defined investment mandate descriptions.
International standards like the Asia Pacific Fund Classification Framework (FCF) similarly
guide factor portfolio reporting practices globally. Robust compliance programs ensure
strategies operate with integrity.
Conclusion
Over the past few decades, an abundance of empirical research has established the importance
of systematic risk factors like size, value, and momentum for driving equity returns. While early
models successfully captured these classic factors, modern portfolio construction has
progressed towards multi-factor risk modeling, factor combination, and dynamic factor-focused
strategies. Leading implementations utilize advanced multi-factor risk models, target balanced
factor tilts complementing overall objectives, and optimize implementation through careful
portfolio design, trading, and oversight practices. Looking ahead, as factor insights continue
adapting based on big data sources, their full potential for superior risk-adjusted performance
will likely be further unlocked. Factor-based portfolio construction remains a cornerstone
approach for investment managers.
Factor investing has become an increasingly important consideration for investors seeking to
drive excess returns. Rather than solely relying on stock returns or market indices, investors are
turning to factor-based strategies that target specific risk premia. This paper will delve into some
of the leading factor models that have emerged in recent years and how they can be
implemented to construct investment portfolios. It will begin by outlining the rationale behind
factor investing and some of the key factors that are commonly targeted. From there it will
explore advanced factor models and how they can enhance portfolio construction beyond
traditional single or multi-factor approaches. The implementation of such models will also be
discussed, including practical considerations for investors. Overall, this paper aims to provide an
in-depth look at state-of-the-art factor investing techniques and how they can empower investing
decisions.
Rational and Historical Basis for Factor Investing
At its core, factor investing is based on the principle that certain risk factors like size, value, and
momentum have historically generated excess returns or premiums over the broader market.
Over long periods of time, empirical research has shown that investing in stocks exhibiting these
"risk premia" characteristics has led to outperformance versus a simple market-capitalization
weighted index. Some of the seminal research establishing the importance of these factors
includes:
- Fama and French (1992,1993,1996): Their three-factor model demonstrated that size (market
capitalization) and value (book-to-market ratio) Premiums beyond market returns help explain
stock returns and should be considered systematic risk factors that compensate investors for
bearing additional risk.
- Carhart (1997): Extended the Fama-French model to include the momentum factor, showing
that stocks with positive past performance tend to continue outperforming for a limited period
(the "momentum premium").
- Asness et al. (2013): Comprehensively analyzed size, value, momentum, quality and low risk
factors across 26 developed and emerging markets over 40+ years, providing compelling
evidence of systematic risk premia globally.
Due to this extensive empirical foundation, factors like size, value, and momentum have
become key inputs for portfolio construction and risk modeling. By targeting stocks exhibiting
the characteristics associated with these premiums, investors seek to generate higher returns
over the long run. However, a limitation of early factor models is they typically involved only a
handful of factors and did not account for complex interactions between them.
Advanced Factor Models
More recently, quantitative researchers have developed progressively more sophisticated multi-
factor models that aim to better capture factor dynamics and generate more robust investment
signals. Several of the most advanced approaches include:
- Barra Models: Developed by MSCI Barra, these are multi-factor risk models that employ up to
30 individual factors including exposure to regions, industries, size, value, and quality
characteristics. The interaction effects between factors are explicitly modeled.
- Axioma Models: Axioma introduced a multi-factor model with over 100 covariates designed to
pick up nuanced industry and country factor exposures. It can distinguish factors like
sustainability versus cyclicality that impact stock returns.
- Northfield Models: Northfield factors represent underlying macroeconomic conditions rather
than just stock attributes. Their multi-region International model identifies over 60 factors
including terms of trade, real rates, and market liquidity dynamics.
- Anthic Factor Zoo Models: This AI-driven platform identifies adaptive factors in real-time using
unsupervised machine learning on vast financial datasets. Novel factors well beyond the classic
ones continually emerge without prescribed definitions.
By incorporating a much broader array of explicitly modeled factors and interactions, these
advanced approaches aim to provide a more complete picture of the true drivers of expected
returns. Their factor signals are also intended to be more robust to changing market
environments over time compared to staticfactor models.
Implementing Factor Investing Strategies
Given the depth of research on factor premia and availability of advanced factor models, the
next area of focus is implementing factor-based strategies effectively. Several crucial
considerations include:
1) Factor Tilt vs. Targeted Exposure
Portfolios can take a modest multi-factor "tilt" approach focusing on a handful of factors like
value, quality and momentum. Or pursue targeted factor "budgeting" constraining exposures to
specific factors and neutralizing others. The appropriate strategy depends on objectives,
constraints and risk tolerance.
2) Factor Combination & Interaction
Factors are not independent and their combination matters. Adding factors may improve
diversification, but too many can dilute signals. Interactions must also be modeled, such as
value stocks exhibiting momentum. Engaging a model provider is important for nuanced factor
blending.
3) Rebalancing Frequency
Rebalancing maintains targeted factor exposures and limits deviations from intended style.
However, frequent trading amplifies transaction costs. Balancing expected benefits against
friction, a quarterly or semi-annual rebalancing regime is often employed.
4) Implementation Vehicle
Strategies can utilize individual stocks and ETFs, factor mutual funds or model portfolios, smart
beta/strategic beta ETFs, or segregated accounts/separately managed accounts. Focus of the
mandate and scale/costs informs vehicle selection.
5) Fees & Liquidity
Transaction costs and management fees must be diligently assessed against gross alpha
potential of the strategy. Strategies with frequent rebalancing or narrow holdings increase
trading expenses. Liquidity screening also prevents bid-ask spread issues.
6) Risk Controls and Performance Tracking
Ongoing monitoring of portfolio factor exposures, risk budgets, tracking error and style drift
relative to intended targets is crucial. Performance attribution analysis decomposes returns by
factor, region and stock-level drivers to evaluate strategy success.
In practice, assembling the optimal mix of individual stock positions or investing vehicles, choice
of a suitable factor model, and setting rebalancing policies requires careful planning. But this
groundwork enables systematically capturing factor premiums over the long-term.
Case Study: International Small Cap Value Factor Portfolio
As an example, consider implementing an international small cap value factor tilt strategy as
follows:
- Universe: MSCI World ex-USA Small Cap Index (~1,500 stocks across 23 developed markets
ex-US)
- Factors: Small Cap exposure boost and Value factor screen via book-to-price ratio targeting
stocks in bottom 30%
- Model: Northfield International Multi-Factor risk model identifies exposures to 60+ macro and
company-specific factors
- Constraints: Target +1.5% small cap exposure and +2% value exposure vs benchmark,
neutralize all other factors
- Holdings: Equally weight 250-300 stocks passing factor criteria with liquidity screen
- Rebalancing: Semi-annually in June and December each year
- Vehicle: Separately managed account with low-cost ETF trades
- Performance: Track information ratio vs benchmark and attributes to factors annually
By leveraging a sophisticated multi-factor model, implementing tight style constraints via factor
budgeting, and minimizing costs through a concentrated portfolio and infrequent trading, this
international small value strategy aims to profitably capture premia in a risk-controlled manner
over market cycles. Ongoing governance ensures it adheres to the intended objective function.
Regulatory and Reporting Considerations
As factor investing grows in prominence, regulatory issues also deserve attention. SEC
guidelines including those from Rule 2a-7 on Money Market Funds and Rule 3c-5 on Diversified
Funds define criteria for concentrated, non-diversified or narrowly-focused investment products.
Strategies pursuing highly concentrated factor tilts must carefully consider these rules.
Additionally, various reporting standards exist for communicating factor portfolio compositions
and performance attribution in a transparent manner. The Global Industry Classification
Standard (GICS) is integral to classifying holdings by industry groups. The United Nations
Principles for Responsible Investment (PRI) provide a framework for assessing Environmental,
Social and Corporate Governance (ESG) integration.
Quantitative regulatory filings involving 13F holdings disclosures and Form N-PORT for US-
registered funds demand factor exposures align well-defined investment mandate descriptions.
International standards like the Asia Pacific Fund Classification Framework (FCF) similarly
guide factor portfolio reporting practices globally. Robust compliance programs ensure
strategies operate with integrity.
Conclusion
Over the past few decades, an abundance of empirical research has established the importance
of systematic risk factors like size, value, and momentum for driving equity returns. While early
models successfully captured these classic factors, modern portfolio construction has
progressed towards multi-factor risk modeling, factor combination, and dynamic factor-focused
strategies. Leading implementations utilize advanced multi-factor risk models, target balanced
factor tilts complementing overall objectives, and optimize implementation through careful
portfolio design, trading, and oversight practices. Looking ahead, as factor insights continue
adapting based on big data sources, their full potential for superior risk-adjusted performance
will likely be further unlocked. Factor-based portfolio construction remains a cornerstone
approach for investment managers.
Factor investing has become an increasingly important consideration for investors seeking to
drive excess returns. Rather than solely relying on stock returns or market indices, investors are
turning to factor-based strategies that target specific risk premia. This paper will delve into some
of the leading factor models that have emerged in recent years and how they can be
implemented to construct investment portfolios. It will begin by outlining the rationale behind
factor investing and some of the key factors that are commonly targeted. From there it will
explore advanced factor models and how they can enhance portfolio construction beyond
traditional single or multi-factor approaches. The implementation of such models will also be
discussed, including practical considerations for investors. Overall, this paper aims to provide an
in-depth look at state-of-the-art factor investing techniques and how they can empower investing
decisions.
Rational and Historical Basis for Factor Investing
At its core, factor investing is based on the principle that certain risk factors like size, value, and
momentum have historically generated excess returns or premiums over the broader market.
Over long periods of time, empirical research has shown that investing in stocks exhibiting these
"risk premia" characteristics has led to outperformance versus a simple market-capitalization
weighted index. Some of the seminal research establishing the importance of these factors
includes:
- Fama and French (1992,1993,1996): Their three-factor model demonstrated that size (market
capitalization) and value (book-to-market ratio) Premiums beyond market returns help explain
stock returns and should be considered systematic risk factors that compensate investors for
bearing additional risk.
- Carhart (1997): Extended the Fama-French model to include the momentum factor, showing
that stocks with positive past performance tend to continue outperforming for a limited period
(the "momentum premium").
- Asness et al. (2013): Comprehensively analyzed size, value, momentum, quality and low risk
factors across 26 developed and emerging markets over 40+ years, providing compelling
evidence of systematic risk premia globally.
Due to this extensive empirical foundation, factors like size, value, and momentum have
become key inputs for portfolio construction and risk modeling. By targeting stocks exhibiting
the characteristics associated with these premiums, investors seek to generate higher returns
over the long run. However, a limitation of early factor models is they typically involved only a
handful of factors and did not account for complex interactions between them.
Advanced Factor Models
More recently, quantitative researchers have developed progressively more sophisticated multi-
factor models that aim to better capture factor dynamics and generate more robust investment
signals. Several of the most advanced approaches include:
- Barra Models: Developed by MSCI Barra, these are multi-factor risk models that employ up to
30 individual factors including exposure to regions, industries, size, value, and quality
characteristics. The interaction effects between factors are explicitly modeled.
- Axioma Models: Axioma introduced a multi-factor model with over 100 covariates designed to
pick up nuanced industry and country factor exposures. It can distinguish factors like
sustainability versus cyclicality that impact stock returns.
- Northfield Models: Northfield factors represent underlying macroeconomic conditions rather
than just stock attributes. Their multi-region International model identifies over 60 factors
including terms of trade, real rates, and market liquidity dynamics.
- Anthic Factor Zoo Models: This AI-driven platform identifies adaptive factors in real-time using
unsupervised machine learning on vast financial datasets. Novel factors well beyond the classic
ones continually emerge without prescribed definitions.
By incorporating a much broader array of explicitly modeled factors and interactions, these
advanced approaches aim to provide a more complete picture of the true drivers of expected
returns. Their factor signals are also intended to be more robust to changing market
environments over time compared to staticfactor models.
Implementing Factor Investing Strategies
Given the depth of research on factor premia and availability of advanced factor models, the
next area of focus is implementing factor-based strategies effectively. Several crucial
considerations include:
1) Factor Tilt vs. Targeted Exposure
Portfolios can take a modest multi-factor "tilt" approach focusing on a handful of factors like
value, quality and momentum. Or pursue targeted factor "budgeting" constraining exposures to
specific factors and neutralizing others. The appropriate strategy depends on objectives,
constraints and risk tolerance.
2) Factor Combination & Interaction
Factors are not independent and their combination matters. Adding factors may improve
diversification, but too many can dilute signals. Interactions must also be modeled, such as
value stocks exhibiting momentum. Engaging a model provider is important for nuanced factor
blending.
3) Rebalancing Frequency
Rebalancing maintains targeted factor exposures and limits deviations from intended style.
However, frequent trading amplifies transaction costs. Balancing expected benefits against
friction, a quarterly or semi-annual rebalancing regime is often employed.
4) Implementation Vehicle
Strategies can utilize individual stocks and ETFs, factor mutual funds or model portfolios, smart
beta/strategic beta ETFs, or segregated accounts/separately managed accounts. Focus of the
mandate and scale/costs informs vehicle selection.
5) Fees & Liquidity
Transaction costs and management fees must be diligently assessed against gross alpha
potential of the strategy. Strategies with frequent rebalancing or narrow holdings increase
trading expenses. Liquidity screening also prevents bid-ask spread issues.
6) Risk Controls and Performance Tracking
Ongoing monitoring of portfolio factor exposures, risk budgets, tracking error and style drift
relative to intended targets is crucial. Performance attribution analysis decomposes returns by
factor, region and stock-level drivers to evaluate strategy success.
In practice, assembling the optimal mix of individual stock positions or investing vehicles, choice
of a suitable factor model, and setting rebalancing policies requires careful planning. But this
groundwork enables systematically capturing factor premiums over the long-term.
Case Study: International Small Cap Value Factor Portfolio
As an example, consider implementing an international small cap value factor tilt strategy as
follows:
- Universe: MSCI World ex-USA Small Cap Index (~1,500 stocks across 23 developed markets
ex-US)
- Factors: Small Cap exposure boost and Value factor screen via book-to-price ratio targeting
stocks in bottom 30%
- Model: Northfield International Multi-Factor risk model identifies exposures to 60+ macro and
company-specific factors
- Constraints: Target +1.5% small cap exposure and +2% value exposure vs benchmark,
neutralize all other factors
- Holdings: Equally weight 250-300 stocks passing factor criteria with liquidity screen
- Rebalancing: Semi-annually in June and December each year
- Vehicle: Separately managed account with low-cost ETF trades
- Performance: Track information ratio vs benchmark and attributes to factors annually
By leveraging a sophisticated multi-factor model, implementing tight style constraints via factor
budgeting, and minimizing costs through a concentrated portfolio and infrequent trading, this
international small value strategy aims to profitably capture premia in a risk-controlled manner
over market cycles. Ongoing governance ensures it adheres to the intended objective function.
Regulatory and Reporting Considerations
As factor investing grows in prominence, regulatory issues also deserve attention. SEC
guidelines including those from Rule 2a-7 on Money Market Funds and Rule 3c-5 on Diversified
Funds define criteria for concentrated, non-diversified or narrowly-focused investment products.
Strategies pursuing highly concentrated factor tilts must carefully consider these rules.
Additionally, various reporting standards exist for communicating factor portfolio compositions
and performance attribution in a transparent manner. The Global Industry Classification
Standard (GICS) is integral to classifying holdings by industry groups. The United Nations
Principles for Responsible Investment (PRI) provide a framework for assessing Environmental,
Social and Corporate Governance (ESG) integration.
Quantitative regulatory filings involving 13F holdings disclosures and Form N-PORT for US-
registered funds demand factor exposures align well-defined investment mandate descriptions.
International standards like the Asia Pacific Fund Classification Framework (FCF) similarly
guide factor portfolio reporting practices globally. Robust compliance programs ensure
strategies operate with integrity.
Conclusion
Over the past few decades, an abundance of empirical research has established the importance
of systematic risk factors like size, value, and momentum for driving equity returns. While early
models successfully captured these classic factors, modern portfolio construction has
progressed towards multi-factor risk modeling, factor combination, and dynamic factor-focused
strategies. Leading implementations utilize advanced multi-factor risk models, target balanced
factor tilts complementing overall objectives, and optimize implementation through careful
portfolio design, trading, and oversight practices. Looking ahead, as factor insights continue
adapting based on big data sources, their full potential for superior risk-adjusted performance
will likely be further unlocked. Factor-based portfolio construction remains a cornerstone
approach for investment managers.
Factor investing has become an increasingly important consideration for investors seeking to
drive excess returns. Rather than solely relying on stock returns or market indices, investors are
turning to factor-based strategies that target specific risk premia. This paper will delve into some
of the leading factor models that have emerged in recent years and how they can be
implemented to construct investment portfolios. It will begin by outlining the rationale behind
factor investing and some of the key factors that are commonly targeted. From there it will
explore advanced factor models and how they can enhance portfolio construction beyond
traditional single or multi-factor approaches. The implementation of such models will also be
discussed, including practical considerations for investors. Overall, this paper aims to provide an
in-depth look at state-of-the-art factor investing techniques and how they can empower investing
decisions.
Rational and Historical Basis for Factor Investing
At its core, factor investing is based on the principle that certain risk factors like size, value, and
momentum have historically generated excess returns or premiums over the broader market.
Over long periods of time, empirical research has shown that investing in stocks exhibiting these
"risk premia" characteristics has led to outperformance versus a simple market-capitalization
weighted index. Some of the seminal research establishing the importance of these factors
includes:
- Fama and French (1992,1993,1996): Their three-factor model demonstrated that size (market
capitalization) and value (book-to-market ratio) Premiums beyond market returns help explain
stock returns and should be considered systematic risk factors that compensate investors for
bearing additional risk.
- Carhart (1997): Extended the Fama-French model to include the momentum factor, showing
that stocks with positive past performance tend to continue outperforming for a limited period
(the "momentum premium").
- Asness et al. (2013): Comprehensively analyzed size, value, momentum, quality and low risk
factors across 26 developed and emerging markets over 40+ years, providing compelling
evidence of systematic risk premia globally.
Due to this extensive empirical foundation, factors like size, value, and momentum have
become key inputs for portfolio construction and risk modeling. By targeting stocks exhibiting
the characteristics associated with these premiums, investors seek to generate higher returns
over the long run. However, a limitation of early factor models is they typically involved only a
handful of factors and did not account for complex interactions between them.
Advanced Factor Models
More recently, quantitative researchers have developed progressively more sophisticated multi-
factor models that aim to better capture factor dynamics and generate more robust investment
signals. Several of the most advanced approaches include:
- Barra Models: Developed by MSCI Barra, these are multi-factor risk models that employ up to
30 individual factors including exposure to regions, industries, size, value, and quality
characteristics. The interaction effects between factors are explicitly modeled.
- Axioma Models: Axioma introduced a multi-factor model with over 100 covariates designed to
pick up nuanced industry and country factor exposures. It can distinguish factors like
sustainability versus cyclicality that impact stock returns.
- Northfield Models: Northfield factors represent underlying macroeconomic conditions rather
than just stock attributes. Their multi-region International model identifies over 60 factors
including terms of trade, real rates, and market liquidity dynamics.
- Anthic Factor Zoo Models: This AI-driven platform identifies adaptive factors in real-time using
unsupervised machine learning on vast financial datasets. Novel factors well beyond the classic
ones continually emerge without prescribed definitions.
By incorporating a much broader array of explicitly modeled factors and interactions, these
advanced approaches aim to provide a more complete picture of the true drivers of expected
returns. Their factor signals are also intended to be more robust to changing market
environments over time compared to staticfactor models.
Implementing Factor Investing Strategies
Given the depth of research on factor premia and availability of advanced factor models, the
next area of focus is implementing factor-based strategies effectively. Several crucial
considerations include:
1) Factor Tilt vs. Targeted Exposure
Portfolios can take a modest multi-factor "tilt" approach focusing on a handful of factors like
value, quality and momentum. Or pursue targeted factor "budgeting" constraining exposures to
specific factors and neutralizing others. The appropriate strategy depends on objectives,
constraints and risk tolerance.
2) Factor Combination & Interaction
Factors are not independent and their combination matters. Adding factors may improve
diversification, but too many can dilute signals. Interactions must also be modeled, such as
value stocks exhibiting momentum. Engaging a model provider is important for nuanced factor
blending.
3) Rebalancing Frequency
Rebalancing maintains targeted factor exposures and limits deviations from intended style.
However, frequent trading amplifies transaction costs. Balancing expected benefits against
friction, a quarterly or semi-annual rebalancing regime is often employed.
4) Implementation Vehicle
Strategies can utilize individual stocks and ETFs, factor mutual funds or model portfolios, smart
beta/strategic beta ETFs, or segregated accounts/separately managed accounts. Focus of the
mandate and scale/costs informs vehicle selection.
5) Fees & Liquidity
Transaction costs and management fees must be diligently assessed against gross alpha
potential of the strategy. Strategies with frequent rebalancing or narrow holdings increase
trading expenses. Liquidity screening also prevents bid-ask spread issues.
6) Risk Controls and Performance Tracking
Ongoing monitoring of portfolio factor exposures, risk budgets, tracking error and style drift
relative to intended targets is crucial. Performance attribution analysis decomposes returns by
factor, region and stock-level drivers to evaluate strategy success.
In practice, assembling the optimal mix of individual stock positions or investing vehicles, choice
of a suitable factor model, and setting rebalancing policies requires careful planning. But this
groundwork enables systematically capturing factor premiums over the long-term.
Case Study: International Small Cap Value Factor Portfolio
As an example, consider implementing an international small cap value factor tilt strategy as
follows:
- Universe: MSCI World ex-USA Small Cap Index (~1,500 stocks across 23 developed markets
ex-US)
- Factors: Small Cap exposure boost and Value factor screen via book-to-price ratio targeting
stocks in bottom 30%
- Model: Northfield International Multi-Factor risk model identifies exposures to 60+ macro and
company-specific factors
- Constraints: Target +1.5% small cap exposure and +2% value exposure vs benchmark,
neutralize all other factors
- Holdings: Equally weight 250-300 stocks passing factor criteria with liquidity screen
- Rebalancing: Semi-annually in June and December each year
- Vehicle: Separately managed account with low-cost ETF trades
- Performance: Track information ratio vs benchmark and attributes to factors annually
By leveraging a sophisticated multi-factor model, implementing tight style constraints via factor
budgeting, and minimizing costs through a concentrated portfolio and infrequent trading, this
international small value strategy aims to profitably capture premia in a risk-controlled manner
over market cycles. Ongoing governance ensures it adheres to the intended objective function.
Regulatory and Reporting Considerations
As factor investing grows in prominence, regulatory issues also deserve attention. SEC
guidelines including those from Rule 2a-7 on Money Market Funds and Rule 3c-5 on Diversified
Funds define criteria for concentrated, non-diversified or narrowly-focused investment products.
Strategies pursuing highly concentrated factor tilts must carefully consider these rules.
Additionally, various reporting standards exist for communicating factor portfolio compositions
and performance attribution in a transparent manner. The Global Industry Classification
Standard (GICS) is integral to classifying holdings by industry groups. The United Nations
Principles for Responsible Investment (PRI) provide a framework for assessing Environmental,
Social and Corporate Governance (ESG) integration.
Quantitative regulatory filings involving 13F holdings disclosures and Form N-PORT for US-
registered funds demand factor exposures align well-defined investment mandate descriptions.
International standards like the Asia Pacific Fund Classification Framework (FCF) similarly
guide factor portfolio reporting practices globally. Robust compliance programs ensure
strategies operate with integrity.
Conclusion
Over the past few decades, an abundance of empirical research has established the importance
of systematic risk factors like size, value, and momentum for driving equity returns. While early
models successfully captured these classic factors, modern portfolio construction has
progressed towards multi-factor risk modeling, factor combination, and dynamic factor-focused
strategies. Leading implementations utilize advanced multi-factor risk models, target balanced
factor tilts complementing overall objectives, and optimize implementation through careful
portfolio design, trading, and oversight practices. Looking ahead, as factor insights continue
adapting based on big data sources, their full potential for superior risk-adjusted performance
will likely be further unlocked. Factor-based portfolio construction remains a cornerstone
approach for investment managers.
Factor investing has become an increasingly important consideration for investors seeking to
drive excess returns. Rather than solely relying on stock returns or market indices, investors are
turning to factor-based strategies that target specific risk premia. This paper will delve into some
of the leading factor models that have emerged in recent years and how they can be
implemented to construct investment portfolios. It will begin by outlining the rationale behind
factor investing and some of the key factors that are commonly targeted. From there it will
explore advanced factor models and how they can enhance portfolio construction beyond
traditional single or multi-factor approaches. The implementation of such models will also be
discussed, including practical considerations for investors. Overall, this paper aims to provide an
in-depth look at state-of-the-art factor investing techniques and how they can empower investing
decisions.
Rational and Historical Basis for Factor Investing
At its core, factor investing is based on the principle that certain risk factors like size, value, and
momentum have historically generated excess returns or premiums over the broader market.
Over long periods of time, empirical research has shown that investing in stocks exhibiting these
"risk premia" characteristics has led to outperformance versus a simple market-capitalization
weighted index. Some of the seminal research establishing the importance of these factors
includes:
- Fama and French (1992,1993,1996): Their three-factor model demonstrated that size (market
capitalization) and value (book-to-market ratio) Premiums beyond market returns help explain
stock returns and should be considered systematic risk factors that compensate investors for
bearing additional risk.
- Carhart (1997): Extended the Fama-French model to include the momentum factor, showing
that stocks with positive past performance tend to continue outperforming for a limited period
(the "momentum premium").
- Asness et al. (2013): Comprehensively analyzed size, value, momentum, quality and low risk
factors across 26 developed and emerging markets over 40+ years, providing compelling
evidence of systematic risk premia globally.
Due to this extensive empirical foundation, factors like size, value, and momentum have
become key inputs for portfolio construction and risk modeling. By targeting stocks exhibiting
the characteristics associated with these premiums, investors seek to generate higher returns
over the long run. However, a limitation of early factor models is they typically involved only a
handful of factors and did not account for complex interactions between them.
Advanced Factor Models
More recently, quantitative researchers have developed progressively more sophisticated multi-
factor models that aim to better capture factor dynamics and generate more robust investment
signals. Several of the most advanced approaches include:
- Barra Models: Developed by MSCI Barra, these are multi-factor risk models that employ up to
30 individual factors including exposure to regions, industries, size, value, and quality
characteristics. The interaction effects between factors are explicitly modeled.
- Axioma Models: Axioma introduced a multi-factor model with over 100 covariates designed to
pick up nuanced industry and country factor exposures. It can distinguish factors like
sustainability versus cyclicality that impact stock returns.
- Northfield Models: Northfield factors represent underlying macroeconomic conditions rather
than just stock attributes. Their multi-region International model identifies over 60 factors
including terms of trade, real rates, and market liquidity dynamics.
- Anthic Factor Zoo Models: This AI-driven platform identifies adaptive factors in real-time using
unsupervised machine learning on vast financial datasets. Novel factors well beyond the classic
ones continually emerge without prescribed definitions.
By incorporating a much broader array of explicitly modeled factors and interactions, these
advanced approaches aim to provide a more complete picture of the true drivers of expected
returns. Their factor signals are also intended to be more robust to changing market
environments over time compared to staticfactor models.
Implementing Factor Investing Strategies
Given the depth of research on factor premia and availability of advanced factor models, the
next area of focus is implementing factor-based strategies effectively. Several crucial
considerations include:
1) Factor Tilt vs. Targeted Exposure
Portfolios can take a modest multi-factor "tilt" approach focusing on a handful of factors like
value, quality and momentum. Or pursue targeted factor "budgeting" constraining exposures to
specific factors and neutralizing others. The appropriate strategy depends on objectives,
constraints and risk tolerance.
2) Factor Combination & Interaction
Factors are not independent and their combination matters. Adding factors may improve
diversification, but too many can dilute signals. Interactions must also be modeled, such as
value stocks exhibiting momentum. Engaging a model provider is important for nuanced factor
blending.
3) Rebalancing Frequency
Rebalancing maintains targeted factor exposures and limits deviations from intended style.
However, frequent trading amplifies transaction costs. Balancing expected benefits against
friction, a quarterly or semi-annual rebalancing regime is often employed.
4) Implementation Vehicle
Strategies can utilize individual stocks and ETFs, factor mutual funds or model portfolios, smart
beta/strategic beta ETFs, or segregated accounts/separately managed accounts. Focus of the
mandate and scale/costs informs vehicle selection.
5) Fees & Liquidity
Transaction costs and management fees must be diligently assessed against gross alpha
potential of the strategy. Strategies with frequent rebalancing or narrow holdings increase
trading expenses. Liquidity screening also prevents bid-ask spread issues.
6) Risk Controls and Performance Tracking
Ongoing monitoring of portfolio factor exposures, risk budgets, tracking error and style drift
relative to intended targets is crucial. Performance attribution analysis decomposes returns by
factor, region and stock-level drivers to evaluate strategy success.
In practice, assembling the optimal mix of individual stock positions or investing vehicles, choice
of a suitable factor model, and setting rebalancing policies requires careful planning. But this
groundwork enables systematically capturing factor premiums over the long-term.
Case Study: International Small Cap Value Factor Portfolio
As an example, consider implementing an international small cap value factor tilt strategy as
follows:
- Universe: MSCI World ex-USA Small Cap Index (~1,500 stocks across 23 developed markets
ex-US)
- Factors: Small Cap exposure boost and Value factor screen via book-to-price ratio targeting
stocks in bottom 30%
- Model: Northfield International Multi-Factor risk model identifies exposures to 60+ macro and
company-specific factors
- Constraints: Target +1.5% small cap exposure and +2% value exposure vs benchmark,
neutralize all other factors
- Holdings: Equally weight 250-300 stocks passing factor criteria with liquidity screen
- Rebalancing: Semi-annually in June and December each year
- Vehicle: Separately managed account with low-cost ETF trades
- Performance: Track information ratio vs benchmark and attributes to factors annually
By leveraging a sophisticated multi-factor model, implementing tight style constraints via factor
budgeting, and minimizing costs through a concentrated portfolio and infrequent trading, this
international small value strategy aims to profitably capture premia in a risk-controlled manner
over market cycles. Ongoing governance ensures it adheres to the intended objective function.
Regulatory and Reporting Considerations
As factor investing grows in prominence, regulatory issues also deserve attention. SEC
guidelines including those from Rule 2a-7 on Money Market Funds and Rule 3c-5 on Diversified
Funds define criteria for concentrated, non-diversified or narrowly-focused investment products.
Strategies pursuing highly concentrated factor tilts must carefully consider these rules.
Additionally, various reporting standards exist for communicating factor portfolio compositions
and performance attribution in a transparent manner. The Global Industry Classification
Standard (GICS) is integral to classifying holdings by industry groups. The United Nations
Principles for Responsible Investment (PRI) provide a framework for assessing Environmental,
Social and Corporate Governance (ESG) integration.
Quantitative regulatory filings involving 13F holdings disclosures and Form N-PORT for US-
registered funds demand factor exposures align well-defined investment mandate descriptions.
International standards like the Asia Pacific Fund Classification Framework (FCF) similarly
guide factor portfolio reporting practices globally. Robust compliance programs ensure
strategies operate with integrity.
Conclusion
Over the past few decades, an abundance of empirical research has established the importance
of systematic risk factors like size, value, and momentum for driving equity returns. While early
models successfully captured these classic factors, modern portfolio construction has
progressed towards multi-factor risk modeling, factor combination, and dynamic factor-focused
strategies. Leading implementations utilize advanced multi-factor risk models, target balanced
factor tilts complementing overall objectives, and optimize implementation through careful
portfolio design, trading, and oversight practices. Looking ahead, as factor insights continue
adapting based on big data sources, their full potential for superior risk-adjusted performance
will likely be further unlocked. Factor-based portfolio construction remains a cornerstone
approach for investment managers.
Factor investing has become an increasingly important consideration for investors seeking to
drive excess returns. Rather than solely relying on stock returns or market indices, investors are
turning to factor-based strategies that target specific risk premia. This paper will delve into some
of the leading factor models that have emerged in recent years and how they can be
implemented to construct investment portfolios. It will begin by outlining the rationale behind
factor investing and some of the key factors that are commonly targeted. From there it will
explore advanced factor models and how they can enhance portfolio construction beyond
traditional single or multi-factor approaches. The implementation of such models will also be
discussed, including practical considerations for investors. Overall, this paper aims to provide an
in-depth look at state-of-the-art factor investing techniques and how they can empower investing
decisions.
Rational and Historical Basis for Factor Investing
At its core, factor investing is based on the principle that certain risk factors like size, value, and
momentum have historically generated excess returns or premiums over the broader market.
Over long periods of time, empirical research has shown that investing in stocks exhibiting these
"risk premia" characteristics has led to outperformance versus a simple market-capitalization
weighted index. Some of the seminal research establishing the importance of these factors
includes:
- Fama and French (1992,1993,1996): Their three-factor model demonstrated that size (market
capitalization) and value (book-to-market ratio) Premiums beyond market returns help explain
stock returns and should be considered systematic risk factors that compensate investors for
bearing additional risk.
- Carhart (1997): Extended the Fama-French model to include the momentum factor, showing
that stocks with positive past performance tend to continue outperforming for a limited period
(the "momentum premium").
- Asness et al. (2013): Comprehensively analyzed size, value, momentum, quality and low risk
factors across 26 developed and emerging markets over 40+ years, providing compelling
evidence of systematic risk premia globally.
Due to this extensive empirical foundation, factors like size, value, and momentum have
become key inputs for portfolio construction and risk modeling. By targeting stocks exhibiting
the characteristics associated with these premiums, investors seek to generate higher returns
over the long run. However, a limitation of early factor models is they typically involved only a
handful of factors and did not account for complex interactions between them.
Advanced Factor Models
More recently, quantitative researchers have developed progressively more sophisticated multi-
factor models that aim to better capture factor dynamics and generate more robust investment
signals. Several of the most advanced approaches include:
- Barra Models: Developed by MSCI Barra, these are multi-factor risk models that employ up to
30 individual factors including exposure to regions, industries, size, value, and quality
characteristics. The interaction effects between factors are explicitly modeled.
- Axioma Models: Axioma introduced a multi-factor model with over 100 covariates designed to
pick up nuanced industry and country factor exposures. It can distinguish factors like
sustainability versus cyclicality that impact stock returns.
- Northfield Models: Northfield factors represent underlying macroeconomic conditions rather
than just stock attributes. Their multi-region International model identifies over 60 factors
including terms of trade, real rates, and market liquidity dynamics.
- Anthic Factor Zoo Models: This AI-driven platform identifies adaptive factors in real-time using
unsupervised machine learning on vast financial datasets. Novel factors well beyond the classic
ones continually emerge without prescribed definitions.
By incorporating a much broader array of explicitly modeled factors and interactions, these
advanced approaches aim to provide a more complete picture of the true drivers of expected
returns. Their factor signals are also intended to be more robust to changing market
environments over time compared to staticfactor models.
Implementing Factor Investing Strategies
Given the depth of research on factor premia and availability of advanced factor models, the
next area of focus is implementing factor-based strategies effectively. Several crucial
considerations include:
1) Factor Tilt vs. Targeted Exposure
Portfolios can take a modest multi-factor "tilt" approach focusing on a handful of factors like
value, quality and momentum. Or pursue targeted factor "budgeting" constraining exposures to
specific factors and neutralizing others. The appropriate strategy depends on objectives,
constraints and risk tolerance.
2) Factor Combination & Interaction
Factors are not independent and their combination matters. Adding factors may improve
diversification, but too many can dilute signals. Interactions must also be modeled, such as
value stocks exhibiting momentum. Engaging a model provider is important for nuanced factor
blending.
3) Rebalancing Frequency
Rebalancing maintains targeted factor exposures and limits deviations from intended style.
However, frequent trading amplifies transaction costs. Balancing expected benefits against
friction, a quarterly or semi-annual rebalancing regime is often employed.
4) Implementation Vehicle
Strategies can utilize individual stocks and ETFs, factor mutual funds or model portfolios, smart
beta/strategic beta ETFs, or segregated accounts/separately managed accounts. Focus of the
mandate and scale/costs informs vehicle selection.
5) Fees & Liquidity
Transaction costs and management fees must be diligently assessed against gross alpha
potential of the strategy. Strategies with frequent rebalancing or narrow holdings increase
trading expenses. Liquidity screening also prevents bid-ask spread issues.
6) Risk Controls and Performance Tracking
Ongoing monitoring of portfolio factor exposures, risk budgets, tracking error and style drift
relative to intended targets is crucial. Performance attribution analysis decomposes returns by
factor, region and stock-level drivers to evaluate strategy success.
In practice, assembling the optimal mix of individual stock positions or investing vehicles, choice
of a suitable factor model, and setting rebalancing policies requires careful planning. But this
groundwork enables systematically capturing factor premiums over the long-term.
Case Study: International Small Cap Value Factor Portfolio
As an example, consider implementing an international small cap value factor tilt strategy as
follows:
- Universe: MSCI World ex-USA Small Cap Index (~1,500 stocks across 23 developed markets
ex-US)
- Factors: Small Cap exposure boost and Value factor screen via book-to-price ratio targeting
stocks in bottom 30%
- Model: Northfield International Multi-Factor risk model identifies exposures to 60+ macro and
company-specific factors
- Constraints: Target +1.5% small cap exposure and +2% value exposure vs benchmark,
neutralize all other factors
- Holdings: Equally weight 250-300 stocks passing factor criteria with liquidity screen
- Rebalancing: Semi-annually in June and December each year
- Vehicle: Separately managed account with low-cost ETF trades
- Performance: Track information ratio vs benchmark and attributes to factors annually
By leveraging a sophisticated multi-factor model, implementing tight style constraints via factor
budgeting, and minimizing costs through a concentrated portfolio and infrequent trading, this
international small value strategy aims to profitably capture premia in a risk-controlled manner
over market cycles. Ongoing governance ensures it adheres to the intended objective function.
Regulatory and Reporting Considerations
As factor investing grows in prominence, regulatory issues also deserve attention. SEC
guidelines including those from Rule 2a-7 on Money Market Funds and Rule 3c-5 on Diversified
Funds define criteria for concentrated, non-diversified or narrowly-focused investment products.
Strategies pursuing highly concentrated factor tilts must carefully consider these rules.
Additionally, various reporting standards exist for communicating factor portfolio compositions
and performance attribution in a transparent manner. The Global Industry Classification
Standard (GICS) is integral to classifying holdings by industry groups. The United Nations
Principles for Responsible Investment (PRI) provide a framework for assessing Environmental,
Social and Corporate Governance (ESG) integration.
Quantitative regulatory filings involving 13F holdings disclosures and Form N-PORT for US-
registered funds demand factor exposures align well-defined investment mandate descriptions.
International standards like the Asia Pacific Fund Classification Framework (FCF) similarly
guide factor portfolio reporting practices globally. Robust compliance programs ensure
strategies operate with integrity.
Conclusion
Over the past few decades, an abundance of empirical research has established the importance
of systematic risk factors like size, value, and momentum for driving equity returns. While early
models successfully captured these classic factors, modern portfolio construction has
progressed towards multi-factor risk modeling, factor combination, and dynamic factor-focused
strategies. Leading implementations utilize advanced multi-factor risk models, target balanced
factor tilts complementing overall objectives, and optimize implementation through careful
portfolio design, trading, and oversight practices. Looking ahead, as factor insights continue
adapting based on big data sources, their full potential for superior risk-adjusted performance
will likely be further unlocked. Factor-based portfolio construction remains a cornerstone
approach for investment managers.
Factor investing has become an increasingly important consideration for investors seeking to
drive excess returns. Rather than solely relying on stock returns or market indices, investors are
turning to factor-based strategies that target specific risk premia. This paper will delve into some
of the leading factor models that have emerged in recent years and how they can be
implemented to construct investment portfolios. It will begin by outlining the rationale behind
factor investing and some of the key factors that are commonly targeted. From there it will
explore advanced factor models and how they can enhance portfolio construction beyond
traditional single or multi-factor approaches. The implementation of such models will also be
discussed, including practical considerations for investors. Overall, this paper aims to provide an
in-depth look at state-of-the-art factor investing techniques and how they can empower investing
decisions.
Rational and Historical Basis for Factor Investing
At its core, factor investing is based on the principle that certain risk factors like size, value, and
momentum have historically generated excess returns or premiums over the broader market.
Over long periods of time, empirical research has shown that investing in stocks exhibiting these
"risk premia" characteristics has led to outperformance versus a simple market-capitalization
weighted index. Some of the seminal research establishing the importance of these factors
includes:
- Fama and French (1992,1993,1996): Their three-factor model demonstrated that size (market
capitalization) and value (book-to-market ratio) Premiums beyond market returns help explain
stock returns and should be considered systematic risk factors that compensate investors for
bearing additional risk.
- Carhart (1997): Extended the Fama-French model to include the momentum factor, showing
that stocks with positive past performance tend to continue outperforming for a limited period
(the "momentum premium").
- Asness et al. (2013): Comprehensively analyzed size, value, momentum, quality and low risk
factors across 26 developed and emerging markets over 40+ years, providing compelling
evidence of systematic risk premia globally.
Due to this extensive empirical foundation, factors like size, value, and momentum have
become key inputs for portfolio construction and risk modeling. By targeting stocks exhibiting
the characteristics associated with these premiums, investors seek to generate higher returns
over the long run. However, a limitation of early factor models is they typically involved only a
handful of factors and did not account for complex interactions between them.
Advanced Factor Models
More recently, quantitative researchers have developed progressively more sophisticated multi-
factor models that aim to better capture factor dynamics and generate more robust investment
signals. Several of the most advanced approaches include:
- Barra Models: Developed by MSCI Barra, these are multi-factor risk models that employ up to
30 individual factors including exposure to regions, industries, size, value, and quality
characteristics. The interaction effects between factors are explicitly modeled.
- Axioma Models: Axioma introduced a multi-factor model with over 100 covariates designed to
pick up nuanced industry and country factor exposures. It can distinguish factors like
sustainability versus cyclicality that impact stock returns.
- Northfield Models: Northfield factors represent underlying macroeconomic conditions rather
than just stock attributes. Their multi-region International model identifies over 60 factors
including terms of trade, real rates, and market liquidity dynamics.
- Anthic Factor Zoo Models: This AI-driven platform identifies adaptive factors in real-time using
unsupervised machine learning on vast financial datasets. Novel factors well beyond the classic
ones continually emerge without prescribed definitions.
By incorporating a much broader array of explicitly modeled factors and interactions, these
advanced approaches aim to provide a more complete picture of the true drivers of expected
returns. Their factor signals are also intended to be more robust to changing market
environments over time compared to staticfactor models.
Implementing Factor Investing Strategies
Given the depth of research on factor premia and availability of advanced factor models, the
next area of focus is implementing factor-based strategies effectively. Several crucial
considerations include:
1) Factor Tilt vs. Targeted Exposure
Portfolios can take a modest multi-factor "tilt" approach focusing on a handful of factors like
value, quality and momentum. Or pursue targeted factor "budgeting" constraining exposures to
specific factors and neutralizing others. The appropriate strategy depends on objectives,
constraints and risk tolerance.
2) Factor Combination & Interaction
Factors are not independent and their combination matters. Adding factors may improve
diversification, but too many can dilute signals. Interactions must also be modeled, such as
value stocks exhibiting momentum. Engaging a model provider is important for nuanced factor
blending.
3) Rebalancing Frequency
Rebalancing maintains targeted factor exposures and limits deviations from intended style.
However, frequent trading amplifies transaction costs. Balancing expected benefits against
friction, a quarterly or semi-annual rebalancing regime is often employed.
4) Implementation Vehicle
Strategies can utilize individual stocks and ETFs, factor mutual funds or model portfolios, smart
beta/strategic beta ETFs, or segregated accounts/separately managed accounts. Focus of the
mandate and scale/costs informs vehicle selection.
5) Fees & Liquidity
Transaction costs and management fees must be diligently assessed against gross alpha
potential of the strategy. Strategies with frequent rebalancing or narrow holdings increase
trading expenses. Liquidity screening also prevents bid-ask spread issues.
6) Risk Controls and Performance Tracking
Ongoing monitoring of portfolio factor exposures, risk budgets, tracking error and style drift
relative to intended targets is crucial. Performance attribution analysis decomposes returns by
factor, region and stock-level drivers to evaluate strategy success.
In practice, assembling the optimal mix of individual stock positions or investing vehicles, choice
of a suitable factor model, and setting rebalancing policies requires careful planning. But this
groundwork enables systematically capturing factor premiums over the long-term.
Case Study: International Small Cap Value Factor Portfolio
As an example, consider implementing an international small cap value factor tilt strategy as
follows:
- Universe: MSCI World ex-USA Small Cap Index (~1,500 stocks across 23 developed markets
ex-US)
- Factors: Small Cap exposure boost and Value factor screen via book-to-price ratio targeting
stocks in bottom 30%
- Model: Northfield International Multi-Factor risk model identifies exposures to 60+ macro and
company-specific factors
- Constraints: Target +1.5% small cap exposure and +2% value exposure vs benchmark,
neutralize all other factors
- Holdings: Equally weight 250-300 stocks passing factor criteria with liquidity screen
- Rebalancing: Semi-annually in June and December each year
- Vehicle: Separately managed account with low-cost ETF trades
- Performance: Track information ratio vs benchmark and attributes to factors annually
By leveraging a sophisticated multi-factor model, implementing tight style constraints via factor
budgeting, and minimizing costs through a concentrated portfolio and infrequent trading, this
international small value strategy aims to profitably capture premia in a risk-controlled manner
over market cycles. Ongoing governance ensures it adheres to the intended objective function.
Regulatory and Reporting Considerations
As factor investing grows in prominence, regulatory issues also deserve attention. SEC
guidelines including those from Rule 2a-7 on Money Market Funds and Rule 3c-5 on Diversified
Funds define criteria for concentrated, non-diversified or narrowly-focused investment products.
Strategies pursuing highly concentrated factor tilts must carefully consider these rules.
Additionally, various reporting standards exist for communicating factor portfolio compositions
and performance attribution in a transparent manner. The Global Industry Classification
Standard (GICS) is integral to classifying holdings by industry groups. The United Nations
Principles for Responsible Investment (PRI) provide a framework for assessing Environmental,
Social and Corporate Governance (ESG) integration.
Quantitative regulatory filings involving 13F holdings disclosures and Form N-PORT for US-
registered funds demand factor exposures align well-defined investment mandate descriptions.
International standards like the Asia Pacific Fund Classification Framework (FCF) similarly
guide factor portfolio reporting practices globally. Robust compliance programs ensure
strategies operate with integrity.
Conclusion
Over the past few decades, an abundance of empirical research has established the importance
of systematic risk factors like size, value, and momentum for driving equity returns. While early
models successfully captured these classic factors, modern portfolio construction has
progressed towards multi-factor risk modeling, factor combination, and dynamic factor-focused
strategies. Leading implementations utilize advanced multi-factor risk models, target balanced
factor tilts complementing overall objectives, and optimize implementation through careful
portfolio design, trading, and oversight practices. Looking ahead, as factor insights continue
adapting based on big data sources, their full potential for superior risk-adjusted performance
will likely be further unlocked. Factor-based portfolio construction remains a cornerstone
approach for investment managers.
Factor investing has become an increasingly important consideration for investors seeking to
drive excess returns. Rather than solely relying on stock returns or market indices, investors are
turning to factor-based strategies that target specific risk premia. This paper will delve into some
of the leading factor models that have emerged in recent years and how they can be
implemented to construct investment portfolios. It will begin by outlining the rationale behind
factor investing and some of the key factors that are commonly targeted. From there it will
explore advanced factor models and how they can enhance portfolio construction beyond
traditional single or multi-factor approaches. The implementation of such models will also be
discussed, including practical considerations for investors. Overall, this paper aims to provide an
in-depth look at state-of-the-art factor investing techniques and how they can empower investing
decisions.
Rational and Historical Basis for Factor Investing
At its core, factor investing is based on the principle that certain risk factors like size, value, and
momentum have historically generated excess returns or premiums over the broader market.
Over long periods of time, empirical research has shown that investing in stocks exhibiting these
"risk premia" characteristics has led to outperformance versus a simple market-capitalization
weighted index. Some of the seminal research establishing the importance of these factors
includes:
- Fama and French (1992,1993,1996): Their three-factor model demonstrated that size (market
capitalization) and value (book-to-market ratio) Premiums beyond market returns help explain
stock returns and should be considered systematic risk factors that compensate investors for
bearing additional risk.
- Carhart (1997): Extended the Fama-French model to include the momentum factor, showing
that stocks with positive past performance tend to continue outperforming for a limited period
(the "momentum premium").
- Asness et al. (2013): Comprehensively analyzed size, value, momentum, quality and low risk
factors across 26 developed and emerging markets over 40+ years, providing compelling
evidence of systematic risk premia globally.
Due to this extensive empirical foundation, factors like size, value, and momentum have
become key inputs for portfolio construction and risk modeling. By targeting stocks exhibiting
the characteristics associated with these premiums, investors seek to generate higher returns
over the long run. However, a limitation of early factor models is they typically involved only a
handful of factors and did not account for complex interactions between them.
Advanced Factor Models
More recently, quantitative researchers have developed progressively more sophisticated multi-
factor models that aim to better capture factor dynamics and generate more robust investment
signals. Several of the most advanced approaches include:
- Barra Models: Developed by MSCI Barra, these are multi-factor risk models that employ up to
30 individual factors including exposure to regions, industries, size, value, and quality
characteristics. The interaction effects between factors are explicitly modeled.
- Axioma Models: Axioma introduced a multi-factor model with over 100 covariates designed to
pick up nuanced industry and country factor exposures. It can distinguish factors like
sustainability versus cyclicality that impact stock returns.
- Northfield Models: Northfield factors represent underlying macroeconomic conditions rather
than just stock attributes. Their multi-region International model identifies over 60 factors
including terms of trade, real rates, and market liquidity dynamics.
- Anthic Factor Zoo Models: This AI-driven platform identifies adaptive factors in real-time using
unsupervised machine learning on vast financial datasets. Novel factors well beyond the classic
ones continually emerge without prescribed definitions.
By incorporating a much broader array of explicitly modeled factors and interactions, these
advanced approaches aim to provide a more complete picture of the true drivers of expected
returns. Their factor signals are also intended to be more robust to changing market
environments over time compared to staticfactor models.
Implementing Factor Investing Strategies
Given the depth of research on factor premia and availability of advanced factor models, the
next area of focus is implementing factor-based strategies effectively. Several crucial
considerations include:
1) Factor Tilt vs. Targeted Exposure
Portfolios can take a modest multi-factor "tilt" approach focusing on a handful of factors like
value, quality and momentum. Or pursue targeted factor "budgeting" constraining exposures to
specific factors and neutralizing others. The appropriate strategy depends on objectives,
constraints and risk tolerance.
2) Factor Combination & Interaction
Factors are not independent and their combination matters. Adding factors may improve
diversification, but too many can dilute signals. Interactions must also be modeled, such as
value stocks exhibiting momentum. Engaging a model provider is important for nuanced factor
blending.
3) Rebalancing Frequency
Rebalancing maintains targeted factor exposures and limits deviations from intended style.
However, frequent trading amplifies transaction costs. Balancing expected benefits against
friction, a quarterly or semi-annual rebalancing regime is often employed.
4) Implementation Vehicle
Strategies can utilize individual stocks and ETFs, factor mutual funds or model portfolios, smart
beta/strategic beta ETFs, or segregated accounts/separately managed accounts. Focus of the
mandate and scale/costs informs vehicle selection.
5) Fees & Liquidity
Transaction costs and management fees must be diligently assessed against gross alpha
potential of the strategy. Strategies with frequent rebalancing or narrow holdings increase
trading expenses. Liquidity screening also prevents bid-ask spread issues.
6) Risk Controls and Performance Tracking
Ongoing monitoring of portfolio factor exposures, risk budgets, tracking error and style drift
relative to intended targets is crucial. Performance attribution analysis decomposes returns by
factor, region and stock-level drivers to evaluate strategy success.
In practice, assembling the optimal mix of individual stock positions or investing vehicles, choice
of a suitable factor model, and setting rebalancing policies requires careful planning. But this
groundwork enables systematically capturing factor premiums over the long-term.
Case Study: International Small Cap Value Factor Portfolio
As an example, consider implementing an international small cap value factor tilt strategy as
follows:
- Universe: MSCI World ex-USA Small Cap Index (~1,500 stocks across 23 developed markets
ex-US)
- Factors: Small Cap exposure boost and Value factor screen via book-to-price ratio targeting
stocks in bottom 30%
- Model: Northfield International Multi-Factor risk model identifies exposures to 60+ macro and
company-specific factors
- Constraints: Target +1.5% small cap exposure and +2% value exposure vs benchmark,
neutralize all other factors
- Holdings: Equally weight 250-300 stocks passing factor criteria with liquidity screen
- Rebalancing: Semi-annually in June and December each year
- Vehicle: Separately managed account with low-cost ETF trades
- Performance: Track information ratio vs benchmark and attributes to factors annually
By leveraging a sophisticated multi-factor model, implementing tight style constraints via factor
budgeting, and minimizing costs through a concentrated portfolio and infrequent trading, this
international small value strategy aims to profitably capture premia in a risk-controlled manner
over market cycles. Ongoing governance ensures it adheres to the intended objective function.
Regulatory and Reporting Considerations
As factor investing grows in prominence, regulatory issues also deserve attention. SEC
guidelines including those from Rule 2a-7 on Money Market Funds and Rule 3c-5 on Diversified
Funds define criteria for concentrated, non-diversified or narrowly-focused investment products.
Strategies pursuing highly concentrated factor tilts must carefully consider these rules.
Additionally, various reporting standards exist for communicating factor portfolio compositions
and performance attribution in a transparent manner. The Global Industry Classification
Standard (GICS) is integral to classifying holdings by industry groups. The United Nations
Principles for Responsible Investment (PRI) provide a framework for assessing Environmental,
Social and Corporate Governance (ESG) integration.
Quantitative regulatory filings involving 13F holdings disclosures and Form N-PORT for US-
registered funds demand factor exposures align well-defined investment mandate descriptions.
International standards like the Asia Pacific Fund Classification Framework (FCF) similarly
guide factor portfolio reporting practices globally. Robust compliance programs ensure
strategies operate with integrity.
Conclusion
Over the past few decades, an abundance of empirical research has established the importance
of systematic risk factors like size, value, and momentum for driving equity returns. While early
models successfully captured these classic factors, modern portfolio construction has
progressed towards multi-factor risk modeling, factor combination, and dynamic factor-focused
strategies. Leading implementations utilize advanced multi-factor risk models, target balanced
factor tilts complementing overall objectives, and optimize implementation through careful
portfolio design, trading, and oversight practices. Looking ahead, as factor insights continue
adapting based on big data sources, their full potential for superior risk-adjusted performance
will likely be further unlocked. Factor-based portfolio construction remains a cornerstone
approach for investment managers.
