economics 2 part
WilSte78
peerreview2.pdf
Journal of Accounting, Auditing & Finance
2022, Vol. 37(1) 39–76 �The Author(s) 2018
Article reuse guidelines: sagepub.com/journals-permissions DOI: 10.1177/0148558X18781144
journals.sagepub.com/home/JAF
Controlled Currency Regime and Pricing of Exchange Rate Risk: Evidence From China
Xiuping Hua1, Wei Huang2, and Ying Jiang1
Abstract
Extant research finds that exchange rate pegs do little to reduce firms’ exposure to exchange rate risk in emerging markets. We study whether exchange rate risk exposures under a pegged/controlled floating currency regime can be priced in asset returns using unique data on exchange rate regime changes in China. As the only currency in the basket of International Monetary Fund’s (IMF) Special Drawing Rights that is not fully market- driven for its exchange rate formation, the Chinese RMB (renminbi) offers an ideal case to study the pricing of exchange rate risk in pegged and controlled floating regimes. We find a negative stock market reaction to the announcement of changes from a pegged exchange rate to a controlled floating regime, suggesting a negative stock valuation effect of expecta- tion of increased exchange rate exposure to currency appreciation. Under the managed floating exchange rate regime, the volatilities of accounting performance measures are sig- nificantly greater for the companies with stronger exposure to foreign-exchange rate risk. More importantly, foreign-exchange rate risk is priced in equity returns under the con- trolled floating regime but not under the strictly pegged regime. Furthermore, the pricing is stronger for firms with higher sensitivity to economic channels transmitting the exchange rate risk exposure to risk premiums. Overall, our results suggest that currency hedging for a multinational corporation with foreign-exchange exposure is warranted under a con- trolled floating currency regime.
Keywords
exchange rate risk, asset returns, equity risk premium, foreign-exchange rate risk hedging, Chinese currency risk
Introduction
Recent literature has examined the potential link between currency arrangements and the
extent of firm level exchange rate exposure. Among others, Parsley and Popper (2006)
examined the foreign-exchange exposure of firms with and without exchange rate pegs.
1Nottingham University Business School China, The University of Nottingham Ningbo China, China 2University of Hawaii at Manoa, Honolulu, USA
Corresponding Author:
Wei Huang, Shidler College of Business, University of Hawaii at Manoa, 2404 Maile Way, Honolulu, HI 96822, USA.
Email: [email protected]
Article
They found that firms in Asian-Pacific countries had more widespread exposure to foreign-
exchange rate risk than firms in western industrialized countries, particularly to USD fluc-
tuations. Interestingly, their results suggest that exchange rate pegs do not diminish the
widespread exposure to exchange rate risk. By contrast, other studies, such as Jorion
(1990), Bodnar and Gentry (1993), and Muller and Verschoor (2006), found exchange rate
exposures among firms in the United States and other industrialized countries are
moderate.
In view of the significant foreign-exchange risk exposure under a pegged system in
emerging markets, an important question that follows is whether the foreign-exchange rate
exposures under a pegged currency regime can be priced in asset returns. As stated in the
conclusion of Parsley and Popper (2006), this issue matters because it is related to whether
firms under a pegged currency regime should care about their exchange rate exposure and
hedge foreign-exchange rate risk. The issue has become more important following extraor-
dinary events such as the Asian financial crisis that show how exchange rate crises can gen-
erate widespread impact beyond currency markets. Indeed, fixed or inflexible exchange
regimes may contribute to stronger and more rapid domestic credit growth than a flexible
exchange regime, especially when there is large capital inflow and foreign-exchange inter-
vention (Reinhart & Montiel, 2001).
In this article, we explore China’s unique currency arrangement to examine whether and
to what extent the exchange rate risk exposure may be priced in asset returns under a con-
trolled currency regime. After more than a decade of pegging the renminbi (RMB) to the
USD, China adopted a controlled floating regime in 2005, setting its target for the RMB
based on a ‘‘reference basket’’ of currencies. Subsequently, during a 2-year period which
included the global financial crisis of 2008, China effectively pegged its currency against
the USD. In recent years, China has adopted a tightly controlled currency regime and has
kept the Yuan steady against the dollar by setting a midpoint for the value of the USD. In
daily trading, the Yuan is allowed to move 2% above or below that midpoint, which is
called the daily fixing. Despite its tight control mechanism, Chinese currency regime offers
an optimal setting to study the pricing of exchange rate risk under a pegged/controlled
floating regime for several reasons.
First, the regime change from pegging to the USD to a controlled floating that pegs to a
basket of currencies offers an experimental data to examine if the pricing of the exchange
rate exposure is different between strictly pegging and controlled floating regime. Under
the controlled floating regime, although the central bank has implemented a ‘‘daily fixing’’
value for the USD, it does not as regularly influence the exchange rate against other foreign
currencies, allowing the Yuan to follow the dollar’s rise against other currencies in the
basket such as the Euro. Therefore, it is likely that the exposure to exchange rate risk under
two different regimes may have different impact on asset return.
Second, China has robust economic channels that can transmit exchange rate risk expo-
sure to equity market risk premiums. This includes active international trades, cross-border
capital flows, and frequent domestic credit expansion. Third, there are off-shore markets
for the Chinese Yuan. The currency, both spot and derivatives, are traded and on the clear-
ing centers in Hong Kong, London, Frankfurt, and Singapore. The market-driven ‘‘off-
shore’’ exchange rate often deviates from the controlled ‘‘on-shore’’ exchange rate set by
the central bank, thus creating market-driven impact on the otherwise government con-
trolled exchange rate determination process.
Finally, the Chinese Yuan has risen as a major currency for international payments.1 On
November 30, 2015, International Monetary Fund (IMF) announced that it would formally
40 Journal of Accounting, Auditing & Finance
include the RMB as one of the five currencies in the basket of IMF’s special drawing
rights (SDR), effective on October 1, 2016. IMF set the weight for the RMB at 10.92%,
which is behind the weight for the USD (41.73%) and the Euro (30.93%), but is ahead of
that for Japanese Yen (8.33%) and the British pound (8.09%). The IMF’s move will help
the country’s market-oriented reform in its exchange rate formation and ultimately make
the RMB one of the major currencies for international reserve and settlement. Overall,
China’s currency regime offers a unique opportunity to study whether the exchange rate
risk can be priced under a pegged or a controlled floating currency regime. Other pegged
currencies, however, do not offer such an opportunity because of less robust transmission
channels or lack of managed floating mechanism.2
The pricing of exchange rate risk has attracted significant attention in academic
research. International capital asset-pricing theory suggests that the covariance of asset
returns with currency returns should be a priced factor in a world where purchasing power
parity does not hold (Adler & Dumas, 1983). Currency risk has typically been included
among the hypothesized factors in international asset pricing models (e.g., Solnik, 1974,
1997; Stulz, 1981a, 1981b). Studies using multicountry joint test have shown evidence of
pricing of exchange rate risk in asset returns. For example, using a three-factor model that
includes world, country, and exchange risk factors, Choi and Rajan (1997) found that
exchange risk is a significant factor affecting asset returns in addition to the domestic and
world market factors in the seven major industrialized markets, notwithstanding partial
market segmentation and local country factor. Similarly, Dumas and Solnik (1995) and De
Santis and Gérard (1998) found evidence supporting a specification of the international
Capital Asset Pricing Model (CAPM) that includes both market risk and foreign-exchange
risk for major developed markets. Carrieri and Majerbi (2006) conducted a multinational
asset pricing test for emerging markets and find significant unconditional risk premium for
exchange risk.
One limitation in interpreting the multinational asset pricing test is that it is a joint test
of market segmentation and exchange risk pricing. Differences in currency regimes across
countries further confound the interpretation of joint-test results. Considering that the factor
structure of asset returns is generally heterogeneous internationally (Choi & Rajan, 1997)
and currency risk premiums vary over time and across markets, researchers also stress the
importance of a single country test. Studies using a single country test show less consistent
results. For example, Jorion (1990, 1991) found that although U.S. equity values react to
fluctuations of the trade-weighted value of the dollar, the risk premium associated with for-
eign currency exposure is insignificant in the U.S. stock market and therefore active hed-
ging should not affect the cost of capital. Kolari, Moorman, and Sorescu (2008) showed
that foreign-exchange risk is priced in the cross-section of U.S. stock returns. However,
contrary to the multinational tests, they find that stocks most sensitive to foreign-exchange
risk (in absolute value) command lower returns, implying a nonlinear, negative premium
for foreign-exchange risk in the U.S. market.
In this article, we examine the pricing of exchange rate risk under a pegged/controlled
floating currency regime in an emerging market setting.3 Given the findings in prior studies
that firms in emerging markets with pegged currency regime experience significant
foreign-exchange risk exposure, we aim to understand whether and how exchange rate risks
under a pegged as well as controlled floating regime are priced in asset returns and as a
result whether such exchange rate risk exposure should be hedged.
Our findings can be summarized as follows. First, there was a negative stock market
reaction to the announcement of changes from a pegged exchange rate to a controlled
Hua et al. 41
floating regime that subsequently led to waves of currency appreciation. This suggests that
on average there is a negative stock valuation effect of expectation of increased exchange
rate risk exposure. By contrast, Chinese shares listed on the Hong Kong stock exchange do
not show the same effect due to Hong Kong’s pegged currency regime. Second, the volati-
lities of earnings, return on asset (ROA) and price-earnings ratios (P/E ratio) are signifi-
cantly greater for ‘‘export’’ companies than for nonexport companies after the exchange
rate reform in 2005. Third, results from asset pricing tests indicate that foreign-exchange
rate risk is not priced in stock returns under the strictly pegged exchange rate regime, but it
is priced under a controlled floating regime for the period after 2005. In particular, for the
sample period after 2005 reform, the asset pricing models that include exchange rate risk
factor show considerable reductions in asset pricing errors, in the form of decreases in the
absolute and squared alphas of various asset pricing models. Fourth, the pricing of
exchange rate risk is more pronounced for companies that are highly sensitive to economic
channels, such as international trade, hot money, and local credit market condition, which
transmit the exchange rate risk exposure to asset risk premiums. Similarly, Chinese stocks
listed on the Hong Kong stock exchange are not sensitive to the three transmission chan-
nels. Our findings suggest that in the presence of robust transmission channels, currency
hedging against foreign-exchange rate risk exposure is warranted under a controlled float-
ing exchange rate regime. This prediction has been borne out in the increasing use of cur-
rency derivatives in hedging by Chinese companies.
The remainder of the article is structured as follows. The section ‘‘Institutional Details’’
provides institutional background on China’s exchange rate regime and discussion of eco-
nomic channels transmitting the exchange rate risk to asset risk premium. The section
‘‘Data Descriptions’’ describes the data. The section ‘‘Empirical Results’’ presents empiri-
cal results for the link between exchange rate risk and stock returns. The section
‘‘Transmission Channels and Pricing of Exchange Rate Risk’’ analyzes whether the pricing
of exchange rate risk is associated with firms’ sensitivity to possible economic channels
that transmit exchange rate risk exposure to stock returns. The section ‘‘Conclusion’’ offers
concluding remarks.
Institutional Details
Overview of China’s Exchange Rate Arrangements
During the period from 1979 to 1994, China had a dual RMB exchange rate regime, in
which the official rate and the market rate coexisted. The market value of RMB per USD
was always greater than the official rate and both exchange rates depreciated against the
USD continuously.4 In 1994, the Chinese monetary authority adopted exchange rate regime
that was pegged to the USD. During 1994 and 1997, the RMB appreciated steadily against
the USD. The currency value was then pegged to the USD at the level of US$1 to 8.28
RMB Yuan between October 1997 and July 2005.
As part of the reform toward a market-driven exchange rate determination, on July 21,
2005, the People’s Bank of China (PBOC), the nation’s central bank, adopted a new
regime that pegs the RMB to a ‘‘reference basket’’ of foreign currencies, including some
Asian currencies. The PBOC stated that the selection of assigned index weights is in line
with China’s real external sector development and the basket should be composed of cur-
rencies of the countries to which China has a prominent exposure in terms of foreign trade,
external debt, and foreign direct investment (FDI). The PBOC provided only guidelines
42 Journal of Accounting, Auditing & Finance
about the composition and trade weights of the reference basket. Because the United States
is China’s most important trading partner and the majority of exports from China are
denominated in USD, the USD presumably accounts for the largest trade weight in the
basket. Indeed, Spiegel (2005) found that movements in China’s trade-weighted exchange
rate indexes over the long term are relatively insensitive to currency composition. The
RMB appreciated by 2.1% immediately and a cumulative 21% against the USD by July
2008. The appreciation trend ended during the global financial crisis. To reduce uncertainty
and stimulate exports during the global financial crisis, China effectively pegged the Yuan
to the USD at 6.8274 Yuan per USD until June 2010. On June 19, 2010, China resumed
the reform to increase flexibility in exchange rate fluctuations. By June 2015, the RMB
rose to a new record of 6.1161 Yuan per USD, marking a 10.42% appreciation since the
resumption of exchange rate reform in June 2010. In recent years, China has implemented
a tightly controlled floating regime by setting the midpoint for the value of USD and allow-
ing a 2% daily fluctuation. Although still under a heavily managed regime, the value of the
RMB has been more volatile recently. Figure 1 describes the major changes in China’s for-
eign-exchange regime since 1979, including the exchange rate regime change on July 21,
2005.
Economic Channels Transmitting Exchange Rate Exposure to Asset Risk Premium
The exchange rate may affect a firm’s profitability directly, if the firm has either foreign
operations or assets and debt denominated in foreign currency. Moreover, the exchange rate
can have an indirect impact on firms’ profitability even if a firm has no foreign currency rev-
enues because the exchange rate regime can affect foreign competition and domestic macroe-
conomic conditions. Thus, a potentially wide range of firms could be exposed to movements
in foreign-exchange rates, regardless of their direct financial exposure (Parsley & Popper,
2006). To study the pricing of foreign-exchange risk, it is important to understand the
mechanisms that transmit foreign-exchange rate risk exposure to stock returns. The extent of
pricing of exchange rate risk exposure may depend on firms’ sensitivity to the economic
channels that transmit the exchange rate risk to asset returns. In view of related literature and
China’s market-oriented reform as well as prevailing macroeconomic policy, we propose
three economic channels through which exchange rate risk exposure may be transmitted to
stock returns: International trade, hot money channel, and domestic credit supply.
International trade. The most recognized transmission channel for exchange rate risk is
international trade. It is well known that currency appreciation generally raises the relative
price of domestic goods, creating less favorable terms of international trade that is detri-
mental to exports. Dornbusch and Fischer (1980) and Pavlova and Rigobon (2007) sug-
gested that domestic stock prices should fall in response to increasing domestic currency
value and vice versa due to trade linkage. However, currency depreciation would stimulate
exports and curtail imports. Therefore, for industries generating significant earnings from
export, currency appreciation (depreciation) may cause stock valuation to decline
(increase). The opposite may hold for industries with heavy imports.
The number of listed companies with exports in China has continued to rise over the
last decade. Appendix A reports the industry distribution of listed companies (Panel A) and
the percentage of companies with exports in sales (Panel B). It appears that the percentage
of companies that have revenues from exports increases over time for almost all industries
except for real estate and media. The percentage of total listed companies with exports
Hua et al. 43
increased from 6% in 2001 to 54% in 2014. Of those listed companies with high percen-
tages of revenue from exports, changes in sales and profits in international trade provide an
explicit transmission channel for the exchange rate risk to impact stock prices.5 Hence, the
pricing of exchange rate risk should be more pronounced for companies with higher sensi-
tivity to terms of trade.
Hot money channel. Another potential transmission channel through which the exchange
rate may affect asset pricing is speculative capital flow or so-called hot money.
Figure 1. China’s currency regime changes since 1979. Note. This figure documents major changes/events in the reform of China’s currency exchange rate regime since 1979.
44 Journal of Accounting, Auditing & Finance
International capital flow induced by speculation, either because of new information or irra-
tional exuberance or fear, may result in changes in risk premia in asset markets (Duarte &
Stockman, 2005). Furthermore, emerging countries generally have more volatile capital
inflows than developed countries (Broner & Rigobón, 2006). We therefore consider capital
flows as one transmission channel for exchange rate risk, even though empirical evidence
on whether stock market behavior in emerging economies is affected by international capi-
tal flows remains inconclusive (Agosin & Huaita, 2012).
As China slowly liberalizes its capital account over the last decade, foreign investors
have been channeling money into Chinese real estate, stock markets, bank accounts, and
other investments with the expectation of local currency appreciation. The speculative capi-
tal inflow may have fueled inflation, driven up stock prices, and helped accelerate a harm-
ful bubble in the real estate market to some extent. Given the importance of international
capital flow following the development of financial globalization in China, we conjecture
that firms with stock prices that are more sensitive to the hot money channel are more
likely to respond to exchange rate risk.6 However, the responses of stock returns to interna-
tional capital flows may be ambiguous. On one hand, the sudden inflow of large amounts
of hot money would increase the money supply in China and would help create a credit
boom, and consequently, the credit expansion induced by hot money may generate positive
valuation effects on asset prices. On the other hand, hot money generally increases with the
expectation of RMB appreciation, which may have negative effect on the terms of trade for
export-oriented industries. As such, the inflow of hot money may be taken by investors as
a signal to possible RMB appreciation as well as deterioration of terms of trade. Investors
would expect a decline in the profit margin and competitiveness of export-oriented compa-
nies, which may cause negative equity valuation effects on those firms.
Domestic credit supply. Related to the hot money channel, existing studies have also noted
that the exchange rate regime can affect risk premia in asset markets through international
reserve accumulation and local credit expansion. In particular, fixed or inflexible exchange
regimes may contribute to stronger and more rapid domestic credit growth than flexible
regimes, especially in the context of large capital inflows and heavy foreign-exchange
intervention (Reinhart & Montiel, 2001). Empirically, the exchange rate regime flexibility
has a statistically significant impact on domestic credit levels despite controlling for capital
inflows (Magud, Reinhart, & Vesperoni, 2014). Thus exchange rate risk exposures may be
transmitted into asset risk premia through local credit markets.
The persistent trade surplus of China in recent years generated rapid accumulation of
foreign reserves, which led to appreciation pressure on the RMB exchange rate.7 China’s
central bank has long purchased foreign-exchange assets and maintained capital account
control which resulted in a large injection of base money and the excess domestic credit
expansion in the form of domestic broad money supply (M2). The reform in the exchange
regime in July 2005 had not changed this trend. To some extent, an appreciation trend
of RMB since July 2005 may in fact have increasing instead of decreasing pressure on
foreign-exchange market intervention because of the rising inflow of speculative capital.
China’s holdings of foreign reserves have risen from US$140 billion in 1997 to more
than US$3.33 trillion at the end of 2015. Over the same period, the central bank injected
more than 20 trillion RMB Yuan to buy the country’s US$3.33 trillion foreign-exchange
reserves, which forced the base currency in circulation to surge. Based on official statistics,
in the process of this reserve buildup, China’s broad money supply (M2) has expanded at
an average rate of 16.5% a year from 1999 to 2015. Both China’s international reserve and
Hua et al. 45
M2-to-GDP ratio are now the highest in the world. Because the credit expansion may be
highly correlated with expansion of asset bubbles in China’s property, commodity, and
stock markets, for listed companies in some sectors, such as real estate and mining indus-
tries, exchange rate risk exposures may be transmitted to equity risk premium through this
channel.8 In view of the significant effect of exchange rate policy on credit expansion, we
hypothesize that companies with equity prices more sensitive to domestic credit expansion
channels are also more likely to have significant risk premium for exchange rate risk
exposure.
Data Descriptions
Our sample contains all listed ‘‘A’’ shares on the Shanghai Stock Exchange and Shenzhen
Stock Exchange. Data are obtained from Wind Information Co. Ltd. We use monthly divi-
dend adjusted prices for the period from July 1995 to June 2015. We exclude the sample
from 1993 to 1994 because only a limited number of stocks were traded during this period
in China. We also exclude those stocks if more than 8 months of prices are missing for a
given year. Panel A of Appendix A shows that the final number of companies in the
sample ranges from 280 to 2,593 during the sample period. Following the industry classifi-
cation standard of China’s Securities Regulatory Commission, we report the number of
companies in 18 industries: agriculture, forestry, stockbreeding, and fishing; mining; manu-
facturing; utilities including electricity, heating, gas, and water; construction; whole sale
and retail; transportation, storage, and post; hotel and restaurant; information technology;
financial institutions; real estate; renting and business services; science research and tech-
nology services; water conservancy, environment, and public facilities; education; health
and social work; culture, sport, and entertainment; and other industries.9 The Chinese stock
market grew fast in the last two decades, especially in the manufacturing industry. For
example, the number of stocks in manufacturing increased more than 10 times since 1995
and accounts for more than half of the total number of listed companies.
To examine the impact of foreign trade on the pricing of exchange rate risk, we distin-
guish companies with export business from those without each year in each industry. Panel
B of Appendix A reports the percentage of companies with export revenues in each indus-
try.10 In most of industries, the percentage of companies with export revenues steadily
increased over time. For example, for manufacturing industry, this percentage increased
from 8% in 2001 to 70% in 2014, while for construction industry, the percentage increased
from 5% in 2001 to 45% in 2014.
Following the conventions in the exchange rate quote, we adopt two nominal measure-
ments for RMB exchange rates. The first series is the nominal RMB versus USD spot
exchange rate (denoted hereafter as RMB/USD). It is the RMB value of one USD. RMB/
USD goes up (down) when RMB depreciates (appreciates) against the USD. The second
series is the nominal effective exchange rate (NEER), a weighted average of a basket of
foreign currencies per RMB Yuan (hereafter BASKET/RMB). The BASKET/RMB goes up
(down) when RMB appreciate (depreciate). The RMB/USD exchange rate data are released
by the PBOC and are available through WIND database, while the BASKET/RMB series
were obtained from Bank for International Settlements (BIS). Figure 2 plots the time series
trend of RMB/USD and BASKET/RMB, which is the RMB NEER, from July 31, 1995, to
June 30, 2015. The figure shows that RMB has appreciated over the USD and other major
currencies in the last 10 years. Notably, the RMB exchange rate series measured in both
USD and basket of currencies have become very volatile since 2010.
46 Journal of Accounting, Auditing & Finance
For the exchange rate risk transmission channels, we construct relevant macro time
series variables that proxy for the factors discussed in the previous section.11 TRADE is
the monthly percentage changes of total values of exports and imports and is a macroeco-
nomic variable that captures China’s international trade. Hot money (HM) is the difference
between the change in foreign-exchange reserves and the balance of trade and service
minus the FDI. It captures the capital flows unrelated to FDI. Finally, we use M2 broad
money supply to proxy for domestic credit supply.12 The data sources involved in the con-
struction of the above variables are from the National Bureau of Statistics, National
Development and Reform Commission, State Administration of Foreign Exchange,
Bloomberg, and Wind Database.
Table 1 reports descriptive statistics on the percentage change of the two exchange rate
series and proxies for the three transmission channels: percentage changes of international
trade (TRADE), hot money (HM), and broad money supply (M2). The mean percentage
change of the RMB/USD exchange rate is negative, suggesting an appreciation of RMB
against the USD over the sample period. Likewise, the positive average percentage change
in BASKET/RMB also suggests an RMB appreciation against a basket of currencies.
Empirical Results
Industry Exposure to Exchange Rate Movements
Exchange rate exposure is often examined at the industry level because an industry in one
country often competes with the same industry in another country. Industries involved in
extensive international trade are more often exposed to exchange rate risk than industries
with low levels of international trade (Priestley & Ødegaard, 2007). To examine the indus-
try exposure to foreign-exchange rate risk in China, we conduct a regression of 18 industry
returns on the monthly changes of exchange rates. The equation is as follows:
Figure 2. The trend of RMB/USD and BASKET/RMB. Note. The figure shows the trend of the exchange rates measured by RMB/USD (USD in the figure, right scale),
which is the nominal RMB exchange rate against USD, and by BASKET/RMB (NEER in the figure, left scale), which is
the RMB NEER, from July 31, 1995 to June 30, 2015. NEER = nominal effective exchange rate.
Hua et al. 47
rit =ji + giXt + et, ð1Þ
where rit represents 18 value-weighted average industry returns (i = 1, 2, . . . , 18), and Xt is
the percentage change of exchange rates. We use RMB/USD and BASKET/RMB, respec-
tively, as Xt to measure the exchange rate shocks. The estimated exposure coefficient gi
represents the degree of the impact of exchange rate movement on industry stock returns.
For BASKET/RMB, we split the sample into two periods before and after July 2005 to cap-
ture the impact of exchange rate regime change in China. We expect that RMB apprecia-
tion would decrease the stock returns due to its negative impact on exports. Because the
values of the BASKET/RMB (RMB/USD) exchange rates will increase (decrease) when
RMB appreciates, the exposure coefficients gis of BASKET/RMB (RMB/USD) should be
negative (positive).
Table 2 reports the estimation results on Equation 1. The degree of sensitivity to
exchange rate change is measured by coefficient gi. There are a couple of noteworthy
results. First, stock returns at industry level are not sensitive to exchange rate movement
(measured by BASKET/RMB) prior to July 2005 when China adopted a regime of strictly
pegging the RMB to the USD at an exchange rate of 8.28. However, after the exchange
regime was changed to a controlled floating rate pegging to a trade-weighted ‘‘reference
basket’’ of currency, the time series exchange rate movement is significantly correlated
with most of industry return series. The sign is consistent with our prediction, negative for
BASKET/RMB and positive for RMB/USD, suggesting that Chinese currency appreciation
(or depreciation) exerts a negative (or positive) effect on stock returns. Second, for the
USD-based exchange rate series (RMB/USD), the significant sensitivity to exchange rate
movement exists for some industries such as mining, manufacturing, construction, finance,
and other industries. These industries typically have high ratio of export. The results sug-
gest that there are industry variations in exposure to foreign-exchange rate movement, per-
haps due to different industry’s sensitivity to the exchange rate risk transmission channels.
Price Reactions of ‘‘A’’ Shares to Announcement of Exchange Rate Reform
Prior studies suggest that stock prices should fall in response to increasing domestic cur-
rency value and vice versa because currency appreciation may deteriorate the terms of
Table 1. Descriptive Statistics of Percentage Changes in Exchange Rates and Economic Transmission Variables. BASKET/RMB, RMB/USD, TRADE, HM, and M2 represent monthly percentage change of RMB nominal effective exchange rate, nominal RMB exchange rate against the USD, international trade (TRADE), hot money (HM), and broad money supply (M2), respectively. For the BASKET/RMB, the sample starts from July 1995, whereas for the USD, the sample starts from July 2005. The sample for macroeconomic variables starts in July 1995. All variables end at June 2015.
M Median SD Sample variance Kurtosis Skewness
BASKET/RMB 0.0024 0.0029 0.0124 0.0002 1.2416 0.1470 RMB/USD –0.0025 –0.0018 0.0037 0.0000 2.0636 –1.3232 TRADE 0.0219 0.0161 0.1487 0.0221 2.4801 0.3476 HM –0.2657 –0.2094 2.8734 8.2564 24.3162 1.9056 M2 0.0134 0.0123 0.0131 0.0002 8.4673 0.9010
48 Journal of Accounting, Auditing & Finance
trade while currency depreciation may enhance the terms of trade, particularly for indus-
tries generating revenue from export (e.g., Dornbusch & Fischer, 1980; Pavlova &
Rigobon, 2007). China has been under external pressure to appreciate its currency value,
and the ongoing reform to increase the exchange rate flexibility has generally raised the
expectation on currency appreciation. Therefore, the announcement of exchange rate
reform provides an opportunity to study the valuation effect of the exchange rate risk expo-
sure. To that end, we examine stock market reactions to news announcements about
Table 2. Sensitivity of Industry Returns to Exchange Rate Movements. The numbers shown in the panel are estimated gis from the following regression: rit =ji + giXt + et,
where rit represents industry portfolio returns (industries are classified by China Securities Regulatory Commission. The Civil Services, Maintenance, and Other Industry are omitted due to lack of data). The value-weighted industry portfolios are formed at the end of June of each year, t. Companies with less than 4 months prices are omitted from that year. Xt is the percentage change of nominal effective exchange rate (BASKET/RMB), which is a weighted average of a basket of foreign currencies per RMB Yuan, and nominal RMB versus USD spot exchange rate (RMB/USD), which is the RMB value of every USD.
BASKET/RMB (July 1995-June 2005)
BASKET/RMB (July 2005-June 2015)
RMB/USD (July 2005-June 2015)
Agri. 0.60 (0.73) –1.86 (–1.99**) 2.91 (1.09) Min. 0.45 (0.48) –2.57 (–3.43***) 4.43 (2.02**) Manuf. 0.47 (0.79) –2.37 (–3.08***) 4.24 (1.90*) Util. 0.53 (0.89) –1.18 (–1.63) 3.13 (1.52) Con. 0.87 (1.29) –1.29 (–1.53) 5.67 (2.40**) Who.&Ret. 0.70 (1.18) –2.45 (–3.08***) 3.37 (1.45) Trans. 0.46 (0.77) –1.58 (–1.98**) 3.52 (1.54) Hot.&Rest. 0.51 (0.60) –3.17 (–3.56***) 3.43 (1.30) I.T. 0.71 (0.92) –2.53 (–3.05***) 3.50 (1.45) Fin. 0.14 (0.20) –1.93 (–2.73***) 3.78 (1.85*) Real Est. 0.38 (0.56) –2.31 (–2.90***) 3.37 (1.45) Rent.&Bus. 0.04 (0.06) –2.82 (–3.50***) 3.61 (1.52) Scie.&Tech. 1.08 (1.23) –1.70 (–1.62) 1.81 (0.60) Envi. 0.96 (0.89) –2.29 (–2.91**) 3.49 (1.53) Edu. –0.31 (–0.39) –2.16 (–1.87*) 3.58 (1.08) Health 0.55 (0.77) –1.99 (–1.82*) 3.41 (1.09) Enter. 0.56 (0.61) –2.58 (–2.93***) 3.26 (1.27) Gen. –0.14 (–0.18) –2.59 (–3.09***) 4.62 (1.89*)
Note. Numbers in parentheses are t statistics. For the BASKET/RMB case, the sample breaks into two parts at July
2005. The industries include Agriculture, Forestry, Stockbreeding, and Fishing (Agri.); Mining (Min.); Manufacturing
(Manuf.); Utility (Util.) including Electricity, Heating, Gas, and Water; Construction (Con.); Whole sale and Retail
(Who.&Ret.); Transportation, Storage, and Post (Trans.); Hotel and Restaurant (Hot.&Rest.); Information Technology
(I.T.); Financial Institutions (Fin.); Real Estate (Real Est.); Renting and Business Services (Rent.&Bus.); Science
Research and Technology Services (Scie.&Tech.); Water Conservancy, Environment, and Public Facilities (Envi.);
Education (Edu.); Health and Social Work (Health); Culture, Sport, and Entertainment (Enter.); and Other Industries
(Gen.). For Agriculture, Forestry, Stockbreeding, and Fishing (Agri.), the sample starts from July 1997 and for Health
and Social Work (Health) the sample starts from July 1996, due to limited number of companies. Other industries
starts from July 1995 and all series end at June 2015.
*denotes statistical significance at 10% significance level. **denotes statistical significance at 5% significance level.
***denotes statistical significance at 1% significance level.
Hua et al. 49
exchange rate reform on July 21, 2005, and June 19, 2010.13 We use a standard event
study methodology (Brown & Warner, 1985; Mazouz, Joseph, & Joulmer, 2009). We
hypothesize that because of expectation of currency appreciation following exchange rate
regime change, the cumulative abnormal returns (CAR) of companies with export revenue
in sales are more adversely affected by RMB regime reform news than companies that do
not have export revenue in business activities. We estimate the abnormal returns (ARjt) by
applying the following equation:
ARjt =Rjt � âj � b̂jRmt, ð2Þ
where Rjt is the return of company j at day t, and the estimated coefficients âj and b̂j are
from the standard market model for an estimation window (–260, –61) interval prior to the
event (t = 0):
Rjt = aj + bjRmt + ejt, ð3Þ
where Rmt is the returns on Shanghai or Shenzhen composite indexes depending on where
security j was listed. The CAR for each security j, CARj, is the sum of average abnormal
returns over the event period. The formula is as follows:
CARj, T1, T2 = XT2
t = T1
ARjt, ð4Þ
where CARj, T1, T2 is the CAR during the period from t = day T1 to t = day T2. We report
six event windows, namely [–1,0], [–3,+3], [–5,+5], [–6,+6], [–11,+11], and [–21,+21]. We
examine the different reactions to news announcements for either export or nonexport
companies.14
Table 3 shows that the average CARs for export and nonexport companies over six win-
dows for two event days. Average CARs of both export and nonexport companies are nega-
tive for both event days, suggesting that the news announcement of exchange rate reform
caused expectation of RMB appreciation and triggered negative stock market reaction on
average. This shows that the expectation of currency appreciation has a negative valuation
effect on the equity market, reflecting the fact that China has an export-oriented economy.
In fact, the results show that the export companies have a stronger negative market reaction
than nonexport companies. For the announcement day of July 21, 2005, the CARs for
export portfolios are significantly negative for all six windows at either the 1% or 5% sig-
nificance level, while those of nonexport portfolios are not significant for some event win-
dows such as [–3,+3] and [–6,+6]. Furthermore, the magnitude of the negative CARs of
export portfolios are larger than those of the nonexport portfolio and the differences in
CARs between export portfolios and nonexport portfolios are significantly negative at the
1% level except for the event window of [–1,0]. For the event date of June 19, 2010, we
observed similar patterns and in most of the event windows the negative CARs of both
export and nonexport portfolios are significantly different from zero. Again, the differences
in CARs between export and nonexport portfolios are statistically significant across all six
event windows. The results thus show that companies with export revenue in sales are
more adversely affected by the news of Chinese exchange rate regime change, which
50 Journal of Accounting, Auditing & Finance
effectively caused the RMB appreciation process. This implies that international trade is an
important channel for transmitting exchange rate risk to asset returns.
An Examination of Price Reaction of ‘‘H’’ Shares
Some Chinese stocks are listed in Hong Kong as ‘‘H’’ shares with share prices denomi-
nated in Hong Kong dollars. Because the Hong Kong dollar remained pegged to the USD
during the sample period of our study, it is possible that Chinese shares listed on the Hong
Kong stock exchange may respond to exchange rate regime change in China in a different
way from what we observed for the mainland China shares. Thus, a study of ‘‘H’’ shares
can offer further evidence on whether the difference in sensitivities to RMB regime change
may cause a different price reaction.15 To that end, we examine the stock price reaction to
announcement of RMB exchange rate reform for ‘‘H’’ shares. In particular, we compare
the differences in CARs between ‘‘H’’ shares that have corresponding ‘‘A’’ shares traded
simultaneously in China and ‘‘H’’ shares that have no corresponding ‘‘A’’ shares. In addi-
tion, we also compare the CARs between ‘‘H’’ shares and their corresponding ‘‘A’’ shares.
We use Hang Seng China Enterprises Index as the market index for ‘‘H’’ shares.
Table 4 reports the average CARs over six windows around the announcement day on
July 21, 2005. ‘‘H with A’’ (‘‘H without A’’) refers to ‘‘H’’ shares that have (that have no)
corresponding ‘‘A’’ share companies listed on stock exchange in mainland China, respec-
tively. ‘‘A-H’’ shares refer to ‘‘A’’ shares that have ‘‘H’’ shares simultaneously cross-listed
on Hong Kong Stock Exchange. ‘‘Mean Comparison’’ reports the mean differences
between CARs. The left panel shows that for the ‘‘H with A’’ shares, the stock price has a
Table 3. Price Reaction of ‘‘A’’ Shares to Announcements of Exchange Rate Reform.
Event window Export Nonexport
Mean comparison
Event window Export Nonexport
Mean comparison
Announcement day: July 21, 2005 Announcement day: June 19, 2010
[–1,0] –0.0030 –0.0023 –0.0007 [–1,0] –0.0044 –0.0017 –0.0028 (–2.19**) (–2.15**) (–0.53) (–3.40***) (–1.45) (–2.93***)
[–3,+3] –0.0100 –0.0013 –0.0087 [–3,+3] –0.0239 –0.0156 –0.0083 (–3.96***) (–0.66) (–4.31***) (–9.82***) (–7.25***) (–4.66***)
[–5,+5] –0.0180 –0.0044 –0.0136 [–5,+5] –0.0208 –0.0121 –0.0088 (–5.68***) (–1.76*) (–4.92***) (–6.82***) (–4.47***) (–4.14***)
[–6,+6] –0.0121 –0.0012 –0.0110 [–6,+6] –0.0326 –0.0243 –0.0083 (–3.51***) (–0.43) (–3.85***) (–9.81***) (–8.27***) (–3.30***)
[–11,+11] –0.0311 –0.0201 –0.0109 [–11,+11] –0.0559 –0.0427 –0.0132 (–6.67***) (–5.57***) (–3.17***) (–12.65***) (–10.92***) (–3.58***)
[–21,+21] –0.0448 –0.0309 –0.0139 [–21,+21] –0.0593 –0.0361 –0.0232 (–7.13***) (–6.26***) (–3.17***) (–9.82***) (–6.76***) (–3.73***)
Note. This table shows the average cumulative abnormal returns (CAR) for export and nonexport ‘‘A’’ share
companies over six windows for two event days in which the Chinese government announced exchange rate
regime reform news. Export (nonexport) companies refer to companies with (without) export revenues at the
event year. Mean Comparison reports the mean differences between CARs from export and nonexport companies.
The t statistics are shown in the parentheses.
*denotes statistical significance at 10% significance level. **denotes statistical significance at 5% significance level.
***denotes statistical significance at 1% significance level.
Hua et al. 51
T a b
le 4 .
P ri
ce R
ea ct
io n
o f ‘‘H
’’ Sh
ar es
to A
n n o u n ce
m en
ts o f E x ch
an ge
R at
e R
ef o rm
.
E ve
n t
w in
d o w
H w
it h
A
(1 )
H w
it h o u t
A
(2 )
M ea
n co
m p ar
is o n
(1 )
– (2
)
A -H (3 )
M ea
n co
m p ar
is o n
(1 )
– (3
)
E ve
n t
w in
d o w
H w
it h
A
(4 )
H w
it h o ut
A
(5 )
M ea
n co
m p ar
is o n
(4 )
– (5
)
A -H (6 )
M ea
n co
m p ar
is o n
(4 )
– (6
)
A n n o u n ce
m en
t d ay
: Ju
ly 2 1,
2 0 0 5
A n no
u n ce
m en
t d ay
:J u n e
1 9,
2 01
0
[– 1,
0 ]
0 .1
7 86
– 0.
0 6 4 3
0 .2
42 9
– 0 .0
00 0 8
0 .1
7 87
[– 1 ,0
] – 0 .0
00 7
– 0 .0
07 0
0 .0
0 64
0 .0
0 00
3 0 .0
00 7
(2 .1
1* * )
(– 0 .7
7 )
(1 .9
2* )
(– 0 .0
3 )
(2 .2
7 * * )
(– 0 .0
3 )
(– 0 .1
8 )
(0 .1
5) (0
.0 2 )
(0 .0
3)
[– 3,
+ 3 ]
0 .4
0 11
0 .1
4 36
0 .2
57 5
0 .0
06 1
0 .3
9 50
[– 3 ,+
3 ]
– 0 .0
83 7
– 0 .0
13 0
– 0 .0
70 7
– 0.
0 0 4 1
0 .0
79 6
(4 .1
8* * * )
(1 .3
7 )
(1 .8
0* )
(1 .1
2) (3
.9 3 * * * )
(– 1 .0
8 )
(– 0 .1
7 )
(– 0 .6
6 )
(– 1 .1
2 )
(1 .0
5)
[– 5,
+ 5 ]
0 .4
4 49
0 .0
5 74
0 .3
87 5
0 .0
15 7
0 .4
2 92
[– 5 ,+
5 ]
– 0 .1
24 4
0 .0
8 41
– 0 .2
08 5
– 0.
0 0 4 2
0 .1
20 2
(3 .1
3* * * )
(0 .5
2 )
(2 .0
9* * )
(2 .1
6* * )
(2 .8
8 * * * )
(– 1 .1
9 )
(0 .8
8) (–
1 .4
9 )
(– 1 .0
3 )
(1 .1
8)
[– 6,
+ 6 ]
0 .4
8 49
0 .0
3 74
0 .4
47 5
0 .0
14 9
0 .4
7 00
[– 6 ,+
6 ]
– 0 .2
91 0
0 .0
5 74
– 0 .3
48 4
– 0.
0 0 7 8
0 .2
83 2
(3 .0
4* * * )
(0 .3
5 )
(2 .2
8* * )
(2 .0
7* * )
(2 .8
2 * * * )
(– 2 .7
2 * * )
(0 .5
4) (–
2 .3
5 * * )
(– 1 .4
8 )
(2 .7
3* * )
[– 11
,+ 1 1 ]
0 .5
8 97
0 .0
9 24
0 .4
97 3
0 .0
18 1
0 .5
7 10
[– 1 1 ,+
11 ]
– 0 .0
99 5
0 .0
7 18
– 0 .1
71 3
– 0.
0 1 4 1
0 .0
85 4
(1 .9
6* )
(0 .7
7 )
(1 .5
0) (2
.1 6* * )
(1 .8
0 * )
(– 0 .7
1 )
(0 .3
1) (–
0 .8
6 )
(– 1 .7
4 * )
(0 .6
3)
[– 21
,+ 2 1 ]
1 .2
4 33
0 .0
5 74
1 .1
86 0
0 .0
07 2
1 .2
3 61
[– 2 1 ,+
21 ]
– 0 .1
86 5
– 0 .0
74 1
– 0 .1
12 4
– 0.
0 1 9 8
0 .1
66 7
(2 .3
2* * )
(0 .4
0 )
(2 .0
3* * )
(0 .7
5) (2
.1 9 * * )
(– 0 .9
4 )
(– 0 .4
2 )
(0 .4
3) (–
1 .6
0 )
(0 .8
6)
N ot
e. T
h is
ta b le
sh o w
s th
e av
er ag
e cu
m u la
ti ve
ab n o rm
al re
tu rn
s (C
A R )
fo r
‘‘H ’’
sh ar
es an
d th
ei r
co rr
es p o n d in
g ‘‘A
’’ sh
ar es
co m
p an
ie s,
o ve
r si
x w
in d o w
s fo
r tw
o ev
en t
d ay
s in
w h ic
h th
e C
h in
es e
go ve
rn m
en t
an n o u n ce
d ex
ch an
ge ra
te re
gi m
e re
fo rm
n ew
s. ‘‘H
w it h
A ’’
(‘ ‘H
w it h o u t
A ’’)
re fe
r to
‘‘H ’’
sh ar
es th
at h av
e (t
h at
h av
e n o )
co rr
es p o n d in
g ‘‘A
’’
sh ar
e co
m p an
ie s
lis te
d o n
st o ck
ex ch
an ge
in m
ai n la
n d
C h in
a, re
sp ec
ti ve
ly . T
h e
sa m
p le
si ze
fo r
‘‘H w
it h
A ’’
is 4 8
(f o r
‘‘H w
it h o u t
A ’’
is 4 6 )
in 2 0 0 5
an d
it is
7 1
fo r
‘‘H w
it h
A ’’
(7 3
‘‘H w
it h o u t
A ’’)
in 2 0 1 0.
‘‘A -H
’’ sh
ar es
re fe
r to
‘‘A ’’
sh ar
es th
at h av
e ‘‘H
’’ sh
ar es
si m
u lt an
eo u sl
y cr
o ss
-l is
te d
o n
H o n g
K o n g
St o ck
E x ch
an ge
. ‘‘M
ea n
C o m
p ar
is o n ’’
re p o rt
s th
e
m ea
n d iff
er en
ce s
b et
w ee
n C
A R s.
T h e
t st
at is
ti cs
ar e
sh o w
n in
th e
p ar
en th
es es
.
* d en
o te
s st
at is
ti ca
l si
gn ifi
ca n ce
at 1 0 %
si gn
ifi ca
n ce
le ve
l. * * d en
o te
s st
at is
ti ca
l si
gn ifi
ca n ce
at 5 %
si gn
ifi ca
n ce
le ve
l. * * * d en
o te
s st
at is
ti ca
l si
gn ifi
ca n ce
at 1 %
si gn
ifi ca
n ce
le ve
l.
52
positive CAR up to 6 days around the announcement day on July 21, 2005. By contrast, the
market reaction for ‘‘H’’ shares without domestic ‘‘A’’ shares is not significant for any of
the windows examined. The results imply that for investors in ‘‘H’’ shares with the fungi-
ble ‘‘A’’ shares traded on domestic Chinese markets, an expectation of RMB appreciation
against the USD could lead to an increase in expected value of ‘‘H’’ shares in Hong Kong
dollars, because Hong Kong currency remained pegged to USD. On the contrary, the prices
of ‘‘H’’ shares without corresponding ‘‘A’’ shares are not affected by the RMB apprecia-
tion and do not react significantly to the announcement exchange rate regime change.
The same panel also reports the comparison of CARs between ‘‘H’’ shares and their cor-
responding ‘‘A’’ shares. Results suggest that although ‘‘H’’ shares that have corresponding
‘‘A’’ shares experienced significantly positive price reaction, their ‘‘A’’ share counterparts
do not have any significant price reactions over most of windows except for [–5,+5],
[–6,+6], and [–11,11]. On average, the difference in CARs between ‘‘H’’ shares and their
corresponding ‘‘A’’ shares is significantly positive at the 5% level or higher. The difference
in CARs between ‘‘H’’ shares and their corresponding ‘‘A’’ shares suggests that the strictly
pegged currency regime in Hong Kong leads to a different stock price reaction to the
Chinese exchange rate reform for shares cross-listed in Hong Kong and China. The results
in the right panel show that on the event day of June 19, 2010 (announcement of resump-
tion of suspended exchange rate reform), the CARs are generally not significant for ‘‘H’’
shares as well as for the corresponding ‘‘A’’ shares.
Exchange Rate Exposure and Earnings Volatility
The analysis of market reaction in above section shows that investors do expect an increase
in exchange rate exposure following the change from pegged regime to a managed floating
rate. We thus expect that under the managed floating regime, firms with greater foreign-
exchange exposure may experience greater volatility of earnings measure. To examine the
link between exchange rate exposure and volatility of accounting measures, we first group
companies into ‘‘export’’ and ‘‘nonexport’’ stocks. ‘‘Export’’ (nonexport) companies refer
to companies with (without) export revenues each year. As we showed in the article, com-
panies with revenues from export are more sensitive to the change from pegged exchange
rate to a floating rate regime. We thus compute the standard deviation of accounting perfor-
mance and of valuation measures for the two portfolios based on a 3-year rolling window.
The accounting performance measures include earnings scaled by total assets and ROA,
and the valuation measure includes P/E ratio. Three accounting measures are all ratios;
therefore, the standard deviations have the same units.
Table 5 reports the volatilities of accounting measures for each year from 2001 until
2014. We compute the mean differences in three volatilities between ‘‘export’’ and ‘‘non-
export’’ companies for the period before or after exchange rate regime change in 2005. The
results of mean differences test show that for the period after 2005, the volatilities of all
three accounting measures are significantly greater for the ‘‘export’’ companies than for the
‘‘nonexport’’ companies. However, for the period before 2005, the differences between
‘‘nonexport’’ and ‘‘export’’ companies are not significant. This result suggests that the
exchange rate regime change in 2005 led to greater foreign-exchange rate exposure, which
in turn led to greater volatility in earnings, profitability, and P/E ratio for export companies
whose exposure to exchange rate risk is greater.
Hua et al. 53
Pricing of Exchange Rate Risk in Cross-Section of Stock Returns
To test whether the exchange rate risks are priced in stock returns, we use a factor portfolio
approach similar to Kolari et al. (2008) to evaluate whether the constructed exchange rate
risk factor is cross-sectionally related to expected returns and whether the inclusion of the
risk factor reduces the pricing error of existing asset-pricing models. As shown in Kolari
et al. (2008), by constructing a foreign-exchange risk factor in the form of a zero-
investment portfolio that buys stocks sensitive to foreign-exchange rate changes and sells
stocks that are not sensitive to exchange rate changes, the factor portfolio approach over-
comes the weakness of using raw changes in the exchange rate, which may contain infor-
mation irrelevant to asset pricing.
We first measure the extent to which each stock is exposed to foreign-exchange risk.
The foreign-exchange exposure for each firm is measured by regressing its stock returns on
the foreign-exchange return series (Xt) with four other risk factors (Carhart, 1997):
Table 5. Volatility of Firm Performance of Nonexport and Export Companies.
Nonexport Export
Number of
companies
Earnings
volume
ROA
volume
P/E
volume
Number of
companies
Earnings
volume
ROA
volume
P/E
volume
2001 963 0.017 3.138 115.11 115 0.019 3.294 214.82
2002 869 0.016 2.901 124.97 280 0.016 3.485 254.32
2003 807 0.328 1.805 344.09 406 0.327 2.073 210.65
2004 817 0.309 2.083 263.28 492 0.319 2.604 366.87
2005 811 0.330 2.548 677.63 513 0.350 3.241 434.26
Mean comparison –0.006
(–0.06)
–0.44
(–1.23)
8.83
(0.08)
2006 826 0.323 2.553 627.86 560 0.357 3.288 562.93
2007 847 0.321 2.561 510.71 632 0.350 3.166 307.07
2008 875 0.332 3.458 152.89 682 0.369 3.750 369.75
2009 954 0.347 4.160 135.783 765 0.370 8.059 894.99
2010 978 0.314 4.201 245.34 1,051 0.361 10.252 1,294.75
2011 1,087 0.326 3.853 192.84 1,192 0.356 7.342 1,131.68
2012 1,149 0.364 4.088 197.85 1,283 0.375 8.214 975.11
2013 1,155 0.397 5.106 260.29 1,333 0.402 6.568 286.60
2014 1,201 0.405 4.409 314.59 1,390 0.419 4.809 364.68
Mean comparison –0.26
(–1.88*)
–2.34
(–3.35***)
–394.38
(–2.45**)
Note. The table shows average performance volatility of nonexport and export companies. ‘‘nonexport’’ (‘‘export’’)
companies refer to companies without (with) export revenues each year. The data on export revenue for each
company, obtained from annual financial statement, are available from year 2001. Firm performance measures
include earnings, ROA and P/E ratio. Earnings volatility is measured by the standard deviation of earnings to total
assets over the past 3 years. ROA and P/E volatility are measured by the standard deviation of each measure over
the past 3 years. The panels show the number of companies included in each year (note that companies with less
than 4 months prices are eliminated from that year) and cross-sectional mean of volatility of each performance
measure from nonexport and export companies. Mean differences of average volatility of nonexport and export
companies are reported in the row below the subsample period. The t statistics of mean comparison tests are
reported in the parentheses.
*denotes statistical significance at 10% significance level. **denotes statistical significance at 5% significance level.
***denotes statistical significance at 1% significance level.
54 Journal of Accounting, Auditing & Finance
rnt � Rft
� � = an + an1 Rmt � Rft
� � + an2SMBt + an3HMLt + an4MOMt + an5Xt + ent, ð5Þ
where rnt is the log return on stock n in month t, Rmt is the return on all companies’ value-
weighted returns, Rft is the 1-year deposit interest rate at month t, SMB is the return on a
portfolio of small stocks minus the return on a portfolio of large stocks, HML is the return
on a portfolio of high book-to-market stocks minus the return on a portfolio of low book-
to-market stocks, and MOM is the return on a portfolio of past winner stocks minus a port-
folio of past loser stocks. Portfolios are rebalanced each year at the end of June. Xt is the
percentage change of exchange rates. We use RMB/USD and BASKET/RMB, respectively,
as Xt to measure exchange rate risks. The coefficient an5 is a proxy for estimated exposure
to foreign-exchange risk.
We obtain annual foreign-exchange risk exposure coefficients for each stock by estimat-
ing Equation 5 using monthly data over 2-year rolling windows beginning in July each
year.16 Each year, we then rank firms into 25 portfolios based on the value of the firm-
specific exposure coefficient an5.17 Following Kolari et al. (2008), the portfolios ranked 1st
and 25th (i.e., portfolios with the highest foreign-exchange sensitivity in absolute value)
are defined as the sensitive portfolio and those ranked from the 2nd to 24th are grouped as
insensitive portfolios. We construct the exchange rate risk factor, XMI, by creating a portfo-
lio that buys stocks with extreme negative or positive sensitivity to foreign-exchange risk
(portfolios ranked 1st and 25th) and sells all other stocks (portfolios ranked 2nd to 24th).
We evaluate whether the exchange rate is a priced factor by including XMI as an additional
factor into both the Fama-French three-factor model (Fama & French, 1993) and the Fama-
French–Carhart four-factor model and examining the change in pricing errors.18 In particu-
lar, we regress the excess returns on 25 portfolios sorted by the firm-specific exposure coef-
ficient an5 against the Fama-French three- or four-factor (Carhart) models plus the
exchange rate risk factor XMI:
Rit � Rft
� � = ai + bi1 Rmt � Rft
� � + bi2SMBt + bi3HMLt + bi4MOMt + bi5XMIt + eit, ð6Þ
Rit � Rft
� � = ai + bi1 Rmt � Rft
� � + bi2SMBt + bi3HMLt + bi4XMIt + eit, ð7Þ
where Rit represents value-weighted portfolio returns (portfolios from 1 to 25) at month t.
As in Kolari et al. (2008), we argue that the exchange rate risk factor XMI is a priced
factor in stock returns if the average of absolute ai or squaredai (the pricing error) signifi-
cantly decreases in Equation 6 or 7 compared with the base models without XMI. We for-
mally conduct a Gibbons–Ross–Shanken test (Gibbons, Ross, & Shanken, 1989, GRS test
hereafter) with the null hypothesis that model’s intercepts are jointly equal to zero for all
25 portfolios.
Table 6 shows the average of coefficient an5, size (logarithmic), and average annual raw
returns on the 25 portfolios formed on foreign-exchange sensitivity, where the exchange
rates are measured by the RMB/USD and BASKET/RMB, respectively. Portfolios ranked 1st
and 25th are those with the highest foreign-exchange sensitivity in absolute value. We first
calculate a cross-sectional average raw return of each of the 25 portfolios each year. We
then compute the intertemporal average of annual portfolio returns. All returns are annual-
ized. In the last row of each panel, we report the differences between average annual raw
returns of foreign-exchange insensitive portfolios (2U . . . 24) and sensitive portfolios
(1U25).
Hua et al. 55
T a b
le 6 .
Si ze
s an
d R
aw R
et u rn
s o f 2 5
Po rt
fo lio
s Fo
rm ed
o n
Fo re
ig n -E
x ch
an ge
Se n si
ti vi
ty .
Fo r
th e
p er
io d
Ju ly
1 9 9 7
to Ju
n e
2 0 1 5
(f ro
m 2 0 0 7
fo r
th e
R M
B /U
SD ca
se ),
2 5
p o rt
fo lio
s ar
e fo
rm ed
at en
d o f Ju
n e
o f ea
ch ye
ar , t,
b as
ed o n
th e
se n si
ti vi
ty o f
in d iv
id u al
fir m
s to
ch an
ge s
in ex
ch an
ge ra
te s
(R M
B /U
SD o r
B A SK
E T /R
M B ).
Se n si
ti vi
ty is
m ea
su re
d b y
th e
co ef
fic ie
n t
o n
X t in
th e
fo llo
w in
g re
gr es
si o n :
r n t �
R ft
� � =
a n
+ a
n1 R
m t �
R ft
� � +
a n2
SM B
t +
a n3
H M
L t +
a n4
M O
M t +
a n5
X t + e n
t,
ab o ve
eq u at
io n
is es
ti m
at ed
w it h
d at
a fr
o m
1 9 9 5
(2 0 0 5
fo r
th e
R M
B /U
SD ca
se )
th ro
u gh
2 0 1 5
o n
a 2 -y
ea r
ro lli
n g
w in
d o w
. T
h e
r n t is
th e
lo g
re tu
rn o n
st o ck
n in
m o n th
t. R
m t
is th
e va
lu e-
w ei
gh te
d av
er ag
e re
tu rn
o f
al l co
m p an
ie s
in th
e sa
m p le
. R
ft is
th e
1 -y
ea r
d ep
o si
t in
te re
st ra
te in
m o n th
t. SM
B is
th e
re tu
rn o n
a p o rt
fo lio
o f sm
al l st
o ck
s m
in u s
th e
re tu
rn o n
a p o rt
fo lio
o f la
rg e
st o ck
s. H
M L
is th
e re
tu rn
o n
a p o rt
fo lio
o f h ig
h b o o k-
to -m
ar ke
t st
o ck
s m
in u s
th e
re tu
rn o n
a p o rt
fo lio
o f lo
w b o o k-
to -m
ar ke
t st
o ck
s. M
O M
is th
e re
tu rn
o n
a p o rt
fo lio
o f p as
t w
in n er
st o ck
s m
in u s
a p o rt
fo lio
o f p as
t lo
se r
st o ck
s. X
t is
th e
p er
ce n ta
ge ch
an ge
o f ex
ch an
ge ra
te s
(R M
B /U
SD o r
B A SK
E T /R
M B ).
T h e
co ef
fic ie
n t
a n5
o n
X t is
th u s
es ti m
at ed
o ve
r a
2 -y
ea r
ro lli
n g
p er
io d . Si
ze is
th e
lo ga
ri th
m o f m
ar ke
t va
lu e
at th
e en
d o f
Ju n e
o f
ea ch
ye ar
, t.
Po rt
fo lio
re tu
rn s
ar e
fr o m
Ju ly
o f
ye ar
t to
Ju n e
o f
ye ar
t +
1 .
A n n u al
re tu
rn is
o b ta
in ed
b y
m u lt ip
ly in
g th
e va
lu e-
w ei
gh te
d av
er ag
e m
o n th
ly re
tu rn
s o f th
e p o rt
fo lio
s b y
1 2 .
Fi rm
m o nt
h C
o ef
fic ie
n t
o n
X (a
n5 )
Si ze
A ve
ra ge
an n u al
re tu
rn
R M
B /U
SD B A SK
E T /R
M B
R M
B /U
SD B A SK
E T /R
M B
R M
B /U
SD B
A SK
E T /R
M B
R M
B /U
SD B A SK
E T /R
M B
Ju ly
1 99
7 -J u n e
2 00
6
1 —
4 ,6
4 4
— – 4 .2
4 —
2 0 .7
0 —
– 0.
0 8 8
2 —
4 ,6
0 8
— – 2 .6
7 —
2 0 .7
4 —
– 0.
0 1 6
3 —
4 ,6
0 8
— – 2 .0
8 —
2 0 .8
4 —
– 0.
0 8 6
4 —
4 ,6
0 8
— – 1 .7
2 —
2 0 .7
9 —
0 .0
1 4
5 —
4 ,5
9 6
— – 1 .4
5 —
2 0 .7
4 —
– 0.
0 6 4
6 —
4 ,5
8 4
— – 1 .2
2 —
2 0 .7
6 —
– 0.
0 5 0
7 —
4 ,6
3 2
— – 1 .0
2 —
2 0 .7
9 —
– 0.
0 4 8
8 —
4 ,6
0 8
— – 0 .8
4 —
2 0 .6
3 —
– 0.
0 1 3
9 —
4 ,5
7 2
— – 0 .6
7 —
2 0 .7
7 —
– 0.
0 5 6
1 0
— 4 ,6
4 4
— – 0 .5
0 —
2 0 .9
4 —
– 0.
0 5 0
1 1
— 4 ,5
9 6
— – 0 .3
4 —
2 0 .7
9 —
– 0.
0 5 5
1 2
— 4 ,5
8 4
— – 0 .1
9 —
2 0 .7
5 —
– 0.
0 3 3
1 3
— 4 ,6
0 8
— – 0 .0
4 —
2 0 .8
1 —
– 0.
0 2 1
1 4
— 4 ,5
8 4
— 0 .1
1 —
2 0 .7
0 —
0 .0
1 1
1 5
— 4 ,6
0 8
— 0 .2
6 —
2 0 .9
2 —
– 0.
0 2 7
1 6
— 4 ,6
2 0
— 0 .4
0 —
2 0 .8
0 —
– 0.
0 2 2
1 7
— 4 ,5
6 0
— 0 .5
6 —
2 0 .6
6 —
– 0.
0 1 9
(c on
tin ue
d)
56
T a b
le 6 . (c
o n ti nu
ed )
Fi rm
m o nt
h C
o ef
fic ie
n t
o n
X (a
n5 )
Si ze
A ve
ra ge
an n u al
re tu
rn
R M
B /U
SD B A SK
E T /R
M B
R M
B /U
SD B A SK
E T /R
M B
R M
B /U
SD B
A SK
E T /R
M B
R M
B /U
SD B A SK
E T /R
M B
1 8
— 4 ,6
0 8
— 0 .7
2 —
2 0 .9
4 —
0 .0
0 9
1 9
— 4 ,6
2 0
— 0 .9
0 —
2 0 .7
0 —
0 .0
0 7
2 0
— 4 ,6
0 8
— 1 .1
0 —
2 0 .6
7 —
0 .0
2 2
2 1
— 4 ,5
8 4
— 1 .3
4 —
2 0 .8
8 —
– 0.
0 0 8
2 2
— 4 ,6
0 8
— 1 .6
2 —
2 0 .9
0 —
– 0.
0 3 8
2 3
— 4 ,5
9 6
— 1 .9
9 —
2 0 .8
7 —
– 0.
0 4 5
2 4
— 4 ,6
0 8
— 2 .5
6 —
2 0 .8
1 —
– 0.
0 8 5
2 5
— 4 ,6
5 6
— 4 .0
1 —
2 0 .7
5 —
– 0.
0 4 5
Z er
o in
ve st
m en
t p o rt
fo lio
(2 U
.. .2
4 )
– (1
U 2 5)
0 .0
3 5
(1 .0
8 )
Ju ly
2 00
7 -J u n e
2 01
5
1 6 ,7
6 8
7 ,1
5 2
– 1 7 .8
4 – 6 .8
0 2 1.
8 8
2 1 .6
6 – 0.
0 0 8
0 .0
2 5
2 6 ,7
3 2
7 ,1
1 6
– 9 .4
4 – 3 .8
5 2 2.
2 7
2 1 .9
8 0 .0
7 3
0 .0
0 3
3 6 ,7
2 0
7 ,1
1 6
– 6 .6
3 – 3 .0
0 2 2.
0 7
2 1 .8
5 0 .0
4 4
0 .0
5 0
4 6 ,7
2 0
7 ,1
0 4
– 4 .7
9 – 2 .4
1 2 2.
4 1
2 1 .7
0 – 0.
0 0 3
– 0.
0 1 6
5 6 ,7
3 2
7 ,1
2 8
– 3 .3
1 – 2 .0
0 2 1.
8 6
2 2 .0
6 – 0.
0 2 1
0 .0
3 2
6 6 ,7
0 8
7 ,1
0 4
– 2 .0
1 – 1 .6
6 2 2.
2 4
2 2 .0
8 – 0.
0 1 4
0 .0
4 6
7 6 ,7
4 4
7 ,1
2 8
– 0 .9
5 – 1 .3
5 2 1.
7 8
2 2 .0
5 – 0.
0 4 7
0 .0
4 4
8 6 ,7
2 0
7 ,1
1 6
0 .0
5 – 1 .0
9 2 2.
0 8
2 2 .6
5 0 .0
2 9
0 .1
3 5
9 6 ,7
2 0
7 ,1
4 0
0 .9
7 – 0 .8
2 2 3.
1 0
2 2 .3
3 0 .0
0 7
– 0.
0 1 7
1 0
6 ,7
3 2
7 ,1
2 8
1 .8
9 – 0 .5
7 2 2.
2 8
2 2 .2
3 0 .0
1 7
0 .0
9 7
1 1
6 ,7
3 2
7 ,0
9 2
2 .6
8 – 0 .3
4 2 1.
9 3
2 2 .4
7 – 0.
0 3 7
0 .0
5 0
1 2
6 ,7
2 0
7 ,1
1 6
3 .4
7 – 0 .1
3 2 2.
6 2
2 2 .2
5 – 0.
0 2 4
0 .0
2 7
1 3
6 ,7
4 4
7 ,1
2 8
4 .3
4 0 .0
8 2 2.
0 7
2 2 .5
7 0 .0
2 9
– 0.
0 0 2
1 4
6 ,7
2 0
7 ,1
2 8
5 .1
2 0 .3
0 2 1.
6 3
2 2 .8
2 0 .0
5 0
0 .0
5 9
1 5
6 ,7
2 0
7 ,1
1 6
5 .9
3 0 .5
4 2 1.
6 8
2 1 .9
7 0 .0
7 7
0 .0
0 4
1 6
6 ,7
2 0
7 ,1
0 4
6 .8
1 0 .7
9 2 2.
7 4
2 2 .1
1 0 .0
1 2
0 .0
0 8
1 7
6 ,7
2 0
7 ,1
1 6
7 .5
8 1 .0
3 2 1.
5 9
2 2 .1
0 0 .0
3 5
0 .0
1 4
1 8
6 ,7
2 0
7 ,1
0 4
8 .3
6 1 .2
9 2 1.
7 5
2 1 .6
7 0 .0
6 2
0 .0
7 9
1 9
6 ,7
4 4
7 ,1
1 6
9 .2
1 1 .5
8 2 1.
9 4
2 2 .0
9 – 0.
0 0 1
– 0.
0 1 3
(c on
tin ue
d)
57
T a b
le 6 . (c
o n ti nu
ed )
Fi rm
m o nt
h C
o ef
fic ie
n t
o n
X (a
n5 )
Si ze
A ve
ra ge
an n u al
re tu
rn
R M
B /U
SD B A SK
E T /R
M B
R M
B /U
SD B A SK
E T /R
M B
R M
B /U
SD B
A SK
E T /R
M B
R M
B /U
SD B A SK
E T /R
M B
2 0
6 ,7
2 0
7 ,1
2 8
1 0 .2
4 1 .9
0 2 1.
6 1
2 1 .8
8 0 .0
1 0
0 .0
5 5
2 1
6 ,7
2 0
7 ,1
0 4
1 1 .4
0 2 .2
7 2 1.
8 2
2 1 .7
6 0 .0
1 7
– 0.
0 0 7
2 2
6 ,7
3 2
7 ,1
1 6
1 2 .7
4 2 .7
4 2 1.
9 8
2 1 .8
1 – 0.
0 2 7
0 .0
5 4
2 3
6 ,7
2 0
7 ,1
2 8
1 4 .4
7 3 .3
4 2 1.
6 4
2 1 .8
1 0 .0
5 3
0 .0
8 6
2 4
6 ,7
3 2
7 ,1
1 6
1 6 .9
5 4 .1
5 2 2.
0 1
2 1 .7
4 0 .0
2 8
0 .0
2 0
2 5
6 ,7
5 6
7 ,1
5 2
2 5 .0
3 6 .5
6 2 1.
4 9
2 1 .7
2 0 .0
5 8
0 .0
0 9
Z er
o in
ve st
m en
t p o rt
fo lio
(2 U
.. .2
4 )
– (1
U 2 5)
0 .0
3 1
(– 2 .8
8 * * )
0 .0
5 6
(– 3 .3
1 * * * )
N ot
e. T
he ta
b le
sh o w
s av
er ag
es o f
fir m
si ze
, se
n si
ti vi
ty to
fo re
ig n -e
x ch
an ge
ra te
, an
d an
n u al
ra w
re tu
rn s
o f
th e
2 5
p o rt
fo lio
s so
rt ed
o n
se n si
ti vi
ty to
fo re
ig n -e
x ch
an ge
ra te
. Fo
r
B A SK
E T /R
M B
ex ch
an ge
ra te
, th
e sa
m p le
b re
ak s
in to
tw o
p er
io d s,
th at
is ,
Ju ly
1 9 9 7
to Ju
n e
2 0 0 6,
an d
Ju ly
2 0 0 7
to Ju
n e
2 0 1 5.
Fo r
R M
B /U
SD ex
ch an
ge ra
te ,
th e
sa m
p le
is fr
o m
Ju ly
2 0 0 7
to Ju
n e
2 0 1 5.
Fi rm
m o nt
h s
fo r
ea ch
p o rt
fo lio
ar e
al so
re p o rt
ed . T
h e
ze ro
-i nv
es tm
en t
p o rt
fo lio
in th
e b o tt
o m
ro w
o f
ea ch
p an
el is
th e
va lu
e- w
ei gh
te d
m o n th
ly re
tu rn
o f
st o ck
s in
Po rt
fo lio
s 2
th ro
u gh
2 4
m in
u s
th e
va lu
e- w
ei gh
te d
re tu
rn o f
th e
st o ck
s in
p o rt
fo lio
s 1
an d
2 5 . T
h e
t st
at is
ti cs
o f
m ea
n co
m p ar
is o n
te st
s o n
th e
av er
ag e
an n u al
ra w
re tu
rn s
o f th
e ze
ro -i nv
es tm
en t
p o rt
fo lio
ar e
re p o rt
ed in
th e
la st
ro w
in ea
ch p an
el (s
ee th
e la
st tw
o co
lu m
n s)
.
* * d en
o te
s st
at is
ti ca
l si
gn ifi
ca n ce
at 5 %
si gn
ifi ca
n ce
le ve
l. * * * d en
o te
s st
at is
ti ca
l si
gn ifi
ca n ce
at 1 %
si gn
ifi ca
n ce
le ve
l.
58
We divide the whole sample into two subperiods to capture the impact of exchange rate
regime switching. For the risk factors constructed by BASKET/RMB, we report the results
for two subsample periods. The first sample period is from July 1997 to June 2006 and the
second sample period is from July 2007 to June 2015.19 Because the exchange rate was
pegged to the USD prior to 2005, for the exchange rate risk factors measured by RMB/
USD, we report the second time period, namely, after the regime change in 2005. The
results show that for the second time period (July 2007 to June 2015) the average annual
return of the combined portfolio containing the two extreme ranks (1st and 25th) is signifi-
cantly higher than that of the remaining portfolios (2nd to 24th) when the exchange rate is
measured either by RMB/USD (at the 5% level) or by BASKET/RMB (at the 1% level).
However, the difference is not significant for the first sample period (July 1997 to June
2006).
Kolari et al. (2008) found that firms most sensitive to exchange rate fluctuations are on
average smaller. In other words, their exchange rate risk factor has a strong correlation
with the size factor. Lewellen, Nagel, and Shanken (2010) stated that the testing asset is
subject to factor structure if any factors added in the asset pricing model are correlated
with the size (or value) factor. To further examine the relations among risk factors, we
report the correlation of exchange rate risk factors with Fama-French’s three factors and
the momentum factor in Table 7. The strongest negative correlation exists between the
Table 7. Correlation Matrix of Fama-French Factors and Foreign-Exchange Risk Factors.
Panel A rm – rf SMB HML MOM
rm – rf 1.00 SMB 0.10 1.00 HML 0.18 0.06 1.00 MOM –0.12 –0.28 –0.44 1.00 BASKET XMI 0.12 0.12 –0.12 0.33 USD XMI 0.04 0.27 –0.45 0.06
Panel B Mean Median SD Sample variance Kurtosis Skewness
BASKET/RMB XMI (July 1997-June 2015)
–0.0029 (–1.25) –0.0041 0.0338 0.0011 5.2182 –0.6880
BASKET/RMB XMI (July 2007-June 2015)
–0.0029 (–0.66) –0.0051 0.0432 0.0019 3.7901 –0.8259
RMB/USD XMI (July 2007-June 2015)
–0.0004 (–0.09) 0.0006 0.0416 0.0017 7.9155 –1.3923
Note. This table reports the correlation coefficients between foreign-exchange risk factors and Fama-French three
factors and momentum factor. The rm – rf is the market risk premium calculated by the value-weighted average
returns of all companies minus the 1-year deposit interest rate. SMB is the return on a portfolio of small stocks
minus the return on a portfolio of large stocks. HML is the return on a portfolio of high book-to-market stocks
minus the return on a portfolio of low book-to-market stocks. MOM is the return on a portfolio of past winner
stocks minus a portfolio of past loser stocks. BASKET/RMB XMI and RMB/USD XMI are the foreign-exchange risk
factor created by a zero-investment portfolio that takes long positions in portfolios ranked 1st and 25th (highly
sensitive) and short positions in portfolios ranked 2nd to 24th (less or nonsensitive), which are formed at end of
June of each year based on the sensitivity of individual firms to changes in the nominal effective exchange rate
(BASKET/RMB) and nominal RMB exchange rate against USD (RMB/USD). For the BASKET XMI, the sample starts in
July 1997, whereas for the USD XMI, the sample starts in July 2007. Numbers in parentheses below the ‘‘Mean’’
are t statistics. All variables end at June 2015.
Hua et al. 59
exchange rate risk factor measured by RMB/USD with the HML factor, which is –0.45.
Furthermore, the correlation coefficients of exchange rate risk factor with the size factor
(SMB) is 0.12 for BASKET/RMB-based XMI and 0.27 for RMB/USD-based XMI, respec-
tively. These correlation levels should not be subject to the factor structure problem raised
by Lewellen et al. (2010).
In this article, we measure the Chinese exchange rate risk factor (XMI) in two ways:
RMB/USD XMI is the USD risk, which is measured by the RMB exchange rate against the
USD. BASKET/RMB is the ‘‘multicurrency’’ risk, which is measured by the ‘‘reference
basket’’ of currencies. As was discussed in previous section, even though China had
pegged RMB’s value to the dollar until July 2005, there were still fluctuations in exchange
rates between RMB and other currencies. It is sensible to separate the two different risks
and test the ‘‘multicurrency’’ exchange risk before 2005. However, most of the Chinese
export transactions are denominated in USDs and account for at least 20% in the trade-
weighted currency basket index, so it is likely that companies are more exposed to USD
exchange rate movements than other currencies even after China switched to a controlled
floating exchange rate regime that is pegged to a reference basket of currencies since July
2005.
Panel B of Table 7 shows the descriptive statistics of exchange rate risk factors (XMIs)
measured by two different exchange rate series. The mean of RMB/USD-based XMI during
the period of July 2007 to June 2015 is negative, but not significant, with a monthly aver-
age return of –0.04%. The mean of BASKET/RMB XMI is also negative for the overall
sample period as well as the second period of July 2007 to June 2015. However, neither of
them is statistically significant. Because the exchange rate risk factors (XMI) are created by
a zero-investment portfolio that takes long positions in portfolios ranked 1st and 25th and
short positions in portfolios ranked 2nd to 24th, our results are broadly consistent with the
findings in Kolari et al. (2008) for the U.S. market that the zero-investment portfolio
returns are negative.
We next estimate the Fama-French–Carhart four-factor and Fama-French three-factor
models along with the XMI factor as in Equations 6 and 7. If XMI is a priced factor, it
should reduce the mean pricing error of asset pricing models, which is represented by the
absolute or squared value of the intercept. Similar to Kolari et al. (2008), we perform the
GRS test. The null hypothesis is that the models’ intercepts are jointly equal to zero. If the
exchange rate risk is priced, the inclusion of XMI factor should reduce the F statistic or
increase the p value, which indicates that by adding the XMI factor the intercepts of the
model are closer to zero in a statistical sense. In addition, the pricing errors measured by
absolute alphas or squared alphas should reduce with the inclusion of the XMI factor if the
exchange rate risk is priced.
Table 8 reports the results of asset pricing tests using two different XMIs. For the
exchange rate risk factors measured by BASKET/RMB, we report the results for two subper-
iods. The first sample period runs from July 1997 to June 2006 and the second sample
period is from July 2007 to June 2015. We use this setting because the exchange rate
regime change occurred in July 2005 and we require 2 years of data to calculate coeffi-
cients’ loadings on the pricing factors. The model’s intercepts in Equations 6 and 7 are
denoted as alpha. The absolute alpha and squared alpha reported in the table are average
values of 25 estimations.
The results in Table 8 show that the F statistic and alphas reduced markedly in both the
three- and four-factor models for the period from July 2007 to June 2015 when XMI mea-
sured by RMB/USD is included. For example, when XMI measured by RMB/USD is
60 Journal of Accounting, Auditing & Finance
T a b
le 8 .
A ss
et P ri
ci n g
Te st
o f th
e Fo
re ig
n -E
x ch
an ge
R is
k Fa
ct o r.
X M
I m
ea su
re d
b y
B A SK
E T /R
M B
(J u ly
1 9 9 7 -J u ne
2 0 0 6 )
X M
I m
ea su
re d
b y
B A SK
E T /R
M B
(J u ly
2 0 0 7-
Ju n e
2 0 1 5)
M o d el
G R
S F
st at
is ti c
G R
S
p va
lu e
A b so
lu te
al p h a
Sq u ar
ed
al p ha
(3 1 0–
6 )
V ar
ia n ce
o f al
p h a
(3 1 0
– 6 )
G R
S
F st
at is
ti c
G R
S
p va
lu e
A b so
lu te
al p h a
Sq u ar
ed
al p h a
(3 1 0
– 6 )
V ar
ia n ce
o f
al p h a
(3 1 0
– 6 )
3 fa
ct o r
0 .7
1 6
0 .8
6 0 .0
0 17
4 .4
6 4 .5
2 0 .2
42 0 .9
9 0 .0
0 23
6 .8
8 7 .0
9
3 fa
ct o r
+ X
M I
0 .7
1 6
0 .8
6 0 .0
0 17
(0 .5
1) 4 .9
7 (0
.6 7)
5 .0
8 0 .2
38 0 .9
9 0 .0
0 19
(– 1 .8
2 * )
5 .4
8 (–
1 .6
3 )
5 .6
0
4 fa
ct o r
0 .6
9 9
0 .8
8 0 .0
0 17
4 .5
3 4 .5
8 0 .3
02 0 .9
9 0 .0
0 49
1 1.
9 1 2 .4
4 fa
ct o r
+ X
M I
0 .6
8 8
0 .8
9 0 .0
0 18
(0 .5
0) 5 .1
8 (0
.7 4)
5 .2
9 0 .2
90 0 .9
9 0 .0
0 14
(– 1 .7
5 * )
7 .6
1 (–
1 .2
3 )
7 .6
4
X M
I m
ea su
re d
b y
R M
B /U
SD (J u ly
2 0 0 7 -J u ne
2 0 1 5 )
3 fa
ct o r
0 .6
21 0 .9
4 0 .0
0 36
1 6.
6 1 6 .0
3 fa
ct o r
+ X
M I
N A
0 .6
08 0 .9
4 0 .0
0 28
(– 2 .8
5 * * * )
1 2.
3 (–
2 .4
4 * * )
1 1 .4
4 fa
ct o r
0 .5
55 0 .9
7 0 .0
0 49
6 2.
7 6 5 .2
4 fa
ct o r
+ X
M I
N A
0 .5
43 0 .9
7 0 .0
0 22
(– 2 .0
7 * * )
1 6.
2 (–
1 .0
6 )
1 5 .2
N ot
e. T
h is
ta b le
re p o rt
s th
e es
ti m
at io
n re
su lt s
fo r
ad d in
g th
e fo
re ig
n -e
x ch
an ge
ri sk
fa ct
o r
in th
e Fa
m a-
Fr en
ch th
re e-
fa ct
o r
m o d el
s an
d th
e Fa
m a-
Fr en
ch – C
ar h ar
t fo
ur -f
ac to
r
m o d el
s.
R it �
R ft
� � =
a i +
b i1
R m
t �
R ft
� � +
b i2 SM
B t +
b i3 H
M L t
+ b
i4 X
M I t
+ e it
,
R it �
R ft
� � =
a i +
b i1
R m
t �
R ft
� � +
b i2 SM
B t +
b i3 H
M L t
+ b
i4 M
O M
t +
b i5 X
M I t
+ e it
,
R it
is th
e va
lu e-
w ei
gh te
d av
er ag
e re
tu rn
fo r
2 5
p o rt
fo lio
s b as
ed o n
th e
se n si
ti vi
ty o f
in d iv
id u al
fir m
s to
th e
ch an
ge s
o f
ex ch
an ge
ra te
s (e
.g .,
R M
B /U
SD an
d B A SK
E T /R
M B ).
R m
t is
th e
va lu
e- w
ei gh
te d
av er
ag e
re tu
rn o f
al l
co m
p an
ie s
in th
e sa
m p le
. R
ft is
th e
1 -y
ea r
d ep
o si
t in
te re
st ra
te in
m o n th
t. SM
B is
th e
re tu
rn o n
a p o rt
fo lio
o f
sm al
l st
o ck
s m
in u s
th e
re tu
rn o n
a p o rt
fo lio
o f
la rg
e st
o ck
s. H
M L
is th
e re
tu rn
o n
a p o rt
fo lio
o f
h ig
h b o o k-
to -m
ar ke
t st
o ck
s m
in u s
th e
re tu
rn o n
a p o rt
fo lio
o f
lo w
b o o k-
to -m
ar ke
t st
o ck
s. M
O M
is
th e
re tu
rn o n
a p o rt
fo lio
o f
p as
t w
in n er
st o ck
s m
in u s
a p o rt
fo lio
o f
p as
t lo
se r
st o ck
s. X
M I t
is th
e fo
re ig
n -e
x ch
an ge
ri sk
fa ct
o r
cr ea
te d
b y
a ze
ro -i nv
es tm
en t
p o rt
fo lio
th at
ta ke
s
lo n g
p o si
ti o n s
in p o rt
fo lio
s ra
n ke
d 1 st
an d
2 5 th
(h ig
h ly
se n si
ti ve
) an
d sh
o rt
p o si
ti o n s
in p o rt
fo lio
s ra
n ke
d 2 n d
to 2 4 th
(l es
s o r
n o ns
en si
ti ve
). T
h e
ta b le
sh o w
s th
e G
ib b o ns
, R
o ss
,
an d
Sh an
ke n ’s
(1 9 8 9 )
F st
at is
ti c
(G R
S) fo
r te
st in
g th
e n u ll
hy p o th
es is
th at
th e
m o d el
’s in
te rc
ep ts
ar e
jo in
tl y
eq u al
to ze
ro fo
r al
l 2 5
ex ch
an ge
ra te
-s en
si ti ve
p o rt
fo lio
s. T
h e
p
va lu
es o f
th e
G R
S F
te st
ar e
re p o rt
ed in
th e
co lu
m n
n ex
t to
th e
G R
S F
st at
is ti c.
A b so
lu te
al p h a
is th
e m
ea n
o f
th e
ab so
lu te
va lu
e o f
th e
m o d el
’s in
te rc
ep ts
o f
th e
2 5
ex ch
an ge
ra te
-s en
si ti ve
p o rt
fo lio
s. Sq
ua re
d al
p h a
is th
e m
ea n
o f
sq u ar
ed in
te rc
ep ts
o f
th e
2 5
ex ch
an ge
ra te
-s en
si ti ve
p o rt
fo lio
s. T
he ab
so lu
te o r
sq u ar
ed va
lu es
o f
th e
in te
rc ep
ts
re p re
se nt
th e
m ea
n p ri
ci n g
er ro
rs .
V ar
ia n ce
o f
al p h a
is th
e m
ea n
o f
sq u ar
ed d iff
er en
ce s
b et
w ee
n th
e al
p h a
an d
th e
av er
ag e
o f
re gr
es si
o n
in te
rc ep
ts .
T h e
n u m
be rs
in th
e
p ar
en th
es es
ar e
th e
t st
at is
ti c
fo r
te st
in g
th e
d iff
er en
ce s
in ab
so lu
te al
p h as
(o r
sq u ar
ed al
ph as
) b et
w ee
n th
e m
o d el
s w
it h
an d
w it h o u t
th e
X M
I fa
ct o r.
T h e
n u ll
hy p o th
es is
is th
at
th e
d iff
er en
ce o f
ab so
lu te
(s q u ar
ed )
al p h as
b et
w ee
n m
o d el
s w
it h
an d
w it h o u t
X M
Is is
eq u al
to ze
ro . In
th e
ca se
o f
B A SK
E T
X M
I, th
e sa
m p le
b re
ak s
in to
tw o
p er
io d s
(J u ly
1 9 9 7
to Ju
ly 2 0 0 6
an d
Ju ly
2 0 0 7
to Ju
n e
2 0 1 5)
. In
th e
ca se
o f
U SD
X M
I, th
e sa
m p le
is fr
o m
Ju ly
2 0 0 7
to Ju
n e
2 0 1 5 . T
h is
is b ec
au se
th e
ex ch
an ge
ra te
w as
st ri
ct ly
p eg
ge d
to th
e U
SD ,
th e
p ri
ci n g
te st
fo r
U SD
X M
I is
th er
ef o re
n o t
av ai
la b le
p ri
o r
to 2 0 0 7.
61
included in the three-factor models, the decreases of average absolute and squared alphas
are statistically significant at the 1% and 5% levels, respectively. When XMI measured by
RMB/USD is included in the four-factor models, the decreases of average absolute alphas
are statistically significant at the 5% levels. The results suggest that during the period of
the controlled floating exchange rate regime that is pegged to a ‘‘reference basket’’ of cur-
rencies, the exchange rate risk exposure to USD fluctuation is priced in Chinese stock
returns. This is not surprising given that the USD carries the largest weight in the currency
basket. In addition, the majority of international trade in China is denominated in the USD
and most of the foreign reserves are in the USD.
For XMI measured by BASKET/RMB, the average absolute alpha significantly decreased
at the 10% level when XMI is included as an additional factor in asset pricing tests for the
second subperiod (July 2007 to June 2015), but not for the first subperiod (July 1997 to
June 2006). These results suggest that reference basket-based foreign-exchange risk was
priced in asset returns after the July 2005 exchange rate regime change. Overall, asset pric-
ing test results show that the exchange rate risk is not priced in equity returns under the
exchange regime that is strictly pegged to the USD. However, the exchange rate risk, either
measured with the USD or measured with the basket of currencies, is priced under the con-
trolled floating regime that is pegged to an index of the currency basket.
Hedging of Exchange Rate Risk Exposure
Our findings that exchange risk exposures are priced in stock returns suggest that compa-
nies should hedge their currency exposure. We therefore examine if companies have ever
increased their hedging activities through derivatives trading.20 The data on the holdings of
derivative assets began available in 2007, which makes it impossible to compare the
change in derivatives holding before and after the exchange rate regime change in 2005.
To examine the hedging of currency risk, we hand collected the use of currency derivatives
from annual reports of the companies that hold derivatives assets.
Appendix B reports the information on currency-related derivative assets. Although only
a small percentage of companies in China hold derivatives assets, the percentage is increas-
ing. For example, in 2007, only 14 companies out of 1485 companies, or 0.94% of the
stock exchange listed companies, held derivative assets. This number increased to 24 com-
panies out of 2,472, or 0.97% of listed companies in 2013. In 2014, the number of compa-
nies using derivative assets jumped up to 52 out of 2,593, or 2.0% of listed companies.
Prior to 2014, among the companies who use derivative assets, mass majority were finan-
cial institutions. However, the pattern seems to have changed in recent years. For example,
in 2014, out of 52 companies using derivative assets, 24 companies are nonfinancial com-
panies. The summary statistic shows that among the companies that use derivatives assets,
the percentage of currency derivatives account for over 70% of the fair value of the total
derivatives traded.
Figure 3 shows the trend in derivatives holdings by Chinese companies. The figure
shows the quarterly weighted derivative holdings and the weighted number of companies
that hold derivatives assets. The weight for each quarter is equal to the total assets of the
companies using derivatives in this quarter divided by the total assets of all companies
using derivatives over the entire sample period. Clearly, both the derivatives holdings (left
scale) and the number of companies (right scale) using derivatives have been steadily
increasing since 2007. This suggests that companies with exchange rate risk exposure are
cognizant of the risk and have increasingly engaged in the hedging activities.
62 Journal of Accounting, Auditing & Finance
Transmission Channels and Pricing of Exchange Rate Risk
The asset pricing test results in the previous section show that exchange rate risk exposure
is priced under the controlled floating regime that is pegged to a trade-weighted index of
currency basket. To examine what may have affected the strength of the pricing, we
attempt to identify the economic channels that can transmit the exchange rate risk exposure
to asset risk premium under the controlled floating exchange rate regime in China. In par-
ticular, we examine whether the pricing of exchange rate risk is related to the firm’s sensi-
tivity to the economic channels that transmit exchange rate risk exposure to asset risk
premiums. To that end, we first measure the sensitivities of stock returns to the three trans-
mission channels discussed in the section ‘‘Economic Channels Transmitting Exchange
Rate Exposure to Asset Risk Premium’’ and then perform asset pricing tests on the portfo-
lios sorted on the sensitivity to transmission channels. Given the results in the section
‘‘Pricing of Exchange Rate Risk in Cross-Section of Stock Returns,’’ we conduct the test
for the sample from July 2005 to June 2015, the period with the controlled floating rate
regime.
Sensitivities of Stock Returns to Transmission Channels
We first examine the sensitivity of stock returns to the transmission channels. If the stock
returns strongly react to the three transmission channels, then those channels should have
an impact on stock price variation when there is exchange rate movement. We use panel
regression with cross-sectional portfolio fixed effects to test the contemporaneous impact
of exchange rate movements and three transmission channels on stock returns.21 We make
the following specification for the panel regressions:
Figure 3. The derivative holdings of listed companies. Note. The figure shows quarterly weighted derivative holdings and number of companies that use derivatives. The
weight for each given quarter is the ratio of the total assets of the companies using derivatives in this quarter to
the total assets of all companies using derivatives over the entire sample period. The sample period spans from the
first quarter of 2007 (when the derivative holding data becomes available) to the second quarter of 2015. The left
scale is for the weighted derivatives holdings, with unit of 1 billion RMB. The right scale is for the weighted number
of companies.
Hua et al. 63
Rit = vi + d1Xi, t + d2TCi, t + ei, t, ð8Þ
where Rit represents returns on 25 exchange rate-sensitive portfolios and Xi,t is the percent-
age change of exchange rates measured by BASKET/RMB and RMB/USD. TCi,t represents
three transmission channel variables that include percentage changes in international trade,
hot money, and local credit supply. The estimated exposure coefficients d1 and d2 represent
the return sensitivity to the changes in exchange rate and the transmission channels, respec-
tively. Cross-sectional fixed effects are also included and are denoted by vi. The ei, t is the
stochastic error term which is generally allowed to be serially correlated. Table 9 reports
the results of the panel ordinary least squares (OLS) regressions for four combinations of
different variables.
The results in Panel A of Table 9 suggest that returns on the 25 portfolios sorted by
foreign-exchange rate sensitivity are sensitive to international trade, hot money, and local
credit supply (broad money supply). As expected, the coefficient of exchange rate risk is
Table 9. Return Sensitivities to Transmission Channels.
Independent variables Model 1 Model 2 Model 3 Model 4
Panel A: Dependent variable: Returns on foreign-exchange rate-sensitive portfolios
Constant 0.021 (0.002***) 0.010 (0.003***) 0.019 (0.002***) 0.008 (0.003)
BASKET/RMB –2.533 (0.154***) –2.523 (0.153***) — —
RMB/USD — — 2.203 (0.518***) 2.046 (0.518***)
TRADE –0.073 (0.015 ***) –0.093 (0.015***) –0.046 (0.015***) –0.067 (0.016***)
HM –0.003 (0.001***) –0.003 (0.001***) –0.003 (0.001***) –0.003 (0.001***)
M2 — 0.838 (0.172***) — 0.827 (0.180***)
Panel B : Dependent variable: Portfolio returns on ‘‘H’’ share stocks with corresponding ‘‘A’’ shares
Constant 0.074 (2.45**) 0.103 (2.14**) 0.042 (1.09) 0.070 (1.18)
BASKET/RMB –10.645 (–6.47***) –10.670 (–6.50***) — —
RMB/USD — — 1.250 (0.14) 1.650 (0.18)
TRADE 0.100 (0.40) 0.154 (0.62) 0.210 (0.84) 0.261 (1.10)
HM 0.005 (0.82) 0.005 (0.79) 0.007 (0.97) 0.007 (0.95)
M2 — –2.22 (–0.81) — –2.104 (–0.72)
Panel C: Dependent variable: Portfolio returns on ‘‘H’’ share stocks without corresponding ‘‘A’’ shares
Constant –0.016 (–0.54) 0.014 (0.31) –0.018 (–0.51) 0.013 (0.29)
BASKET/RMB –2.440 (–1.03) –2.465 (–1.02) — —
RMB/USD — — 2.535 (0.33) 2.976 (0.39)
TRADE 0.086 (0.34) 0.142 (0.56) 0.112 (0.45) 0.168 (0.67)
HM –0.007 (–1.56) –0.008 (–1.21) –0.007 (–1.02) –0.007 (–1.06)
M2 — –2.283 (–0.89) — –2.32 (–0.92)
Note. In this table, Panel A reports the estimation results from the following panel OLS regression:
Rit = vi + d1Xit + d2TCit + eit,
where Rit represents 25 foreign-exchange sensitive portfolio return series and Xit is percentage change of exchange
rates, expressed in BASKET/RMB and RMB/USD, respectively. TCit represents three transmission channel variables,
which include percentage changes of international trade (TRADE), hot money (HM), and credit expansion (M2).
Panel B and C report the estimation results from the following OLS regression:
rHt = vt + dH1Xt + dH2TCt + et,
where rHt represents portfolio return series of ‘‘H’’ shares with corresponding ‘‘A’’ shares and ‘‘H’’ shares without
corresponding ‘‘A’’ shares. Definitions of Xt and TCt are the same as Xit and TCit.
Numbers in parentheses are t statistics. The sample is from July 2005 to June 2015.
**denotes statistical significance at 5% significance level. ***denotes statistical significance at 1% significance level.
64 Journal of Accounting, Auditing & Finance
positive for RMB/USD and negative for BASKET/RMB series. This suggests that RMB
appreciation on average is associated with decreases in the stock returns. The coefficients
are all statistically significant at the 1% significance level. The coefficient of international
trade variable is generally negative and significant at the 1% significance level, suggesting
that a currency appreciation would harm an export-oriented company’s earnings through
declining international trade. The coefficient of hot money is significantly negative. These
results together suggest that international trade expansion and the increase in hot money
generally lead to a negative impact on stock returns. Empirical evidence here indicates that
Chinese investors have interpreted the changes of speculative cash inflows as a negative
signal to the stock market. The coefficient of broad money supply is positive at the 1% sig-
nificance level, implying that credit expansion positively affect market valuation. Overall,
the results show that the three transmission channels significantly affect stock returns.
If Chinese ‘‘A’’ share stock returns have become sensitive to transmission channels after
the reform of controlled floating exchange rate in 2005, we should expect that returns on
Chinese shares listed in Hong Kong stock exchange as ‘‘H’’ shares are not sensitive to the
transmission channels, because ‘‘H’’ shares are denominated in Hong Kong dollars that
pegs to the USD. We thus examine the return sensitivities of ‘‘H’’ share stocks to the trans-
mission channels. Panel B and C in Table 9 report the results for ‘‘H’’ shares with the cor-
responding ‘‘A’’ shares and ‘‘H’’ shares without the corresponding ‘‘A’’ shares,
respectively. Both Panel B and C show that ‘‘H’’ share stock returns are not sensitive to
the three transmission channels. This difference suggests that the change from pegged to
managed floating exchange rate regime affect the return sensitivity to transmission channels
only for the domestic ‘‘A’’ shares. Panel B shows that, similar to the domestic ‘‘A’’ share
return, ‘‘H’’ shares with the corresponding ‘‘A’’ shares traded on Chinese domestic stock
markets is negatively correlated with exchange rate series measured in BASKET/RMB at
the 1% significance level. By contrast, results in Panel C suggests that for ‘‘H’’ shares
without corresponding ‘‘A’’ shares, there is no significant correlation between the ‘‘H’’
share returns and BASKET/RMB exchange rate series, possibly due to the pegged exchange
rate for Hong Kong dollars.
Pricing of Exchange Rate Risk for Portfolios Sorted by Transmission Channels
We further examine if a company whose stock returns are more sensitive to above three
channels are more likely to respond to exchange rate risk. For each stock we first estimate
the following regression using 2 years of rolling data and obtain annual sensitivity un1s
(same as obtaining annual foreign-exchange risk exposure coefficients described in the sec-
tion ‘‘Pricing of Exchange Rate Risk in Cross-Section of Stock Returns’’) to international
trade (TRADE), hot money (HM), and domestic credit supply (M2) based on Equation 9 as
follows.
rnt � Rft
� � = cn + un1TRADE or HM or M2ð Þ+ ent, ð9Þ
where rnt is the log return on stock n in month t and Rft is the 1-year deposit interest rate
at month t. According to the values of sensitivity to TRADE, we sort stocks at end of each
June into three groups: high, medium, and low groups at breakpoints of the top 30%,
middle 40%, and bottom 30%. The procedure is repeated for the hot money and credit
supply channels. We then estimate Fama-French three-factor or momentum four-factor
regressions with foreign-exchange risk being added as shown in Equations 6 and 7. The
Hua et al. 65
dependent variable Rit is the value-weighted portfolio returns on high, medium, and low
groups ranked by sensitivity coefficients to changes in international trade, hot money, and
broad money supply, respectively. Table 10 reports the results. The high and low sensitivity
coefficient groups consist of those stocks with most positive or negative coefficients and
hence have higher sensitivities (in absolute value) to the transmission channels than those
in the medium sensitivity coefficient group.
There are three main findings based on the empirical results reported in Table 10. First,
Panel A reports portfolios sorted by sensitivity coefficients of international trade. Because
the proxy for international trade, variable TRADE, is the monthly percentage change of
total values of export and imports, our focus is therefore on the absolute value of the sensi-
tivity coefficient. Results show that the ‘‘medium’’ group, which has the smallest sensitiv-
ity to international trade in absolute term, does not have significant beta coefficient of the
exchange rate risk factor. On the contrary, the ‘‘low’’ group, which has the most negative
sensitivity to international trade, show significantly positive coefficient of exchange rate
risk factor measured by the RMB/USD series. By contrast, the ‘‘high’’ group, that is, the
stock with most positive sensitivity to international trade, has a marginally significant nega-
tive beta coefficient of foreign-exchange risk factor measured by RMB/USD series. The
evidence thus suggests that stocks that are affected by international trade carry stronger
risk premium on their exposure to foreign-exchange risk.
Second, Panel B in Table 10 report the results on portfolios sorted by hot money sensi-
tivity. The ‘‘low’’ group, which has the most negative sensitivity to hot money, shows posi-
tive beta coefficient of exchange rate risk factor measured by either BASKET/RMB or
RMB/USD. The ‘‘high’’ group, which has the most positive sensitivity to hot money, shows
positive beta coefficient of exchange rate risk factor measured by BASKET/RMB. In con-
trast, the ‘‘medium’’ group, which has the least (in absolute value term) sensitivity to hot
money does not have significant beta coefficient of exchange rate risk factor.
Third, the results based on portfolios sorted by broad money supply (M2) in Panel C
show that the ‘‘low’’ group respond significantly to the exchange rate risk measured by
BASKET/RMB, while the ‘‘high’’ group has a significantly positive beta coefficient of
exchange risk measured by RMB/USD. In contrast, although the ‘‘medium’’ group has a
negative beta coefficient, the magnitude is small. The positive coefficient of foreign-
exchange rate risk measured by BASKET/RMB (RMB/USD) for ‘‘low’’ (‘‘high’’) group sup-
ports the notion that the increase in risk premium for RMB appreciation is stronger for
companies highly sensitive (in absolute value term) to local credit market expansion. For
instance, highly leveraged (thus sensitive to local credit market condition) and export-
dependent firms may particularly experience stronger increase in equity risk premium on
expectation of RMB appreciation. In sum, evidence suggests that the pricing of foreign-
exchange rate risk is more pronounced for companies that are highly sensitive to the three
transmission channels, namely, international trade, hot money and local credit supply.
Conclusion
We examine whether and how exchange rate risks under a pegged as well as controlled
floating regime can be priced in asset returns. We do so by studying the Chinese currency
and stock return data that cover the regime changes in exchange rate determination. Prior
research reported that Asian-Pacific firms have more widespread and significant exposure
to foreign-exchange rate risk than their counterparts in industrialized economies, and the
exchange rate pegs do not alleviate this exposure. It remains unknown, however, whether
66 Journal of Accounting, Auditing & Finance
T a b
le 1 0 .
P ri
ci n g
o f Fo
re ig
n -E
x ch
an ge
R is
k in
th e
P re
se n ce
o f a
Tr an
sm is
si o n
C h an
n el
. W
e fir
st es
ti m
at e
ea ch
st o ck
’s se
n si
ti vi
ty (u
1 )
to ch
an ge
s o f tr
ad e,
h o t
m o n ey
, o r
cr ed
it ex
p an
si o n
o ve
r a
ro lli
n g
2 -y
ea r
p er
io d :
r n t �
R ft
� � =
c n +
u n1
T R
A D
E o r
H M
o r
M 2
ð Þ+
e n t,
w h er
e r n
t is
th e
in d iv
id u al
st o ck
re tu
rn an
d R
ft is
th e
1 -y
ea r
d ep
o si
t in
te re
st ra
te in
m o n th
t. T
R A
D E
(o r
H M
o r
M 2 )
is th
e p er
ce n ta
ge ch
an ge
o f in
te rn
at io
n al
tr ad
e (o
r h o t
m o n ey
o r
cr ed
it ex
p an
si o n )
in m
o n th
t. St
o ck
s ar
e so
rt ed
in to
h ig
h , m
ed iu
m , an
d lo
w gr
o u p s
ac co
rd in
g to
th e
va lu
es o f
u n1
at th
e en
d o f
Ju n e
o f ea
ch ye
ar at
b re
ak p o in
ts 3 0 th
an d
7 0 th
p er
ce n ti le
s. W
e n ex
t es
ti m
at e
Fa m
a- Fr
en ch
th re
e- fa
ct o r
m o d el
s an
d Fa
m a-
Fr en
ch – C
ar h ar
t fo
u r-
fa ct
o r
m o d el
s w
it h
X M
I in
cl u d ed
: R
it �
R ft
� � =
a i +
b i1
R m
t �
R ft
� � +
b i2 SM
B t +
b i3 H
M L t
+ b
i4 X
M I t
+ e i
t,
R it �
R ft
� � =
a i +
b i1
R m
t �
R ft
� � +
b i2 SM
B t +
b i3 H
M L t
+ b
i4 M
O M
t +
b i5 X
M I t
+ e it
,
R it
is th
e va
lu e-
w ei
gh te
d av
er ag
e re
tu rn
s o n
th re
e p o rt
fo lio
s b as
ed o n
th e
se n si
ti vi
ty o f
in d iv
id u al
fir m
s to
th e
ch an
ge s
in in
te rn
at io
n al
tr ad
e, h o t
m o n ey
, o r
cr ed
it ex
p an
si o n . R
m t is
th e
va lu
e- w
ei gh
te d
av er
ag e
re tu
rn o f al
l co
m p an
ie s
in th
e sa
m p le
. R
ft is
th e
1 -y
ea r
d ep
o si
t in
te re
st ra
te in
m o n th
t. SM
B is
th e
re tu
rn o n
a p o rt
fo lio
o f
sm al
l st
o ck
s m
in u s
th e
re tu
rn o n
a p o rt
fo lio
o f
la rg
e st
o ck
s. H
M L
is th
e re
tu rn
o n
a p o rt
fo lio
o f
h ig
h b o o k-
to -m
ar ke
t st
o ck
s m
in u s
th e
re tu
rn o n
a p o rt
fo lio
o f
lo w
b o o k-
to -m
ar ke
t st
o ck
s. M
O M
is th
e re
tu rn
o n
a p o rt
fo lio
o f
p as
t w
in n er
st o ck
s m
in u s
a p o rt
fo lio
o f
p as
t lo
se r
st o ck
s. X
M I
is th
e fo
re ig
n -e
x ch
an ge
ri sk
fa ct
o r
b as
ed o n
a ze
ro -i nv
es tm
en t
p o rt
fo lio
th at
ta ke
s lo
n g
p o si
ti o n s
in p o rt
fo lio
s ra
n ke
d 1 st
an d
2 5 th
an d
sh o rt
p o si
ti o n s
in p o rt
fo lio
s ra
n ke
d 2 n d
to 2 4 th
. Tw
o fo
re ig
n -e
x ch
an ge
ri sk
fa ct
o rs
(f ro
m B
A SK
E T /R
M B
an d
R M
B /U
SD )
ar e
u se
d in
th e
es ti m
at io
n .
M o d el
H ig
h M
ed iu
m Lo
w
a b
i4 b
i5 ð Þ
a b
i4 b
i5 ð Þ
a b
i4 b
i5 ð Þ
P an
el A
: Po
rt fo
lio s
so rt
ed b y
se n si
ti vi
ty to
T R A D
E
FF M
o d el
3 fa
ct o r
2 .6
1 3
1 0
– 4
(0 .1
4) —
0 .0
0 1 3
(0 .8
5 )
— – 0.
0 0 5 1
(– 1 .3
9 )
—
4 fa
ct o r
9 .6
3 1 0–
4 (0
.5 0 )
— 0 .0
0 1 6
(0 .9
5 )
— – 0.
0 0 5 4
(– 1 .8
4 * )
—
B A SK
E T /R
M B
X M
I 3
fa ct
o r
+ X
M I
2 .7
4 3
1 0
– 4
(0 .1
5) 0 .0
0 53
(0 .1
0) 0 .0
0 1 4
(0 .8
3 )
– 0 .0
09 1
(– 0 .2
0 )
– 0.
0 0 3 2
(– 1 .2
9 )
0 .1
0 55
(1 .4
9)
4 fa
ct o r
+ X
M I
0 .0
0 10
(0 .5
2) 0 .0
1 36
(0 .2
5) 0 .0
0 1 6
(0 .9
3 )
– 0 .0
06 4
(– 0 .1
4 )
– 0.
0 0 4 3
(– 1 .7
8 * )
0 .0
9 22
(1 .3
0)
R M
B /U
SD X
M I
3 fa
ct o r
+ X
M I
5 .0
4 3
1 0
– 4
(0 .2
7) – 0.
1 0 5 0
(– 1 .6
5 * )
0 .0
0 1 3
(0 .8
1 )
0 .0
2 8
(0 .5
0) – 0.
0 0 4 1
(– 1 .7
9 * )
0 .3
0 54
(3 .7
9* * * )
4 fa
ct o r
+ X
M I
0 .0
0 12
(0 .6
1) – 0.
1 0 2 0
(– 1 .6
0 )
0 .0
0 1 6
(0 .9
1 )
0 .0
2 92
(0 .5
2) – 0.
0 0 5 3
(– 2 .2
1 * * )
0 .3
0 01
(3 .7
6* * * )
P an
el B :P
o rt
fo lio
s so
rt ed
b y
se n si
ti vi
ty to
H o t
M o n ey
(H M
)
FF M
o d el
3 fa
ct o r
– 0.
0 0 2 4
(– 1 .2
2 )
— 0 .0
0 1 7
(1 .0
9 )
— – 0.
0 0 3 2
(– 1 .0
6 )
—
4 fa
ct o r
– 0.
0 0 2 1
(– 1 .0
2 )
— 0 .0
0 1 9
(1 .1
2 )
— – 0.
0 0 3 6
(– 1 .0
6 )
—
B A SK
E T /R
M B
X M
I 3
fa ct
o r
+ X
M I
– 0.
0 0 2 1
(– 1 .0
9 )
0 .1
1 61
(2 .0
7* * )
0 .0
0 1 6
(0 .9
8 )
– 0 .0
71 5
(– 1 .5
6 )
– 0.
0 0 2 0
(– 0 .9
4 )
0 .1
1 19
(1 .8
3* )
4 fa
ct o r
+ X
M I
– 0.
0 0 1 6
(– 0 .8
1 )
0 .1
2 15
(2 .1
5* * )
0 .0
0 1 6
(0 .9
7 )
– 0 .0
70 8
(– 1 .5
3 )
– 0.
0 0 2 4
(– 1 .0
5 )
0 .1
0 79
(1 .7
4* )
R M
B /U
SD X
M I
3 fa
ct o r
+ X
M I
– 0.
0 0 2 4
(– 1 .2
2 )
0 .0
0 54
(0 .0
8) 0 .0
0 1 8
(1 .1
2 )
– 0 .0
02 4
(– 0 .4
3 )
– 0.
0 0 2 8
(– 1 .3
3 )
0 .2
0 83
(2 .9
0* * * )
4 fa
ct o r
+ X
M I
– 0.
0 0 2 1
(– 1 .0
2 )
0 .0
0 68
(0 .1
0) 0 .0
0 1 9
(1 .1
4 )
– 0 .0
23 2
(– 0 .4
1 )
– 0.
0 0 3 2
(– 1 .4
0 )
0 .2
0 64
(2 .8
6* * * )
(c on
tin ue
d)
67
T a b
le 1 0 . (c
o n ti n u ed
)
M o d el
H ig
h M
ed iu
m Lo
w
a b
i4 b
i5 ð Þ
a b
i4 b
i5 ð Þ
a b
i4 b
i5 ð Þ
P an
el C
: Po
rt fo
lio s
so rt
ed b y
se n si
ti vi
ty to
D o m
es ti c
C re
d it
Su p pl
y (M
2 )
FF M
o d el
3 fa
ct o r
4 .6
1 3
1 0
– 4
(0 .2
4) —
4 .1
5 3
1 0
– 5
(0 .0
3 )
— – 0.
0 0 2 2
(– 1 .1
4 )
—
4 fa
ct o r
3 .5
6 3
1 0
– 4
(0 .1
8) —
0 .0
0 0 9
(0 .6
7 )
— – 0.
0 0 3 1
(– 1 .5
5 )
—
B A SK
E T /R
M B
X M
I 3
fa ct
o r
+ X
M I
5 .5
9 3
1 0
– 4
(0 .2
9) 0 .0
3 91
(0 .7
1) – 1 .7
0 3
1 0
– 4
(– 0 .1
3 )
– 0 .0
85 5
(– 2 .3
0 * * )
– 0.
0 0 1 7
(– 0 .9
4 )
0 .2
0 44
(3 .9
6* * * )
4 fa
ct o r
+ X
M I
5 .0
2 3
1 0
– 4
(0 .2
5) 0 .0
3 84
(0 .6
9) 6 .1
1 3
1 0
– 4
(0 .4
5 )
– 0 .0
76 7
(– 2 .0
8 * * )
– 0.
0 0 2 3
(– 1 .2
5 )
0 .1
9 68
(3 .8
0* * * )
R M
B /U
SD X
M I
3 fa
ct o r
+ X
M I
– 6.
4 6
3 1 0
– 5
(– 0 .0
4 )
0 .2
2 62
(3 .6
7* * * )
6 .8
9 3
1 0
– 5
(0 .0
5 )
– 0 .0
11 8
(– 0 .2
6 )
– 0.
0 0 2 3
(– 1 .1
7 )
0 .0
3 13
(0 .4
7)
4 fa
ct o r
+ X
M I
– 9
3 1 0
– 5
(– 0 .0
5 )
0 .2
2 61
(3 .6
4* * * )
9 .1
8 3
1 0
– 4
(0 .6
8 )
– 0 .0
07 9
(– 0 .1
7 )
– 0.
0 0 3 2
(– 1 .5
7 )
0 .0
2 72
(0 .4
1)
N ot
e. T
hi s
ta b le
re p o rt
s al
p h a,
w h ic
h is
th e
m o d el
in te
rc ep
t, an
d b
i4 (b
i5 ),
th e
ri sk
p re
m iu
m o f
th e
fo re
ig n -e
x ch
an ge
ri sk
fa ct
o r.
T h e
sa m
p le
p er
io d
is fr
o m
Ju ly
2 0 0 7
to Ju
n e
2 0 1 5.
N u m
b er
s in
p ar
en th
es es
ar e
t st
at is
ti cs
.
* * d en
o te
s st
at is
ti ca
l si
gn ifi
ca n ce
at 5 %
si gn
ifi ca
n ce
le ve
l. * * * d en
o te
s st
at is
ti ca
l si
gn ifi
ca n ce
at 1 %
si gn
ifi ca
n ce
le ve
l.
68
such exchange rate risk exposures under a pegged or controlled floating exchange rate
regime can be priced in asset returns.
Chinese RMB exchange rate determination has gone through a few major changes from
a strictly USD-pegged regime to a controlled floating mechanism. As the only currency in
the basket of IMF’s SDR that is not market-driven for its exchange rate formation, Chinese
RMB offers an ideal case to study the pricing of exchange rate risk in a pegged or con-
trolled floating regime given the country’s robust economic activities and off-shore trading
of the currency.
We find that under a strictly pegged exchange rate regime, exchange rate risk does not
seem to be priced in asset returns. However, the exchange rate risk is priced during the
period of a controlled floating exchange rate regime. There is a negative stock market reac-
tion to the announcement of the change from a strictly pegged exchange rate to a controlled
floating regime, which subsequently led to a series of appreciation of RMB. In general, due
to China’s export-oriented economy, a currency appreciation adversely affects the econ-
omy; the negative stock market reaction suggests a negative stock valuation effect of
expectation of increased exchange rate risk exposure. We also find that volatilities of earn-
ings, ROA and P/E ratios are significantly greater for ‘‘export’’ companies with larger
exposure to exchange rate risk than for ‘‘nonexport’’ companies after the exchange rate
reform in 2005. In addition, the time series exchange rate movement is significantly corre-
lated with stock returns after the change to the controlled floating regime.
We further find that the pricing of exchange rate risk is more pronounced for stocks that
are more sensitive to international trade, hot money and local credit supply, the three main
economic channels that transmit the exchange rate risk exposure to asset risk premiums.
Our analyses with the ‘‘H’’ shares show that Chinese shares listed on the Hong Kong stock
exchange do not exhibit the same effect to the announcement of exchange rate reform, and
more importantly, they are not sensitive to the three transmission channels due to Hong
Kong’s pegged currency regime. We also find that the use of financial derivatives
by Chinese companies have increased over recent years. Overall, our results suggest that
foreign-exchange rate risks under a controlled floating exchange rate regime can be priced
in asset returns and currency risk hedging under such regime is warranted.
Hua et al. 69
A p
p e n
d ix
A .
In d u st
ry D
is tr
ib u ti o n
o f Li
st ed
C o m
p an
ie s
an d
Pe rc
en ta
ge o f C
o m
p an
ie s
W it h
E x p o rt
s in
Sa le
s.
A gr
i. M
in .
M an
u f.
U ti l.
C o n.
W h o .&
R et
. Tr
an s.
H o t. &
R es
t. I.T
. Fi
n .
R ea
l E st
. R
en t. &
B us
. Sc
ie .&
Te ch
. E nv
i. E d u.
H ea
lt h
E n te
r. G
en .
To ta
l
P an
el A
: N
u m
b er
o f co
m p an
ie s
b y
ye ar
an d
in d us
tr y
1 9 9 5
0 4
1 20
1 8
6 3 9
1 2
3 1 1
6 4 1
3 2
2 2
0 4
7 2 8 0
1 9 9 6
1 1 4
2 05
3 2
8 6 2
1 6
6 1 7
9 6 4
7 3
6 2
2 1 0
1 4
4 7 8
1 9 9 7
6 2 0
3 02
4 4
1 1
7 4
2 3
7 2 0
1 7
7 8
1 3
3 9
2 2
1 4
1 5
6 6 0
1 9 9 8
7 2 3
3 64
5 1
1 4
7 9
2 8
7 2 0
1 8
8 3
1 4
3 1 1
2 2
1 6
1 6
7 5 8
1 9 9 9
8 2 4
4 19
5 4
1 8
8 2
3 2
7 2 2
1 9
9 5
1 4
3 1 1
2 2
1 6
2 2
8 5 0
2 0 0 0
1 2
3 0
4 96
6 2
2 1
9 6
4 0
8 2 5
2 3
1 0 0
1 5
3 1 3
2 3
1 8
2 2
9 8 9
2 0 0 1
1 3
3 4
5 39
6 5
2 2
1 0 1
4 3
8 2 8
2 4
1 0 6
1 5
3 1 3
2 3
1 8
2 2
1 ,0
59
2 0 0 2
1 5
3 7
5 76
6 9
2 4
1 0 4
4 8
8 3 1
2 6
1 1 0
1 6
3 1 4
2 3
2 0
2 2
1 ,1
28
2 0 0 3
1 7
4 1
6 18
7 1
2 8
1 0 5
5 1
8 3 5
2 7
1 1 1
1 6
3 1 4
2 3
2 1
2 2
1 ,1
93
2 0 0 4
2 0
4 4
6 84
7 8
3 1
1 0 9
5 3
8 3 7
2 7
1 1 3
1 7
3 1 5
2 3
2 1
2 2
1 ,2
87
2 0 0 5
2 1
4 3
6 93
8 0
3 2
1 0 9
5 4
8 3 7
2 7
1 1 3
1 7
3 1 5
2 4
2 1
2 3
1 ,3
02
2 0 0 6
2 2
4 7
7 35
8 0
3 6
1 1 0
6 0
8 4 1
2 8
1 1 3
1 7
2 1 6
2 4
2 1
2 3
1 ,3
65
2 0 0 7
2 3
5 1
7 95
8 0
3 8
1 1 0
6 2
1 0
4 9
3 7
1 1 1
2 2
5 1 7
2 4
2 1
2 3
1 ,4
60
2 0 0 8
2 5
5 4
8 54
7 9
4 0
1 1 4
6 3
1 0
5 5
3 7
1 1 3
2 2
5 1 7
2 4
2 0
2 3
1 ,5
37
2 0 0 9
2 7
5 5
9 40
8 2
4 7
1 1 8
6 7
1 1
7 1
4 0
1 1 9
2 4
7 1 7
2 5
2 3
2 3
1 ,6
78
2 0 1 0
3 5
6 0
1 ,1
9 4
8 4
5 2
1 2 7
7 4
1 1
9 9
4 4
1 1 9
2 8
1 0
1 9
2 5
2 7
2 3
2 ,0
13
2 0 1 1
3 7
6 3
1 ,3
9 2
8 6
6 2
1 3 6
7 5
1 1
1 2 2
4 9
1 2 2
3 0
1 2
2 3
2 6
3 1
2 2
2 ,2
81
2 0 1 2
4 0
6 4
1 ,4
9 3
8 9
6 6
1 3 9
7 9
1 1
1 4 5
5 1
1 2 3
3 1
1 2
2 4
2 7
3 5
2 3
2 ,4
34
2 0 1 3
4 1
6 8
1 ,5
2 6
9 1
6 7
1 3 9
8 0
1 1
1 5 5
5 1
1 2 2
3 4
1 2
2 5
2 7
3 6
2 3
2 ,4
90
2 0 1 4
4 1
7 0
1 ,6
0 3
9 3
6 9
1 4 1
8 0
1 0
1 5 8
5 3
1 2 4
3 4
1 9
2 8
2 7
3 8
2 3
2 ,5
93
(c on
tin ue
d)
70
A p
p e n
d ix
A . (c
o n ti nu
ed )
A gr
i. M
in .
M an
u f.
U ti l.
C o n.
W h o .&
R et
. Tr
an s.
H o t. &
R es
t. I.T
. Fi
n .
R ea
l E st
. R
en t. &
B us
. Sc
ie .&
Te ch
. E nv
i. E d u.
H ea
lt h
E n te
r. G
en .
To ta
l
P an
el B :P
er ce
n ta
ge o f co
m p an
ie s
w it h
ex p o rt
re ve
n u e
(n u m
b er
s ar
e in
d ec
im al
p o in
ts )
w it h in
ea ch
in d u st
ry b y
ye ar
2 0 0 1
0 .1
5 0 .0
3 0 .0
8 0 .0
0 0 .0
5 0 .0
5 0 .0
5 0 .0
0 0 .0
0 0 .0
8 0 .0
3 0 .0
7 0 .0
0 0 .0
8 0 .0
0 0 .3
3 0 .1
1 0 .0
5 0 .0
6
2 0 0 2
0 .2
0 0 .1
4 0 .2
5 0 .0
4 0 .1
3 0 .0
6 0 .0
4 0 .0
0 0 .1
0 0 .0
8 0 .1
1 0 .1
9 0 .0
0 0 .0
7 0 .0
0 0 .3
3 0 .1
5 0 .2
3 0 .1
7
2 0 0 3
0 .1
8 0 .2
4 0 .3
4 0 .0
6 0 .2
1 0 .1
3 0 .0
6 0 .0
0 0 .1
1 0 .1
5 0 .1
3 0 .1
9 0 .0
0 0 .0
0 0 .0
0 0 .3
3 0 .1
9 0 .3
2 0 .2
4
2 0 0 4
0 .2
0 0 .2
3 0 .3
8 0 .0
6 0 .1
9 0 .2
0 0 .0
6 0 .0
0 0 .1
4 0 .1
5 0 .1
2 0 .2
4 0 .0
0 0 .0
0 0 .5
0 0 .3
3 0 .1
9 0 .4
1 0 .2
7
2 0 0 5
0 .1
4 0 .2
6 0 .4
1 0 .0
6 0 .1
6 0 .2
0 0 .0
6 0 .0
0 0 .1
4 0 .1
9 0 .1
5 0 .2
9 0 .0
0 0 .0
0 0 .5
0 0 .5
0 0 .1
9 0 .3
9 0 .2
9
2 0 0 6
0 .2
7 0 .2
6 0 .5
3 0 .1
0 0 .2
2 0 .2
6 0 .0
7 0 .0
0 0 .1
5 0 .1
8 0 .2
4 0 .3
5 0 .0
0 0 .0
0 0 .0
0 0 .5
0 0 .1
9 0 .4
3 0 .3
8
2 0 0 7
0 .3
9 0 .2
5 0 .5
8 0 .1
0 0 .4
5 0 .3
2 0 .1
6 0 .0
0 0 .2
0 0 .3
0 0 .2
1 0 .5
9 0 .2
0 0 .0
0 0 .0
0 0 .2
5 0 .2
9 0 .4
8 0 .4
3
2 0 0 8
0 .4
0 0 .3
1 0 .6
0 0 .1
0 0 .4
8 0 .3
0 0 .1
7 0 .0
0 0 .1
8 0 .3
2 0 .1
9 0 .5
5 0 .2
0 0 .0
0 0 .0
0 0 .5
0 0 .2
5 0 .4
3 0 .4
4
2 0 0 9
0 .3
7 0 .3
6 0 .6
3 0 .1
3 0 .4
9 0 .3
0 0 .1
6 0 .0
0 0 .2
8 0 .3
3 0 .1
4 0 .5
0 0 .4
3 0 .0
0 0 .0
0 0 .4
0 0 .3
0 0 .3
5 0 .4
7
2 0 1 0
0 .4
0 0 .3
8 0 .6
7 0 .1
2 0 .4
8 0 .3
0 0 .1
9 0 .0
0 0 .2
4 0 .3
2 0 .1
4 0 .4
3 0 .5
0 0 .0
5 0 .0
0 0 .4
0 0 .2
6 0 .4
3 0 .5
0
2 0 1 1
0 .4
3 0 .4
1 0 .6
9 0 .1
0 0 .4
4 0 .3
1 0 .1
9 0 .0
0 0 .2
7 0 .3
1 0 .1
1 0 .4
3 0 .4
2 0 .0
9 0 .0
0 0 .3
3 0 .2
6 0 .5
0 0 .5
2
2 0 1 2
0 .4
0 0 .3
6 0 .7
0 0 .1
1 0 .4
1 0 .2
9 0 .1
8 0 .0
0 0 .3
0 0 .2
7 0 .1
1 0 .4
5 0 .3
3 0 .0
8 0 .0
0 0 .4
3 0 .3
1 0 .3
5 0 .5
3
2 0 1 3
0 .3
7 0 .3
4 0 .7
1 0 .1
1 0 .4
0 0 .3
0 0 .1
6 0 .0
0 0 .3
1 0 .2
9 0 .1
1 0 .3
8 0 .4
2 0 .0
4 0 .0
0 0 .4
3 0 .3
1 0 .4
3 0 .5
4
2 0 1 4
0 .4
1 0 .3
6 0 .7
0 0 .1
1 0 .4
5 0 .3
0 0 .1
4 0 .0
0 0 .3
1 0 .3
2 0 .1
0 0 .3
8 0 .4
2 0 .1
4 0 .0
0 0 .4
3 0 .3
2 0 .4
8 0 .5
4
N ot
e. P an
el A
sh o w
s th
e n u m
b er
o f
co m
p an
ie s
in ea
ch in
d u st
ry in
th e
fin al
sa m
p le
. T
h e
in d us
tr ie
s in
cl u d e
A gr
ic u lt u re
, Fo
re st
ry , St
o ck
br ee
d in
g an
d Fi
sh in
g (A
gr i.)
; M
in in
g (M
in .) ;
M an
u fa
ct ur
in g
(M an
uf .) ;
U ti lit
ie s
in cl
u d in
g E le
ct ri
ci ty
, H
ea ti n g,
G as
, an
d W
at er
(U til
.) ;
C o n st
ru ct
io n
(C on
.) ;
W h o le
sa le
an d
R et
ai l
(W ho
.& R et
.) ;
Tr an
sp o rt
at io
n ,
St o ra
ge an
d Po
st
(T ra
ns .) ;
H o te
l an
d R
es ta
u ra
n t
(H ot
.& R es
t. );
In fo
rm at
io n
Te ch
n o lo
gy (I
.T .) ;
Fi n an
ci al
In st
it u ti o n s
(F in
.) ;
R ea
l E st
at e
(R ea
l E st
.) ;
R en
ti n g
an d
B u si
n es
s Se
rv ic
es (R
en t.&
B us
.) ;
Sc ie
n ce
R es
ea rc
h an
d Te
ch n o lo
gy Se
rv ic
es (S
ci e.
& Te
ch .) ; W
at er
C o n se
rv an
cy , E nv
ir o n m
en t
an d
P ub
lic Fa
ci lit
ie s
(E nv
i.) ; E d u ca
ti o n
(E du
.) ; H
ea lt h
an d
So ci
al W
o rk
(H ea
lth );
C u lt u re
, Sp
o rt
an d
E n te
rt ai
n m
en t
(E nt
er .) ; an
d O
th er
In d u st
ri es
(G en
.) . T
h e
C iv
il Se
rv ic
e, M
ai n te
n an
ce , an
d O
th er
se rv
ic es
in d u st
ry ar
e n o t
sh o w
n d ue
to la
ck o f
d at
a. T
h e
in d u st
ry p o rt
fo lio
s
ar e
fo rm
ed at
th e
en d
o f Ju
n e
o f ea
ch ye
ar , t.
Fo r
ex am
p le
, ye
ar 1 9 9 5
re p re
se nt
s th
e p er
io d
fr o m
Ju ly
1 9 9 5
to Ju
n e
1 9 9 6.
C o m
p an
ie s
w it h
le ss
th an
4 m
o n th
s p ri
ce s
ar e
o m
it te
d
fr o m
th at
ye ar
. P an
el B
sh o w
s th
e p er
ce n ta
ge o f
co m
p an
ie s
w it h
ex p o rt
re ve
n u es
w it h in
ea ch
in d us
tr y
fr o m
2 0 0 1
to 2 0 1 4.
T he
d at
a o n
ex p o rt
re ve
n u e
fo r
ea ch
co m
p an
y,
o b ta
in ed
fr o m
ea ch
co m
p an
y’ s
an n u al
fin an
ci al
st at
em en
t, ar
e av
ai la
b le
fr o m
ye ar
2 0 0 1.
71
A p
p e n
d ix
B .
D er
iv at
iv e
H o ld
in gs
o f A
ll Li
st ed
C o m
p an
ie s.
To ta
l n u m
b er
o f co
m p an
ie s
N u m
b er
o f
co m
p an
ie s
h o ld
in g
d er
iv at
iv es
N u m
b er
o f fin
an ci
al In
st it u ti o n s
h o ld
in g
d er
iv at
iv es
N u m
b er
o f
co m
p an
ie s
h o ld
in g
cu rr
en cy
d er
iv at
iv es
To ta
l h o ld
in gs
o f
cu rr
en cy
d er
iv at
iv es
To ta
l h o ld
in gs
o f al
l d er
iv at
iv es
Pe rc
en ta
ge o f cu
rr en
cy d er
iv at
iv es
h o ld
in gs
in th
e to
ta l d er
iv at
iv es
h o ld
in gs
2 0 0 7
1 ,4
8 5
1 4
9 1 1
7 .5
0 1 0 .2
6 0 .7
3 2 0 0 8
1 ,5
6 4
1 8
1 1
1 3
1 1 .3
1 1 3 .6
1 0 .8
3 2 0 0 9
1 ,7
2 4
1 5
1 3
1 2
4 .4
2 5 .4
1 0 .8
2 2 0 1 0
2 ,0
3 1
1 8
1 6
1 5
6 .8
9 9 .4
3 0 .7
3 2 0 1 1
2 ,2
8 2
2 0
1 8
1 8
7 .5
6 1 0 .6
2 0 .7
1 2 0 1 2
2 ,4
3 4
2 2
2 1
2 0
1 1 .6
8 1 3 .4
2 0 .8
7 2 0 1 3
2 ,4
7 2
2 4
2 4
2 3
1 1 .5
9 1 5 .9
0 0 .7
3 2 0 1 4
2 ,5
9 3
5 2
2 8
2 9
1 1 .9
0 1 5 .4
1 0 .7
7
N ot
e. T
h e
ta b le
sh o w
s d er
iv at
iv e
h o ld
in gs
o f co
m p an
ie s
in cl
u d ed
in o u r
sa m
p le
ea ch
ye ar
. T
h e
to ta
l h o ld
in gs
o f cu
rr en
cy d er
iv at
iv es
an d
to ta
l h o ld
in gs
o f al
l d er
iv at
iv es
ar e
in 1 0
b ill
io n
R M
B .
T h e
la st
co lu
m n
sh o w
s th
e p er
ce n ta
ge o f
to ta
l h o ld
in gs
o f
cu rr
en cy
d er
iv at
iv es
in to
ta l
h o ld
in gs
o f
al l
d er
iv at
iv es
ea ch
ye ar
. To
ta l
d er
iv at
iv es
h o ld
in gs
ar e
th e
fa ir
va lu
e o f
th e
d er
iv at
iv es
as se
ts .
A cc
o rd
in g
to th
e A
cc o u n ti n g
St an
d ar
d s
ad o p te
d b y
C h in
a M
in is
tr y
o f
Fi n an
ce in
2 0 1 4,
co m
p an
ie s
sh o u ld
re p o rt
th e
fa ir
va lu
e o f
th ei
r fin
an ci
al
d er
iv at
iv es
as se
ts in
b al
an ce
sh ee
t. T
he fa
ir va
lu e
o f
th e
fin an
ci al
d er
iv at
iv es
in cl
u d ed
in th
e h el
d -f
o r-
tr ad
in g
p o rt
fo lio
s is
b as
ed o n
d ai
ly q u o te
d p ri
ce if
th er
e is
an ac
ti ve
m ar
ke t
fo r
th es
e fin
an ci
al d er
iv at
iv es
. If
fo r
an y
re as
o n
th ei
r q u o te
d p ri
ce is
n o t
av ai
la b le
o n
a gi
ve n
d at
e, th
es e
fin an
ci al
d er
iv at
iv es
ar e
m ea
su re
d u si
n g
m et
h o d s
si m
ila r
to th
o se
u se
d in
o ve
r- th
e- co
u n te
r (O
T C
) m
ar ke
ts . T
h e
fa ir
va lu
e o f O
T C
d er
iv at
iv es
(‘ ‘p
re se
n t
va lu
e’ ’ o r
‘‘t h eo
re ti ca
l p ri
ce ’’)
is eq
u al
to th
e su
m o f fu
tu re
ca sh
flo w
s ar
is in
g fr
o m
th e
in st
ru m
en t,
d is
co u n te
d at
th e
m ea
su re
m en
t d at
e; th
es e
d er
iv at
iv es
ar e
va lu
ed u si
n g
m et
h o d s
re co
gn iz
ed b y
in te
rn at
io n al
fin an
ci al
m ar
ke ts
: th
e ‘‘n
et p re
se nt
va lu
e’ ’
(N P V
) m
et h o d ,
o p ti o n
p ri
ce ca
lc ul
at io
n m
o d el
s, an
d so
o n .
72
Acknowledgments
The authors gratefully acknowledge the thoughtful and constructive comments from Chris Adcock,
Warren Bailey, Vihang Errunza, Ghon Rhee, Trevor Rogers, Bharat Sarath (the editor), Chu Zhang,
an anonymous referee, and seminar participants at University of Hawaii at Manoa, Deakin University,
Hunan University, Jinan University, Nottingham University Business School China. They are grateful
to Menglong Yang and Mingyuan You for their research assistance.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/
or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or
publication of this article: Wei Huang gratefully acknowledges financial support from John and Sue
Dean professorship at the Shidler College of Business, University of Hawaii Manoa, National Natural
Science Foundation of China (Project #71773027), the Center for Economics, Finance and
Management Studies (CEFMS) at Hunan University. Xiuping Hua gratefully acknowledges financial
support from Center for Inclusive Finance at the People’s Bank of China Ningbo Branch and Ningbo
Science and Technology Bureau for Innovation Team (No. 2011B81006).
Notes
1. According to the press release of the Society for Worldwide Interbank Financial
Telecommunication (SWIFT, October 6, 2015), Chinese Yuan has entered the top four of world
payment currencies by value in August 2015 along with USD, Euro, and British pound, overtak-
ing the Japanese yen and reaching a record high share of 2.79% in global payments. In the last 3
years, the RMB has overtaken seven currencies rising from position number 12 with a share of
0.84% in August 2012.
2. Besides China, other major economies, such as Bahrain, Cuba, Djibouti, Eritrea, Hong Kong,
Jordan, Lebanon, Oman, Panama, Qatar, Saudi Arabia, United Arab Emirates, and Venezuela,
have pegged their currencies to the USD (see www.investopedia.com/).
3. No prior study has examined the pricing of exchange rate risk under a pegged regime in an emer-
ging market environment. Bartov, Bodnar, and Kaul (1996) studied the regime change in the
United States. They studied the relation between exchange rate variability and stock return vola-
tility for U.S. multinational firms over two 5-year periods around the 1973 switch from fixed to
floating exchange rates and find a significant increase in the volatility of monthly stock returns.
4. The official rate for the USD price was adjusted from 2.80 RMB in 1985 to 3.72 in 1986 and to
5.20 in 1990. After fluctuation between 1992 and 1994, it depreciated to 5.80. Correspondingly,
the market exchange rate changed from 3.08 Yuan per dollar to 4.20 in 1985, 5.70 in 1988, and
8.70 in January 1994.
5. For example, Henan Rebecca Hair Products (stock code: 600439.SH) is one of the listed compa-
nies with high ratios of exports to sales. By June 2005, most of its products were sold to overseas
markets, with 69.30% of total sales in the North American market, 10.45% in the European
market, and 18.20% in the African market. With the RMB currency appreciation trend since July
2005, the company has faced higher external uncertainties and has worked very hard to increase
domestic sales and distribution channels. Despite its effort, most of the company’s sales are still
in the overseas market. In mid-2009, 59.65% of its total sales were in the North American
market, while 6.01% were in the European market and 29.55% in the African market. Our model
estimation at firm level confirms that international trade was a driving force of their stock returns
during 2005 and 2015. The company is highly sensitive to international trade. The company,
Hua et al. 73
along with other companies with high export revenues, such as Qingdao Haier Co. (600690.SH)
and Ningbo Electric Appliance (600724.SH), are among the group with high sensitivity to Trade
in Table 10 during some sample years.
6. For example, one type of industry prone to the hot money effect is mining companies, such as
Yunnan Copper (code: 000878.SZ). The company has principally engaged in the production and
sale of copper concentrates, precious metals, and related products, as well as chemical products.
The company, which operates its businesses mainly in the domestic markets, is popularly per-
ceived in the market and has attracted a large amount of short-term speculative funds, especially
during exchange rate reform periods. The other sector that may have attracted a lot of hot money
is real estate. However, the government had adopted strict regulation measures to prevent the
influx of hot money into this sector. For instance, foreign institutions with a physical presence in
China are only allowed to buy commercial property for their own use and it must be in the city
in which they are registered. Thus, not all real estate companies are affected by hot money. The
stock mentioned above, Yunnan Copper, along with other companies such as Tibet Mining
(000762.SZ), China Merchants Property Development Co. (000024.SZ), and Overseas Chinese
Town (000069.SZ), are part of ‘‘Hot Money (HM)’’ sensitive group in Table 10 during some
sample years.
7. China has recently surpassed Japan to become the world’s largest reserve holder (Ouyang, Rajan,
& Willett, 2010).
8. China Vanke Co. (000002.SZ), the country’s largest real estate developer by sales, provides an
example. The company has focused primarily on residential property development and its home
sales have always been affected by the government’s credit expansion or tightening measures.
Other similar stocks, including China Guanyu Development Co. (000537.SZ), Zhongjiang
Property Co. (600053.SH), and Huayuan Property Co. (600743.SH) are part of M2 high-
sensitivity group in Table 10 during some sample years.
9. According to classification by China’s Securities Regulatory Commission in 2012, there are total
19 industries. We omit Civil service, maintenance, and other services industry in our sample due
to lack of data.
10. We obtained the export revenue information from companies’ financial reports and the data are
only available from 2001.
11. All variables are expressed in percentage changes.
12. M2 broad money supply includes savings deposits, money market mutual funds, and other time
deposits. They are less liquid but can be quickly converted into cash or checking deposits.
13. As aforementioned, on July 21, 2005, the People’s Bank of China (PBOC), China’s central bank,
made the announcement to switch Yuan from strictly pegged to the USD to a regime in which
the RMB was pegged to a basket of foreign currencies, including some Asian currencies. The
RMB appreciated by 2.1% immediately and a cumulative 21% against the USD by July 2008.
The appreciation process took a toll on Chinese exports, especially during the global financial
crisis. To help the country’s exporters during the financial crisis, China had effectively pegged
the Yuan at about 6.83 per dollar until June 2010. However, during this period, speculation
about an imminent move in the value of the RMB intensified amid rising trade tensions between
the United States and China. On June 19, 2010, the PBOC announced a new reform to improve
the flexibility in exchange rate determination by setting a midpoint and daily variation for the
value of USD. By June 2015, the RMB rose to a new record of 6.1161 Yuan per USD, marking
a 10.42% appreciation since the resumption of exchange rate reform in June 2010. Although the
RMB exchange rate regime is still heavily managed, it tends to be more volatile.
14. We obtained sales distribution data from companies’ financial statements. Export companies are
those with export revenues in their sales.
15. We thank the referee for suggesting this test.
16. For example, we estimate Equation 5 for each firm using data from July 1995 to June 1997 to
obtain firm-specific values of an5 coefficient for 1997. We repeat this procedure until 2015.
74 Journal of Accounting, Auditing & Finance
17. We also tried ranking the firms into 10 portfolios based on the value of an5; the test results are
similar to using the 25 grouping method.
18. As in Kolari, Moorman, and Sorescu (2008), because the emphasis is on the change in pricing
error, the risk factor can therefore be constructed based on the absolute value of foreign-
exchange sensitivity.
19. We use this setting because the exchange rate regime switching occurred in July 2005 and we
require 2 years of data to calculate coefficients’ loadings on the pricing factors.
20. We thank the referee for suggesting this analysis.
21. The random effects are also estimated and the results are very similar.
References
Adler, M., & Dumas, B. (1983). International portfolio choice and corporation finance: A synthesis.
The Journal of Finance, 38, 925-984.
Agosin, M. R., & Huaita, F. (2012). Overreaction in capital flows to emerging markets: Booms and
sudden stops. Journal of International Money and Finance, 31, 1140-1155.
Bartov, E., Bodnar, G. M., & Kaul, A. (1996). Exchange rate variability and the riskiness of U.S.
multinational firms: Evidence from the breakdown of the Bretton Woods system. Journal of
Financial Economics, 42, 105-132.
Bodnar, G. M., & Gentry, W. M. (1993). Exchange rate exposure and industry characteristics: Evidence
from Canada, Japan, and the USA. Journal of International Money and Finance, 12, 29-45.
Broner, F. A., & Rigobón, R. (2006). Why are capital flows so much more volatile in emerging than
in developed countries? In R. J. Caballero, C. Calderón, & L. F. Céspedes (Eds.), External vulner-
ability and preventive policies (pp. 15-40). Santiago: Central Bank of Chile.
Brown, S. J., & Warner, J. B. (1985). Using daily stock returns: The case of event studies. Journal of
Financial Economics, 14, 3-31.
Carhart, M. M. (1997). On persistence in mutual fund performance. The Journal of Finance, 52, 57-82.
Carrieri, F., & Majerbi, B. (2006). The pricing of exchange risk in emerging stock markets. Journal
of International Business Studies, 37, 372-391.
Choi, J. J., & Rajan, M. (1997). A joint test of market segmentation and exchange risk factor in inter-
national capital markets. Journal of International Business Studies, 28, 29-49.
De Santis, G., & Gérard, B. (1998). How big is the premium for currency risk? Journal of Financial
Economics, 49, 375-412.
Dornbusch, R., & Fischer, S. (1980). Exchange rates and the current account. The American
Economic Review, 70, 960-971.
Duarte, M., & Stockman, A. C. (2005). Rational speculation and exchange rates. Journal of Monetary
Economics, 52, 3-29.
Dumas, B., & Solnik, B. (1995). The world price of foreign exchange risk. The Journal of Finance,
50, 445-479.
Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds.
Journal of Financial Economics, 33, 3-56.
Gibbons, M. R., Ross, S. A., & Shanken, J. (1989). A test of the efficiency of a given portfolio.
Econometrica, 57, 1121-1152.
Jorion, P. (1990). The exchange-rate exposure of U.S. multinationals. The Journal of Business, 63,
331-345.
Jorion, P. (1991). The pricing of exchange rate risk in the stock market. The Journal of Financial and
Quantitative Analysis, 26, 363-376.
Kolari, J. W., Moorman, T. C., & Sorescu, S. M. (2008). Foreign exchange risk and the cross-section
of stock returns. Journal of International Money and Finance, 27, 1074-1097.
Lewellen, J., Nagel, S., & Shanken, J. (2010). A skeptical appraisal of asset pricing tests. Journal of
Financial Economics, 96, 175-194.
Hua et al. 75
Magud, N. E., Reinhart, C. M., & Vesperoni, E. R. (2014). Capital inflows, exchange rate flexibility
and credit booms. Review of Development Economics, 18, 415-430.
Mazouz, K., Joseph, N. L., & Joulmer, J. (2009). Stock price reaction following large one-day price
changes: UK evidence. Journal of Banking and Finance, 33, 1481-1493.
Muller, A., & Verschoor, W. (2006). Asymmetric foreign exchange risk exposure: Evidence from
U.S. multinational firms. Journal of Empirical Finance, 13, 495-518.
Ouyang, A. Y., Rajan, R. S., & Willett, T. D. (2010). China as a reserve sink: The evidence from
offset and sterilization coefficients. Journal of International Money and Finance, 29, 951-972.
Parsley, D. C., & Popper, H. A. (2006). Exchange rate pegs and foreign exchange exposure in East
and South East Asia. Journal of International Money and Finance, 25, 992-1009.
Pavlova, A., & Rigobon, R. (2007). Asset prices and exchange rates. The Review of Financial
Studies, 20, 1139-1181.
Priestley, R., & Ødegaard, B. A. (2007). Linear and nonlinear exchange rate exposure. Journal of
International Money and Finance, 26, 1016-1037.
Reinhart, C., & Montiel, P. (2001). The dynamics of capital movements to emerging economies
during the 1990s. In S. Griffith-Jones, M. Montes, & A. Nasution (Eds.), Short-term capital flows
and economic crises (pp. 3-28). Oxford, UK: Oxford University Press.
Solnik, B. H. (1974). An equilibrium model of the international capital market. Journal of Economic
Theory, 8, 500-524.
Solnik, B. H. (1997). The world price of foreign exchange risk: Some synthetic comments. European
Financial Management, 3, 9-22.
Spiegel, M. M. (2005, September 9). A look at China’s new exchange rate regime. FRBSF Economic
Letter. Retrieved from http://www.frbsf.org/economic-research/files/el2005-23.pdf
Stulz, R. M. (1981a). A model of international asset pricing. Journal of Financial Economics, 9, 383-
406.
Stulz, R. M. (1981b). On the effects of barriers to international investment. The Journal of Finance,
36, 923-934
76 Journal of Accounting, Auditing & Finance
Copyright of Journal of Accounting, Auditing & Finance is the property of Sage Publications Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
