Exchange—rate exposure and abnormal stock performance： Evidence from China

Xiaoyuan+Liu+HanXu+Li

摘 要：The article using quarterly data presents the empirical results of the relation between abnormal stock performance and contemporaneous and lagged changes in the value of RMB based on 25 listed trading firms with exchange rate exposure in China from 2002Q2 to 2013Q4. Regression results show us that both the contemporaneous and lagged exchange rate changes are negatively associated with the abnormal stock performance， and that the contemporaneous exchange rate changes explain more about the mispricing.

Introduction

The purpose of this paper is to examine the relationship between changes of exchange rate and abnormal stock performanc. Also we take the China exchange rate reform into consideration to see the structural change in the relation.

In order to see clear in this problem， we need to choose firms which have similar exchange exposures. In the previous studies， there are two major problems exists. First is the sample selection， this paper argues that if the sample contains firms with different relationship between exchange rate and stock performance， that will make the empirical result ambiguous. Also， for some big trading firms， because they have big exchange rate exposure， so they always take some strategy to hedge the risk. They think that investors do not freely use all the available information is another reason to the not significant evidence in prior studies. We explore the relation between the exchange rate changes with abnormal stock performance in Chinese market.

Model and Variable Definitions

To set up our regression model， we should first get the abnormal stock returns （ASP） of firms. The ASPs are derived from a series of abnormal return and we use the similar method to ELI BARTOV and GORDON M.BODNAR （1994） to measure the abnormal return.

We use CAPM model r-rf=β（rm-rf）+α to estimated daily beta and alpha of each stock. Data used to calculate day ts beta and alpha is from 200 days prior of day t to 100 days prior day t. rf is calculated from 3-month time deposit and rm is SSE Composite Index.

Compute daily abnormal returns （ARi，t） of each firm using the market model： ARi，t= ri，t - αi - βi*rm，t，where ri，t is the real return for security i in day t， rm，t is the market return in day t

According to the formula， we use the first 60 trading days AR to capture the ASP. If there is any quarter doesnt have 60 trading days data， we use all the days we have got in that quarter. The first ASP is from 2002Q2， we skip the first quarter of the new fiscal year to allow the abnormal return period to be a period after the investors have received information about the past performance of the firm.

Our key regression：

Where ASPi，t is the abnormal stock performance for security i in period t； ？CURi，t-j is the change of trade-weighted exchange rate for the period t to j； Rt is dummy variable represents China exchange rate reform， a variable that is 1 if the 60-trading-day window for calculating ASPi，t starts on July 1， 2005， or later， and zero otherwise； d_j id the coefficient that measures the difference between the coefficient of the jth lag （j = 0，1） of the change in the dollar for 2005Q3 to 2013Q4 as compared to 2002Q2 to 2005Q2；

Data Description

We obtain the daily data of 25 listed trading firms in Chinese stock market from 2001Q2 to 2013Q4 from Wind. We also use quarter trade-weighted exchange rate index （REER index） from 2002Q2 to 2013Q4 to capture the changes of exchange rate， the increase of real effective exchange rate means the appreciation of RMB. The data of exchange rate are obtained from DataStream.

Tests and Results

Contemporaneous and lagged relation

We begin our testing by exploring whether the contemporaneous relation between exchange rate changes in RMB and abnormal stock returns is evident for our sample firms. We choose the firms with negative exposure to exchange rate changes， so we expected there would exist negative relationship between the exchange rate changes and the abnormal stock returns.

In regression （1）， we mainly test the contemporaneous relation and the lagged relation. We expected both the coefficients of ΔCUR are negative， which means that the appreciation of RMB will decrease the values of firms with foreign sales or operations. Also， we expected that the lagged exchange rate changes would have significant influence on the current abnormal stock returns because the delayed information is disseminated for investors.

Table 1. Results of regression of abnormal stock returns on contemporaneous and lagged changes in the RMB

*， **， and *** denote significance at 10， 5， and 1 percent levels， respectively.

The numbers in the parenthesis are standard errors of estimators.

We run the regression and Table 1 displays the results of contemporaneous and lagged relation between exchange rates changes and the firms abnormal stock performance. These results suggest that when the RMB appreciate， the stocks of those firms with foreign sales tend to have negative abnormal performance. Though the freely available information concerning the past changes can help explain the abnormal future stock performance， the contemporaneous changes in RMB explain more in the mispricing. Whats more， because less information is disclosed in the current period， the scale of pricing errors tends to be higher when the exchange rate changes.

Structural break in the relation

In the second part， we will test for the stability of the relation between changes in firm values and movement of the RMB. Chinese exchange rate reform is the event that we think can change the stability of the relation， so we use 2005Q3 as the break point to divide our subsamples.

Table 11. Tests for a structural break in the relation between abnormal stock returns and RMB changes

*， **， and *** denote significance at 10， 5， and 1 percent levels， respectively.

The numbers in the parenthesis are standard errors of estimators.

Table 11 displays the results of regression （2）. From the results， we can see that in the period before the exchange rate reform， the lagged changes in RMB explain more in the mispricing， while the contemporaneous changes are not as significant as the lagged changes. The estimate of the coefficient for the structural break in the second period d_1is positive and significant， suggesting the decrease in the size of the negative relation between lagged changes in RMB and abnormal stock performance. Although the lagged RMB change is significantly negatively related to abnormal stock performance in the second period， the degree of the pricing errors is lower in the second period. That results inspire us that the Chinese exchange rate reform help decrease the degree of pricing errors when the exchange rates fluctuate.

Conclusion

The article using quarterly data presents the empirical results of the relation between abnormal stock performance and contemporaneous and lagged changes in the value of RMB based on 25 listed trading firms with exchange rate exposure in China from 2002Q2 to 2013Q4. We choose those listed firms with foreign sales and the exchange-rate exposure. Regression results show us that both the contemporaneous and lagged exchange rate changes are negatively associated with the abnormal stock performance， with is consistent with the hypothesis. Compared to lagged exchange rate changes， the contemporaneous exchange rate changes explain more about the mispricing. Because less information is disclosed in the current period， the scale of pricing errors tends to be higher when the exchange rate changes.

Taking the consideration of Chinese exchange rate reform， we divide the whole sample into two subsamples. The regression results show us that the lagged changes in RMB explain more in the mispricing before the exchange rate reform. In the period after the reform， we find the significant decrease in the size of the relation between lagged changes in RMB and abnormal stock performance. Although the lagged RMB change is significantly negatively related to abnormal stock performance in the second period， the degree of the pricing errors is lower in the second period.

Reference

Eli Bartov， Gordon M. Bodnar， （1994）， Firm Valuation， Earnings Expectations， and the Exchange-Rate Exposure Effect， Journal of Finance 5， 1755-1785.

Abuaf， N.， & Jorion， P. （1990）. Purchasing power parity in the long run. Journal of Finance， 45（1）， 157-174.

Amihud， Y. （1993）. Evidence on exchange rates and valuation of equity shares.