SHORT-TERM STOCK PRICE REACTION TO SHOCKS: EVIDENCE FROM AMMAN STOCK EXCHANGE

Transcription

1 SHORT-TERM STOCK PRICE REACTION TO SHOCKS: EVIDENCE FROM AMMAN STOCK EXCHANGE Dima Walee Hanna Alrabai Assistant Professor, Finance an Banking Sciences Department, Faculty of Economics an Business Aministration, Yarmouk University, Irbi, Joran. (P.O Box 566, ZIP Coe 21163). Abstract This stuy examine the short term price reaction to shocks for all the stocks trae in Amman Stock Exchange (ASE) using aily ata over the perio The abnormal return is calculate as the ifference between the aily return an an average return over a winow of 50 ays ening 10 ays prior to the ay of the shock. ASE applies a aily price limit of 5% up an own. Therefore, the stuy uses a shock of 5% trigger value. The finings provie evience in favor of the unerreaction hypothesis. Highly significant return continuation is ocumente up to 10 ays following both positive an negative price shocks. The evience is robust to size an liquiity effects. Moreover, the results are not time varying an consistent over the years of the stuy. Abnormal returns exist even after ajusting for market risk, however, in lower values. Keywors: Short term stock reaction, Efficient market hypothesis, Overreaction, Unerreaction, Amman Stock Exchange, price reversal, return continuation, price shocks. 1. Introuction Accoring to the efficient market hypothesis, stock prices shoul incorporate new information quickly an accurately. Thus, no investor can preict prices an achieve abnormal returns. Financial markets are foun to be inefficient. Price reversals an return continuation following large one-ay price changes are well ocumente anomalies against the efficient market hypothesis. Reversal is first reporte by DeBont an Thaler (1985) who argue that investors overweigh new information, resulting in a price reversal when the market corrects itself (what is also calle overreaction phenomenon). On the contrary, Jegaeesh an Titman (1993, 2001) fin that investors unerweight the new information signals, causing significant return continuation (what is calle unerreaction phenomenon). We examine the short term price reaction to shocks for all the stocks trae in Amman Stock Exchange (ASE) using aily ata over the perio In particular, we consier the cumulative abnormal returns up to 10 ays following a price shock. The shock is efine as an increase (ecrease) of 5% or more (less). The abnormal return is calculate as the ifference between the aily return an an average return over a winow of 50 ays ening 10 ays prior to the ay of the shock. ASE applies a aily price limit of 5% up an own. Therefore, we use a shock of 5% only. Our results provie evience in favor of the unerreaction hypothesis. We ocument highly significant return continuation up to 10 ays following both positive an negative price shocks. The evience is robust to size an liquiity effects. Five portfolios sorte accoring to firm size show consistent evience of momentum. Moreover, analogous portfolios sorte accoring to the number of traes show consistent results. Our results are not time varying, they are consistent over the years of the stuy. Abnormal returns exist even after ajusting for market risk, however, in lower values. The remainer of the stuy is organize as follows. Section 2 reviews the relate literature. Section 3 escribes ata an research methoology. Section 4 reports the empirical results an Section 5 conclues on the main finings. 2. Literature Review Fama (1970) introuces the efficient market hypothesis. It states that stock prices reflect all past, publically available, an private information. Thus, no investor can use any piece of information to achieve an abnormal return. Researchers fin many anomalies challenging the efficient market hypothesis. Price reversals an return continuation are well ocumente phenomena which enable investors to formulate investment strategies earning abnormal returns. Price reversal was originally explore by DeBont an Thaler (1985); they fin that stocks which COPY RIGHT 2012 Institute of Interisciplinary Business Research 770

2 ha goo/ba performance over the last one to three years (winners/losers) reverse it over the coming analogous perios. Howe (1986) an Lehmann (1990) report consistent results over the short run. On the other han, Jegaeesh an Titman (1993) were the first to fin a return continuation over the short run (3 months to one year) for winners an losers portfolios. Schnusenberg an Maura (2001) ocument similar results over a sixty-ay interval. Researchers who examine the short run price reaction to shocks foun inconsistent results. Bremer an Sweeney (1991) examine the stock price reaction to a shock of -10% for all the stocks that are liste in Fortune 500 over the perio 1962 to They efine the abnormal return as the stock's aily return in excess of its average return over the sample perio. Their results show large price reversals after negative price shocks. Particularly, Bremer an Sweeney (1991) ocument that a significant positive abnormal return of 1.773% can be realise in the ay after a -10% price ecline an it accumulates to 2.215% in the secon ay after the price shock. They argue that the ocumente price reversal is istinct from any other anomaly. Thus, it is not a January effect, a Monay effect, a weeken effect, or a turn of the year effect. It is robust to ifferent trigger values, sub-perios, an subsamples. Lasfer et al. (2003) examine the short-run price reaction for the market inexes of 39 stock exchanges aroun the worl through 1989 to They efine a positive (negative) price shock as the one where the return on a particular ay is above (below) two stanar eviations the average market aily return calculate over [60 to 11] ays prior to the ay of the price shock. The abnormal return is compute as the ifference between aily return an the average return over a winow of 50 ays ening 10 ays prior to the ay of the shock. Lasfer et al. (2003) fin that for both evelope an emerging markets, investors show unerreaction following positive an negative price shocks. That is, abnormal returns can be achieve up to 10 ays following the shock. However, they argue that some of these abnormal returns have ecrease in the late 1990s. Furthermore, they report that the quantity of the momentum profits is larger for emerging an less liqui markets. Aitionally, the abnormal returns are foun to be robust to seasonal month/year effects. Spyrou et al. (2007) stuy the same issue for four FTSE inexes in Lonon Stock Exchange over the perio They argue that each inex can be consiere as a value weighte portfolio of stocks which represent a certain size segment of the market. Their finings show that large capitalization stock portfolios react efficiently to extreme shocks. However, small an meium capitalization stock portfolios show significant unerreaction to both positive an negative price shocks. Thus, abnormal returns can be achieve while investing in these stocks for many ays after the shock. Moreover, Spyrou et al. (2007) fin that the abnormal returns exist even after ajusting for Fama an French (1993) factors an consiering bi-ask biases an global financial crises. Along the lines with the previous stuies, Mazouz et al. (2009) investigate the short run stock price reaction to a large one-ay price shock of 424 stocks of Lonon Stock Exchange over the perio of January 1992 to December This stuy focuses on the iniviual stock behavior rather than the market inex behavior. Mazouz et al. (2009) argue that the momentum or contrarian abnormal returns may exit ue to the limite ability of the CAPM to explain the cross section of stock returns. They efine the abnormal returns as the resiuals from three competing moels CAPM, the Fama an French (1993) three-factor moel, an the Carhart (1997) four-factor moel. They consier a price change of 5%, 10%, 20% (or more) as a shock. They fin significant abnormal returns following positive price shocks at all trigger values an negative price shocks of a 5% or more. These abnormal profits o not significantly iffer across the three examine moels. Moreover, a significant one-ay positive abnormal profit of 1% can be achieve after a positive shock of a 5% or more uner all the three moels. Similarly, a negative abnormal return of -.43% uner the CAPM an -.34% uner the other two moels can be earne after a negative shock of -5% or less. Thus, these results support the unerreaction hypothesis an are robust to ifferent subperios. Mazouz et al. (2012) examine the price reaction to one-ay price shocks of FTSEALL shares inex stocks over the perio Their results support the return continuation hypothesis. However, they show that stocks with low systematic liquiity risk react efficiently to both positive an negative price shocks, whereas stocks with high systematic liquiity risk unerreact to both positive an negative shocks. Their results are vali irrespective of various robustness tests such as size of the shock, size of the firm, month-of-the-year an ay-of-the-week effects. Researchers suggest many factors to explain the observe price reversals an return continuations. Zarwin (1990) argue that overreaction is a emonstration of the size anomaly 1. Chan (1988), Ball an Kothari (1989), Berk et al. (1999), Choria an Shivakumar (2002), Wu (2002), Wang (2003) an Li et al. (2008) argue that markets are efficient an the observe abnormal returns are ue to the failure to ajust for time-varying risk. A number of stuies fin that the abnormal returns are not large enough to cover the transaction cost of traing (see, for example, Atkins an Dyi, 1990; Cox an Peterson, 1994; Park, 1995; Korajczyk an Saka, 2004; Lesmon et al. 2004). COPY RIGHT 2012 Institute of Interisciplinary Business Research 771

3 Moskowitz an Grinblatt (1999) emonstrate that inustry effects 2 explain the iniviual stock momentum (return continuation). Saka (2006) suggests liquiity risk as a plausible explanation for momentum profits. Lo an Mackinlay (1990) ocument that the overtime positive cross-autocorrelation between stocks an its lea-lag structure rather than the negative autocorrelation in iniviual stock returns that explain the price reversals. Daniel et al. (1998), Barberis et al. (1998), Hong an Stein (1999) an Hong et al. (2000) employ psychological concepts such as overconfience, biase self-attribution, conservatism an representativeness to valiate the investor behavior in response to new information signals. This stuy is the first to examine the behavior of stock prices following large one ay shocks in ASE. We inspect that behavior an its plausible explanations. 3. Data an methoology: Our ata set consists of the aily traing ata for all the stocks trae in ASE over the perio Each stock shoul be trae at least once each 10 traing ays in orer to be inclue in the analysis. This filtering process which was carrie out on a yearly basis yiels a number of stocks range from 64 in 2002 to 171 in The abnormal return for stock i on ay t is calculate as follows: AR i, t Ri, t AVG( R) i, t (1) Where AR i, t is the abnormal return, R i, t is the return of stock i on ay t, AVG ( R) i, t is the average return over a winow of 50 ays ening 10 ays prior to the ay of the shock. We calculate the cumulative abnormal returns starting one ay after a shock as: CAR i AR CAR i,, i, t (2) t 1 for stock i over a winow of ays The average cumulative abnormal return associate with N stocks over a winow of ays following a shock is calculate as: CAAR n CAR i, i 1 n (3) We etermine the significance of CAAR using the t-test ajuste to serial correlation an heteroskeasticity by Newey-West HAC Stanar Errors & Covariance. Moreover, a non-parametric sign test is use to juge the significance of the average cumulative abnormal returns assuming non-normality. We check if results are le by size or liquiity. Particularly, we sort stocks accoring to the firm size an assign them to five portfolios. Next, we use equations (1) an (2) to calculate the abnormal returns AR i, t an the cumulative abnormal returns CAR i, of each stock in each portfolio. In this case, we calculate the average cumulative abnormal return for each portfolio separately. Thus: CAAR n CAR i, i 1 n where n is the number of stocks in a certain portfolio an is the number of ays following a shock. We repeat the same analysis for portfolios sorte accoring to the number of traes. Moreover, as a robustness check we repeat the analysis on a year by year basis in orer to know whether the results are time varying or consistent over time. Finally, the methoology use might not fully account for the risk element. We use a ifferent methoology to calculate CAARs. In particular, we efine the abnormal returns as the resiuals from the market moel. (4) COPY RIGHT 2012 Institute of Interisciplinary Business Research 772

4 4. Empirical Results 4.1. Results of short term price reaction to shocks of all sample stocks Table 1 reports the CAARs of all sample stocks following positive an negative price shocks over the perio The negative price shocks are larger in number an value than positive shocks. The number of positive shocks is 2387 while the number of negative shocks is Table 1 illustrates evience in favor of the unerreaction hypothesis. Stocks show significant return continuation that extens for at least 10 ays following both positive an negative price shocks. Joranian investors seem to unerreact to both goo an ba news causing return momentum. The CAAR reaches its maximum of 7% on the thir ay following positive shocks. On the other han, a CAAR of approximately -8% exists along ten ays following negative shocks. Our results are consistent with the finings of Lasfer et al. (2003), Spyrou et al. (2007) an Mazouz et al. (2009, 2012). Lasfer et al. (2003) fin that investors in both evelope an emerging markets realize significant positive CAARs up to 10 ays following both positive an negative price shocks. Spyrou et al. (2007) fin evience supporting the unerreaction hypothesis for the small an meium capitalisation stocks of FTSE250 an FTSE SmallCap, respectively. Mazouz et al. (2009, 2012) argue that UK stocks shows significant momentum following positive shocks of all magnitues (5%, 10%, an 20%) an negative shocks of -5% or less. Figure 1 shows the CAARs of all sample stocks following positive an negative price shocks over the perio Are the results size-epenent? Table 2 shows the CAARs of stocks in portfolios sorte accoring to firm size following both positive an negative price shocks. Each year, we assign the sample stocks to 5 portfolios accoring to company size. Port 1 inclues the smallest stocks while port 5 inclues the largest stocks. Size is measure by the firm's market capitalization. The number of stocks in each portfolio ranges from 64 in 2002 to 171 in Next, we calculate the CAARs of the stocks of each portfolio following both positive an negative shocks. The results show that the evience of return continuation is robust to firm size. Table 2 shows that the stocks of both Port 1 an Port 5 show significant return continuation extens for at least 10 ays following both positive an negative price shocks. However, Port 1 shows higher number of positive an negative shocks than Port 5. The values of positive CAARs are higher for Port1 while the values of negative CAARs are higher for Port Are the results liquiity epenent? Table 3 ocuments the CAARs of stocks in portfolios sorte accoring to stock liquiity following both positive an negative price shocks. Each year, we assign the sample stocks to 5 portfolios accoring to stock liquiity. Port 1 inclues the least liqui stocks while port 5 inclues the most liqui stocks. Number of traes is use as a proxy for stock liquiity. The number of stocks in each portfolio ranges from 64 in 2002 to 171 in Next, we calculate the CAARs of the stocks of each portfolio following both positive an negative price shocks. The results emonstrate that the evience of return continuation is also robust to stock liquiity. Table 3 shows that the stocks of both Port 1 an Port 5 show significant return continuation extens for at least 10 ays following both positive an negative shocks. However, Port 1 shows higher number of positive an negative shocks than Port 5. The values of positive CAARs are higher for Port1 while the values of negative CAARs are higher for Port Are the results time-varying? Table 4 reports the CAARs of all sample stocks on a yearly basis. Panel A reports the yearly CAARs of all stocks following positive price shocks. The results are consistent over years. Stocks show significant return continuation for 10 ays following a positive shock. Panel B reports the yearly CAARs of all stocks following negative price shocks. The results are also consistent over ifferent years. Negative shocks are followe by significant negative CAARs for at least 10 ays following a shock Are the results robust to risk ajustment? Table 5 shows the CAARs after ajusting them to risk. Stocks show significant return continuation following both positive an negative price shocks even after ajusting their returns to risk. However, the CAARs compute accoring to the market moel are less in number an amount. They range between 4-5% following both positive an negative shocks. 5. Conclusions Literature has ocumente many anomalies challenging the efficient market hypothesis of Fama (1970). One important anomaly is the abnormal returns following large one-ay price shocks. Some stuies fin significant return continuation following positive an negative price shocks. This phenomenon is calle the unerreaction COPY RIGHT 2012 Institute of Interisciplinary Business Research 773

5 hypothesis which assumes that investors unerreact to new information causing a return momentum over several ays subsequent to a price shock. On the other han some researchers provie evience in favor of what is calle the overreaction hypothesis. This hypothesis is originally propose by DeBont an Thaler (1985) an states that investors overreact to new information causing a price reversal when the market corrects itself. We examine the stock price reaction to large one-ay shocks of all the stocks liste in ASE over the perio Shocks are efine as price increase (ecrease) of 5% or more (less). Abnormal return are compute as the ifference between the aily return an an average return over a winow of 50 ays ening 10 ays prior to the ay of the shock. We fin evience in favor of the unerreaction hypothesis in ASE. Significant return continuation is foun for at least 10 ays following both positive an negative price shocks. These results are consistent with Lasfer et al. (2003) who reach to the same conclusion for the inexes of 39 evelope an emerging markets. Moreover, our results are analougos to those of Spyrou et al. (2007) an Mazouz et al. (2009, 2012) for Lonon Stock Exchange. Abnormal returns accumulate to a maximum of 7% on the thir ay following positive price shocks in ASE. On the other han, a CAAR of approximately -8% exists along ten ays following negative shocks. Furthermore, we show that our results are not relate to size or liquiity effects an are consistent over the ifferent years of the sample perio. In particular, both small an large stocks an least an most liqui stocks show significant return momentum following shocks of 5% (-5%) or more (less). Abnormal returns exist even after ajusting to market risk. Our finings are very important to both acaemicians an practitioners in ASE. For acaemicians, we provie new evience against the efficient market hypothesis. Practitioners mainly iniviual an institutional investors can formulate their investment strategies taking into consieration the stock that experience a large one-ay change of 5%(-5%) or more (less) will continue to achieve abnormal returns (losses) of aroun 7-8% up to 10 ays following that change. Future research may consier other methoologies to efine abnormal returns an may suggest other explanations to the behavior of stock prices following one-ay shocks in ASE. COPY RIGHT 2012 Institute of Interisciplinary Business Research 774

6 References 1. Atkins, A. & Dyl, E. (1990). Price reversals, bi-ask spreas an market efficiency. Journal of Financial an Quantitative Analysis, 25, Ball, R. & Kothari, S. P. (1989). Nonstationary expecte returns: implications for tests of market efficiency an serial correlations in returns. Journal of Financial Economics, 25, Banz, R. W. (1981). The relationship between return an market value of common stocks. Journal of financial Economics, 9, Barberis, N., Shleifer, A. & Vishny, R. (1998). A moel of investor sentiment. Journal of Financial Economics, 49, Berk, J., Green, R. & Naik, V. (1999). Optimal investment, growth options an security returns. Journal of Finance, 54, Bremer, M. & Sweeney, R.J. (1991). The reversal of large stock-price ecreases. Journal of Finance, 46, Carhart, M. (1997). On persistence in mutual fun performance. Journal of Finance, 52(1), Chan, K. C. (1988). On the contrarian investment strategy. Journal of Business, 61, Choria, T. & Shivakumar, L. (2002). Momentum, business cycle an time-varying expecte returns. Journal of Finance, 57, Conra, J. & Kaul, G. (1998). An anatomy of traing strategies. Review of Financial Stuies, 11, Cox, D. R. & Peterson, D.R. (1994). Stock returns following large one-ay eclines: evience on shortterm reversals an longer-term performance. Journal of Finance, 49, Daniel, K., Hirshleifer, D. & Subrahmanyam A. (1998). A theory of overconfience, self-attribution an security market uner- an over-reactions. Journal of Finance, 53(6), DeBont, W. & Thaler, R. (1985). Does the stock market overreact?. Journal of Finance, 40, DeBont, W. & Thaler, R. (1987). Further evience on investor overreaction an stock market seasonality. Journal of Finance, 42, Fama, E. & French, K. (1993). Common risk factors in the returns on stocks an bons. Journal of Financial Economics, 33, Fama, E. F. (1970). Efficient capital markets: A review of theory an empirical work. Journal of Finance, 25(2), Hong, H. & Stein, J. (1999). A unifie theory of unerreaction, momentum traing an overreaction in asset markets. The Journal of Finance, 54(6), Hong, H., Lim, T. & Stein, J. C. (2000). Ba news travels slowly: size, analyst coverage an the profitability of momentum strategies. Journal of Finance, 55, Howe, J. (1986). Evience on stock market overreaction. Financial Analyst Journal, 42(4), Jegaeesh, N. & Titman, S. (1993). Returns to buying winners an selling losers: implications for stock market efficiency. Journal of Finance, 48, Jegaeesh, N. & Titman, S. (2001). Profitability of momentum strategies: an evaluation of alternative explanations. Journal of Finance, 56, Korajczyk, R. A. & R. Saka. (2004). Are momentum profits robust to traing costs?. Journal of Finance, 59, Lasfer, M. A., Melnik, A. & Thomas, D. C. (2003). Short term reaction of stock markets in stressful circumstances. Journal of Banking an Finance, 27, Lehmann, B. N. (1990). Fas, martingales an market efficiency. Quarterly Journal of Economics, 105, COPY RIGHT 2012 Institute of Interisciplinary Business Research 775

7 25. Lesmon, D. A., Schill, M. J. & Zhou, C. (2004). The illusory nature of momentum profits. Journal of Financial Economics, 71, Li, X., Brooks, C. & Miffre, J. (2008). Low-cost momentum strategies. Unpublishe Working Paper. [Online] Available: SSRN: Lo, A. W. & MacKinlay, A.C. (1990). When are contrarian profits ue to market overreaction?. Review of Financial Stuies, 3, Mazouz, K., Joseph, N. L. & Joulmer, J. (2009). Stock price reaction following large one-ay price changes: U.K. evience. Journal of Banking an Finance, Mazouz, K., Alrabai, D., & Yin, S. (2012). Systematic liquiity risk an stock price reaction to shocks. Journal of Accounting an finance, 52(2), Moskowitz, T. J. & Grinblatt, M. (1999). Do inustries explain momentum?. Journal of Finance, 54, Park, J. (1995). A market microstructure explanation for preictable variations in stock returns following large price changes. The Journal of Financial an Quantitative Analysis, 30(2), Saka, R. (2006). Momentum an post-earnings-announcement rift anomalies: the role of liquiity risk. Journal of Financial Economics, 80, Schnusenberg, O. & Maura, J. (2001). Do U.S. stock market inexes over- or unerreact?. The Journal of Financial Research, 24, Spyrou, S., Kassimatis, K. & Galariotis, E. (2007). Short-term overreaction, unerreaction an efficient reaction: evience from the Lonon Stock Exchange. Journal of Applie Financial Economics, 17(3), Wang, K.Q. (2003). Asset pricing with conitional information: a new test. Journal of Finance, 58, Wu, X. (2002). A conitional multifactor moel of return momentum. Journal of Banking an Finance, 26, Zarowin, P. (1990). Size, seasonality an stock market overreaction. Journal of Financial an Quantitative Analysis, 25, Notes: Note1: Banz (1981) is the first to ientify the size anomaly. He fins that small firms earn higher risk- ajuste returns, on average, than large firms. Note 2: Moskowitz an Grinblatt (1999) clarify that investors may her towars (away from) hot (col) inustries, causing price pressure that coul create return continuation. COPY RIGHT 2012 Institute of Interisciplinary Business Research 776

8 Annexure Figure 1: CAARs following positive an negative shocks. COPY RIGHT 2012 Institute of Interisciplinary Business Research 777

9 Table 1: CAARs following shocks Positive Shocks Stats # of shocks CAAR 0 CAAR 1 CAAR 2 CAAR 3 CAAR 4 CAAR 5 CAAR 6 CAAR 7 CAAR 8 CAAR 9 CAAR 10 Mean t-test Wilcoxon Negative Shocks Stats # of shocks CAAR 0 CAAR 1 CAAR 2 CAAR 3 CAAR 4 CAAR 5 CAAR 6 CAAR 7 CAAR 8 CAAR 9 CAAR 10 Mean t-value Wilcoxon Table 2: CAARs of size-portfolios Size-ports # of shocks CAAR 0 CAAR 1 CAAR 2 CAAR 3 CAAR 4 CAAR 5 CAAR 6 CAAR 7 CAAR 8 CAAR 9 CAAR 10 PORT1-POS PORT5-POS PORT1-NEG PORT5-NEG COPY RIGHT 2012 Institute of Interisciplinary Business Research 778

10 Table 3: CAARs of size-portfolios Size-ports # of shocks CAAR 0 CAAR 1 CAAR 2 CAAR 3 CAAR 4 CAAR 5 CAAR 6 CAAR 7 CAAR 8 CAAR 9 CAAR 10 PORT1-POS PORT5-POS PORT1-NEG PORT5-NEG Table 4: CAARs sorte on a year by year basis. Panel A: CAARs following positive shocks Year Shocks CAAR0 CAAR1 CAAR2 CAAR3 CAAR4 CAAR5 CAAR6 CAAR7 CAAR8 CAAR9 CAAR COPY RIGHT 2012 Institute of Interisciplinary Business Research 779

11 Panel B: CAARs following negative shocks Year Shocks CAAR 0 CAAR 1 CAAR 2 CAAR 3 CAAR 4 CAAR 5 CAAR 6 CAAR 7 CAAR 8 CAAR 9 CAAR Table 5: Risk ajuste CAARs following price shocks Shocks CAAR0 CAAR1 CAAR2 CAAR3 CAAR4 CAAR5 CAAR6 CAAR7 CAAR8 CAAR9 CAAR10 Positive shocks Negative shocks COPY RIGHT 2012 Institute of Interisciplinary Business Research 780