دانلود رایگان مقاله موشکافی پازل نوسانات نامتقارن

Abstract

This paper discusses a novel explanation for asymmetric volatility based on the anchoring behavioral pattern. Anchoring as a heuristic bias causes investors to focus on recent price changes and price levels, which leads to a belief in continuing trend and mean-reversion, respectively. The empirical results support our theoretical explanation through an analysis of large price fluctuations in the S&P 500 and the resulting effects on implied and realized volatility. These results indicate that asymmetry (a negative relationship) between shocks and volatility in the subsequent period indeed exist. Moreover, contrary to previous research, our empirical tests also suggest that implied volatility is not simply an upward biased predictor of future deviation compensating for the variance of the volatility but rather, due to investors’ systematic anchoring to losses and gains in their volatility forecasts, a co-integrated yet asymmetric over-/under-estimated financial instrument. We also provide results indicating that the medium-term implied volatility (measured by the VIX Index) is an unbiased though inefficient estimation of realized volatility, while in contrast short-term volatility (measured by the recently introduced VXST Index representing the 9-day implied volatility) is also unbiased and yet efficient.

5. Concluding remarks

Despite asymmetric volatility attracting a great deal of attention, there still seems to be no consensual solution to the puzzle. Our paper presents novel findings in two fields: First, we contribute to the existing literature on realized and implied volatility by studying their relationship in the context of asymmetric volatility and analyzing the recently introduced shortterm implied volatility indicator, the VXST. Second, supported by our analysis, we propose a novel explanation for the asymmetry puzzle by applying the behavioral pattern known as anchoring. Our regression tests of the relationship between RV and IV yields that, in contrast to the prior literature, the two processes are co-integrated both in the long and short term. Furthermore, our event studies confirm the zero-reverting difference between the two in extreme cases as well as in general. Through an economic interpretation of the analysis, we conclude that IV is an unbiased estimator of RV. However, using the VIX, we find significant autocorrelation between the residuals in the long term. Therefore, our results indicate that only the short term estimation using the VXST is efficient. Our empirical results also support our proposed theory for asymmetric volatility. We find that the asymmetric effect of price falls and jumps is less significant on RV (which is measured ex post over the following period) than their effect on IV (that is estimated ex ante at the beginning of the following period). Hence, we conclude that both the asymmetric effect on volatility and the fading property of anchoring are supported by empirical results. Moreover, our short-term event studies suggest that, in contrast to the long-term results, the asymmetric effect is slightly present in the following period as well. Thus, the time-related characteristic of our proposed explanation is valid. Nevertheless, there are still many questions on the topic which require further research. Less is known about the transition of IV of options into the RV of spot prices, the effect of changes in investors’ risk aversion on the asymmetry or about the implementation of the asymmetric effect into derivative pricing. These problems remain to be solved in future studies.