Forecasting daily variability of the S&P 100 stock index using historical,realised and implied volatility measurements

The increasing availability of financial market data at intraday frequencies has not only led to the development of improved volatility measurements but has also inspired research into their potential value as an information source for volatility forecasting. In this paper, we explore the forecasting value of historical volatility (extracted from daily return series), of implied volatility (extracted from option pricing data) and of realised volatility (computed as the sum of squared high frequency returns within a day). First, we consider unobserved components (UC-RV) and long memory models for realised volatility which is regarded as an accurate estimator of volatility. The predictive abilities of realised volatility models are compared with those of stochastic volatility (SV) models and generalised autoregressive conditional heteroskedasticity (GARCH) models for daily return series. These historical volatility models are extended to include realised and implied volatility measures as explanatory variables for volatility. The main focus is on forecasting the daily variability of the Standard & Poor's 100 (S&P 100) stock index series for which trading data (tick by tick) of almost 7 years is analysed. The forecast assessment is based on the hypothesis of whether a forecast model is outperformed by alternative models. In particular, we will use superior predictive ability tests to investigate the relative forecast performances of some models. Since volatilities are not observed, realised volatility is taken as a proxy for actual volatility and is used for computing the forecast error. A stationary bootstrap procedure is required for computing the test statistic and its p-value. The empirical results show convincingly that realised volatility models produce far more accurate volatility forecasts compared to models based on daily returns. Long memory models seem to provide the most accurate forecasts.     

Market imperfections and the information content of implied and realized volatility

The information content of option implied volatility and realized volatility under market imperfections are studied in the context of GARCH modeling and volatility forecasts of Taiwan stock market (TAIEX) returns. Consistent with most studies, we find that the Taiwan implied volatility index (TVIX) calculated from the TAIEX option prices contains most of the information, and that White's [White, H., 2000. A reality check for data snooping. Econometrica 68, 1097–1126] reality check test cannot reject the null hypothesis that the TVIX provides the best forecast. Possibly due to market imperfections, however, the incremental information content of realized volatility as well as daily returns cannot be ruled out. Finally, we also find that the information is found only in the most recent TVIX, indicating information is being efficiently impounded on the TAIEX option prices. This finding suggests that appropriately designed derivative products can alleviate the problems caused by market imperfections.     

The information content of implied volatilities and model-free volatility expectations: Evidence from options written on individual stocks

We measure the volatility information content of stock options for individual firms using option prices for 149 US firms and the S&P 100 index. We use ARCH and regression models to compare volatility forecasts defined by historical stock returns, at-the-money implied volatilities and model-free volatility expectations for every firm. For 1-day-ahead estimation, a historical ARCH model outperforms both of the volatility estimates extracted from option prices for 36% of the firms, but the option forecasts are nearly always more informative for those firms that have the more actively traded options. When the prediction horizon extends until the expiry date of the options, the option forecasts are more informative than the historical volatility for 85% of the firms. However, at-the-money implied volatilities generally outperform the model-free volatility expectations.     

Information content and other characteristics of the daily cross-sectional dispersion in stock returns

We study the cross-sectional dispersion in daily stock returns, or daily return dispersion (RD). Our primary empirical contribution is to demonstrate that RD contains reliable incremental information about the future traditional volatility of both firm-level and portfolio-level returns. The relation between RD and future stock volatility is pervasive across time and across different industry portfolios, size-based portfolios, and beta-based portfolios. Further, our results suggest that RD contains more incremental information about the future volatility of firm-level stock returns than do lagged market-level return shocks. To further characterize RD and assist in interpretation, we also document how dispersion varies with stock turnover and macroeconomic news.     

The Impact of Trades on Daily Volatility

This article proposes a trading-based explanation for the asymmetriceffect in daily volatility of individual stock returns. Previousstudies propose two major hypotheses for this phenomenon: leverageeffect and time-varying expected returns. However, leveragehas no impact on asymmetric volatility at the daily frequencyand, moreover, we observe asymmetric volatility for stocks withno leverage. Also, expected returns may vary with the businesscycle, that is, at a lower than daily frequency. Trading activityof contrarian and herding investors has a robust effect on therelationship between daily volatility and lagged return. Consistentwith the predictions of the rational expectation models, thenon-informational liquidity-driven (herding) trades increasevolatility following stock price declines, and the informed(contrarian) trades reduce volatility following stock priceincreases. The results are robust to different measures of volatilityand trading activity. (JEL C30, G11, G12)     

Relationship between gold and stock markets during the global financial crisis: Evidence from nonlinear causality tests

This paper investigates the nonlinear dynamic co-movements between gold returns, stock market returns and stock market volatility during the recent global financial crisis for the UK (FTSE 100), the US (S&P 500) and Japan (Nikkei 225). Initially, the bivariate dynamic relationships between i) gold returns and stock market returns and ii) gold returns and stock market volatility are tested; both of these relationships are further investigated in the multivariate nonlinear settings by including changes in the three-month LIBOR rates. In this paper correlation integrals based on the bivariate model show significant evidence of nonlinear feedback effect among the variables during the financial crisis period for all the countries understudy. Very limited evidence of significant feedback is found during the pre-crisis period. Results from the multivariate tests including changes in the LIBOR rates provide results similar to the bivariate results. These results imply that gold may not perform well as a safe haven during the financial crisis period due to the bidirectional interdependence between gold returns and, stock returns as well as stock market volatility. However, gold may be used as a hedge against stock market returns and volatility in stable financial conditions.     

MACROECONOMIC NEWS,STOCK TURNOVER,AND VOLATILITY CLUSTERING IN DAILY STOCK RETURNS

We study volatility clustering in daily stock returns at both the index and firm levels from 1985 to 2000. We find that the relation between today's index return shock and the next period's volatility decreases when important macroeconomic news is released today and increases with the shock in today's stock market turnover. Collectively, our results suggest that volatility clustering tends to be stronger when there is more uncertainty and disperse beliefs about the market's information signal. Our findings also contribute to a better understanding of the joint dynamics of stock returns and trading volume.     

Forecasting global stock market volatilities in an uncertain world

We investigate the predictive relationship between uncertainty and global stock market volatilities from a high-frequency perspective. We show that uncertainty contains information beyond fundamentals (volatility) and strongly affects stock market volatility. Using several crucial uncertainty measures (i.e., uncertainty and implied volatility indices), we prove that the CBOE volatility index (VIX) performs best in point (density) forecasting; the financial stress index (FSI) in directional forecasting. Furthermore, VIX's predictive power improved dramatically after the COVID-19 outbreak, and the VIX-based portfolio strategy enables mean-variance investors to achieve higher returns. There are two empirical properties of VIX: (i) it helps reduce significantly forecast variance rather than bias; and (ii) its forecasts encompass other uncertainty forecasts well. Overall, we highlight the importance of considering uncertainty when exploring the expected stock market volatility.     

Volatility Clustering, Asymmetry and Hysteresis in Stock Returns: International Evidence

Encompassing a very broad family of ARCH-GARCH models, we show that the AT-GARCH (1,1) model, where volatility rises more in response to bad newsthan to good news, and where news are considered bad only below a certain level, is a remarkably robust representation of worldwide stock market returns. The residual structure is then captured by extending ATGARCH (1,1) to an hysteresis model, HGARCH, where we modelstructured memory effects from past innovations. Obviously, this feature relates to the psychology of the markets and the way traders process information. For the French stock market we show that votalitity is affected differently, depending on the recent past being characterized by returns all above or below a certain level. In the same way a longer term trend may also influence volatility. It is found that bad news are discounted very quickly in volatility, this effect being reinforced when it comes after a negative trend in the stock index. On the opposite, good news have a very small impact on volatility except when they are clustered over a few days, which in this case reduces volatility.     

The usefulness of cross-sectional dispersion for forecasting aggregate stock price volatility

Does cross-sectional dispersion in the returns of different stocks help forecast volatility of the S&P 500 index? This paper develops a model of stock returns where dispersion in returns across different stocks is modeled jointly with aggregate volatility. Although specifications that allow for feedback from cross-sectional dispersion to aggregate volatility have a better fit in sample, they prove not to be robust for purposes of out-of-sample forecasting. Using a full cross-section of stock returns jointly, however, I find that use of cross-sectional dispersion can help improve parameter estimates of a GARCH process for aggregate volatility to generate better forecasts both in sample and out of sample. Given this evidence, I conclude that cross-sectional information helps predict market volatility indirectly rather than directly entering in the data-generating process.     

STOCK RETURNS AND VOLATILITY ON CHINA'S STOCK MARKETS

We examine time‐series features of stock returns and volatility, as well as the relation between return and volatility in four of China's stock exchanges. Variance ratio tests reject the hypothesis that stock returns follow a random walk. We find evidence of long memory of returns. Application of GARCH and EGARCH models provides strong evidence of time‐varying volatility and shows volatility is highly persistent and predictable. The results of GARCH‐M do not show any relation between expected returns and expected risk. Daily trading volume used as a proxy for information arrival time has no significant explanatory power for the conditional volatility of daily returns. JEL classification: G15     

The incremental volatility information in one million foreign exchange quotations

The volatility information found in high-frequency exchange rate quotations and in implied volatilities is compared by estimating ARCH models for DM/$ returns. Reuters quotations are used to calculate five-minute returns and hence hourly and daily estimates of realised volatility that can be included in equations for the conditional variances of hourly and daily returns. The ARCH results show that there is a significant amount of information in five-minute returns that is incremental to options information when estimating hourly variances. The same conclusion is obtained by an out-of-sample comparison of forecasts of hourly realised volatility.     

Social media bots and stock markets

This study examines the link between information spread by social media bots and stock trading. Based on a large sample of tweets mentioning 55 companies in the FTSE 100 composites, we find significant relations between bot tweets and stock returns, volatility, and trading volume at both daily and intraday levels. These results are also confirmed by an event study of stock response following abnormal increases in the volume of tweets. The findings are robust to various specifications, including controlling for traditional news channel, alternative measures of volatility, information flows in pretrading hours, and different measures of sentiment.     

Incomplete information,idiosyncratic volatility and stock returns

When investors have incomplete information, expected returns, as measured by an econometrician, deviate from those predicted by standard asset pricing models by including a term that is the product of the stock’s idiosyncratic volatility and the investors’ aggregated forecast errors. If investors are biased this term generates a relation between idiosyncratic volatility and expected stocks returns. Relying on forecast revisions from IBES, we construct a new variable that proxies for this term and show that it explains a significant part of the empirical relation between idiosyncratic volatility and stock returns.     

The Determinants of Conditional Autocorrelation in Stock Returns

We investigate whether return volatility, trading volume, return asymmetry, business cycles, and day‐of‐the‐week are potential determinants of conditional autocorrelation in stock returns. Our primary focus is on the role of feedback trading and the interplay of return volatility. We present empirical evidence using conditional autocorrelation estimates generated from multivariate generalized autoregressive conditional heteroskedasticity (M‐GARCH) models for individual U.S. stock and index data. In addition to return volatility, we find that trading volume and market returns are important in explaining the time‐varying patterns of return autocorrelation.     

Historic risk and implied volatility

This study extends the volatility prediction literature with (1) new intraday realized volatility measures and (2) various implied volatility indexes for commodities, currencies, and equities. Predicting volatility is important for academics, investors, and regulators. Applications range from forecasting stock and option returns to constructing early warning systems. Using twenty-three Chicago Board Options Exchange VIX indexes, as opposed to the common S&P 100 and S&P 500 equity indexes, we find a bidirectional lead-lag relationship between implied volatility and realized volatility. The lead-lag relationships are more robust and stronger using suggested intraday volatility measures than using the interday volatility measures that are common in the literature.     

A High‐Frequency Investigation of the Interaction between Volatility and DAX Returns

One of the most noticeable stylised facts in finance is that stock index returns are negatively correlated with changes in volatility. The economic rationale for the effect is still controversial. The competing explanations have different implications for the origin of the relationship: Are volatility changes induced by index movements, or inversely, does volatility drive index returns? To differentiate between the alternative hypotheses, we analyse the lead‐lag relationship of option implied volatility and index return in Germany based on Granger causality tests and impulse‐response functions. Our dataset consists of all transactions in DAX options and futures over the time period from 1995 to 2005. Analyzing returns over 5‐minute intervals, we find that the relationship is return‐driven in the sense that index returns Granger cause volatility changes. This causal relationship is statistically and economically significant and can be clearly separated from the contemporaneous correlation. The largest part of the implied volatility response occurs immediately, but we also observe a smaller retarded reaction for up to one hour. A volatility feedback effect is not discernible. If it exists, the stock market appears to correctly anticipate its importance for index returns.     

VARIANCE,RETURN, AND HIGH-LOW PRICE SPREADS

We report three new findings that rely upon the high-low price range as an estimate of stock return variance. The predictability of variance is associated with persistence in high prices and with correlated shocks to high and low prices. Excess stock returns are positively related to anticipated variance and inversely related to unanticipated variance. Lagged squared residuals in GARCH(1,1) models have no incremental explanatory power in the presence of forecasts of conditional volatility generated from high-low price spread models.     

Box-Cox stochastic volatility models with heavy-tails and correlated errors

This paper presents a Markov chain Monte Carlo (MCMC) algorithm to estimate parameters and latent stochastic processes in the asymmetric stochastic volatility (SV) model, in which the Box-Cox transformation of the squared volatility follows an autoregressive Gaussian distribution and the marginal density of asset returns has heavy-tails. We employed the Bayes factor and the Bayesian information criterion (BIC) to examine whether the Box-Cox transformation of squared volatility is favored against the log-transformation. When applying the heavy-tailed asymmetric Box-Cox transformed SV model, three competing SV models and the t-GARCH(1,1) model to continuously compounded daily returns of the Australian stock index, we find that the Box-Cox transformation of squared volatility is strongly favored by Bayes factors and BIC against the log-transformation. While both criteria strongly favor the t-GARCH(1,1) model against the heavy-tailed asymmetric Box-Cox transformed SV model and the other three competing SV models, we find that SV models fit the data better than the t-GARCH(1,1) model based on a measure of closeness between the distribution of the fitted residuals and the distribution of the model disturbance. When our model and its competing models are applied to daily returns of another five stock indices, we find that in terms of SV models, the Box-Cox transformation of squared volatility is strongly favored against the log-transformation for the five data sets.     

Volatility as an Asset Class: European Evidence

Abstract

Volatility movements are known to be negatively correlated with stock index returns. Hence, investing in volatility appears to be attractive for investors seeking risk diversification. The most common instruments for investing in pure volatility are variance swaps, which now enjoy an active over-the-counter (OTC) market. This paper investigates the risk-return tradeoff of variance swaps on the Deutscher Aktienindex and Euro STOXX 50 index over the time period from 1995 to 2004. We synthetically derive variance swap rates from the smile in option prices. Using quotes from two large investment banks over two months, we validate that the synthetic values are close to OTC market prices. We find that variance swap returns exhibit an option-like profile compared to returns of the underlying index. Given this pattern, it is crucial to account for the non-normality of returns in measuring the performance of variance swap investments. As in the US, the average returns of selling variance swaps are found to be strongly positive and too large to be compatible with standard equilibrium models. The magnitude of the estimated risk premium is related to variance uncertainty and past index returns. This indicates that the variance swap rate does not seem to incorporate all past information relevant for forecasting future realized variance.