EXTREME VALUE AT RISK: A SCENARIO FOR RISK MANAGEMENT

This paper uses the generalised extreme value (GEV) distribution to model the extreme losses that are likely to occur during market crashes, in the case of an investor who has long positions in stocks and currencies. The null hypothesis – which tests for normality of asset returns – is rejected due to asymmetry of these returns. We assume that the asymmetric behaviour and volatility of the returns are captured by the shape and scale parameters, respectively, of a GEV distribution. The data set includes stock indices for the United States, Japan, the United Kingdom, Germany, France and South Africa, and the South African rand exchange rates against the US dollar observed from 3 January 2005 to 30 December 2009. In addition, we divide this sample period into two periods: the pre‐crisis period, from 3 January 2005 to 31 December 2007 and the crisis period, from 1 January 2008 to 30 December 2009. We compared the estimates of value at risk (VaR) using an extreme value theory (EVT) model, with the estimates derived from the traditional variance–covariance method and found that during the crisis the 99% extreme VaR estimates are more reliable as they lie within the Basel II green zone. These results suggest that, at higher quintiles, the VaR estimates based on EVT are reliable and more accurate than estimates from the traditional method.     

THE PREDICTABILITY OF STOCK MARKET RETURNS IN SOUTH AFRICA: PARAMETRIC VS. NON‐PARAMETRIC METHODS

This paper compares the forecasting performance of a sub‐class of univariate parametric and non‐parametric models in predicting stock market returns in South Africa. To account for conditional heteroskedasticity in stock returns data, the non‐parametric model is generated by the conditional heteroskedastic non‐linear autoregressive (NAR) model, while the parametric model is produced by the generalised autoregressive conditional heteroskedastic in mean (GARCH‐M) model. The results of the paper show that the NAR as a non‐parametric model performs better than the GARCH‐M model in short‐term forecasting horizon, and this indicates the importance of a distribution‐free model in predicting stock returns in South Africa.     

RISK–RETURN TRADE‐OFF AND BEHAVIOUR OF VOLATILITY ON THE SOUTH AFRICAN STOCK MARKET: EVIDENCE FROM BOTH AGGREGATE AND DISAGGREGATE DATA

The study analyses the nature and behaviour of volatility, the risk–return relationship and the long‐term trend of volatility on the South African equity markets using aggregate level, industrial level and sectoral level daily data for the period 1995‐2009. By employing dummy variables for the Asian and the sub‐prime financial crises and the 11 September political shock, the study further examines whether the long‐term trend of volatility structurally breaks during financial crises and major political shocks. Three time‐varying generalised autoregressive conditional heteroskedasticity models were employed: one of them symmetric, and the other two asymmetric. Each of these models was estimated based on three error distributional assumptions. The findings of the study are as follows: First, volatility is largely persistent and asymmetric. Second, risk at both aggregate and disaggregate level is generally not a priced factor on the South Africa (SA) stock market. Third, the threshold autoregressive conditional heteroscedasticity (TARCH) model under the generalised error distribution is the most appropriate model for conditional volatility of the SA stock market. Fourth, volatility generally increases over time, and its trend structurally breaks during financial crises and major global shocks. The policy and investment implications of the findings are outlined.     

Does Terrorism Affect the Stock‐Bond Covariance? Evidence from European Countries

Using daily stock and bond returns data from four European countries—France, Germany, Spain, and Great Britain—that have been the victims of significant terrorist activity, this study addresses the issue of whether transnational and/or domestic terrorist attacks have affected in any significant manner the time‐varying stock–bond covariance, their returns, and their variances. Stock and bond markets can be influenced and determined not only by the usual array of macroeconomic factors but also by security shocks, such as a terrorist incident, that have the potential to affect investors' sentiment and portfolio allocation decisions. The issue at hand is addressed using a VAR(p)‐GARCH(1,1)‐in‐mean model, and the results reported herein indicate that terrorist attacks trigger a flight‐to‐safety effect primarily in France and Germany and to a smaller degree in Great Britain and Spain.     

中国股市截面收益率再研究:分位数回归方法

分位数回归方法因为考虑了分布函数的各局部信息而比只考虑条件期望的普通最小二乘回归方法更具有优势，特别是在具有厚尾分布的金融数据分析方面，提供了更详尽的信息。本文通过分位数回归方法重新审视中国股市截面收益率的共同风险因子，查看是否存在规模效应与帐面市值比效应。结果发现，分位数回归结果与普通最小二乘结果显著不同，不同分位数下回归系数及其统计显著性都存在巨大差异。股票收益率与规模正相关的规模效应显著，且高收益率部分的正规模效应更加强烈。帐面市值比效应在低收益率部分正相关，高收益率阶段负相关．中间部分不显著。     

Do Monetary Policy Announcements Affect Exchange Rate Returns and Volatility of Returns? Some Evidence from High‐Frequency Intra‐Day South African Data

This paper examines the temporal effect of domestic monetary policy surprises on both returns and volatility of returns of the South African rand/U.S. dollar exchange rate. The analysis in this “event study” proceeds using intra‐day minute‐by‐minute exchange rate data, repo rate data from the South African Reserve Bank's scheduled monetary policy announcements, and market consensus repo rate forecasts. A carefully selected sample over the period August 2003 to November 2017 ensures that the change in monetary policy is exogenous to the exchange rate. We find statistically and economically significant responses in intra‐day high‐frequency South African rand/U.S. dollar exchange rate returns and volatility of exchange rate returns to domestic interest rate surprises, but anticipated changes have no bearing on exchange rate returns and their volatility. The empirical results also indicate that there is an instantaneous response of the rand/dollar exchange rate to monetary policy surprises and that monetary policy news is an important determinant of the exchange rate until at least 42 minutes after the pronouncement – suggesting a high degree of market “efficiency” in its mechanical sense (although not necessarily in the deeper economic‐informational sense) in processing this information. Essentially, the asymmetric GARCH results exhibit no leverage effects – positive and negative information shocks have symmetric effects on conditional variance.     

THE PRICE OF RISK IN THE SOUTH AFRICAN EQUITY MARKET

This paper investigates domestic risk–return behaviour by focussing on the intertemporal relationship between the conditional domestic equity market premium, its conditional variance and its conditional covariance with the international equity market. The paper finds that the domestic equity market prices in both domestic and international diversification risk. The estimated daily price of domestic variance risk is 0.0279% (EAR: 7.28%) for every one unit of expected domestic variance. The estimated daily price of covariance risk is 0.0111% (EAR: 2.83%) for every unit of expected covariance risk. The representative domestic investor values domestic variance more than covariance risk. The variances of domestic and international equity returns are found to be time‐varying, as is the covariance between the two. Evidence is found that the Johannesburg Securities Exchange is not perfectly integrated with the world economy, in an absolute sense. The volatility spillover effect is observed to be both significant and positive. The standard Capital Asset Pricing Model misspecifies the return to domestic risk, biasing the risk–return coefficient upwards. Domestic investors are rewarded for holding internationally diversified portfolios, with an internationally diversified portfolio expected to have an additional daily return of 0.0238% (EAR: 6.29%) for the same level of risk as an entirely domestic equity portfolio.     

Mean‐reverting and Asymmetric Volatility Switching Properties of Stock Price Index,Exchange Rate and Foreign Capital in Taiwan*

The main purpose of the present study is to explore the relationships among stock price index, exchange rate and foreign capital in Taiwan and to detect whether the mean‐reverting and asymmetric volatility switching properties exist in these markets. The multivariate asymmetric nonlinear smooth transition generalized autoregressive conditional heteroskedastic in mean model is used in this study. The empirical results indicate that overbuy and oversell rates of foreign capital influence the movements of the stock price index and the exchange rate. All three conditional means exhibit asymmetric mean‐reverting behavior, with negative returns reverting quicker than positive returns in terms of both speed and magnitude. The empirical results also demonstrate that the conditional heteroskedasticities of these markets are asymmetric, generating different volatility persistence under a prior positive and negative return shock.     

Federal Reserve Communications and Emerging Equity Markets

Work on the impact of U.S. monetary policy on emerging financial markets mostly focuses on official target rate announcements; empirical evidence using data on informal communication channels, such as speeches, is scant. Employing a unique data set covering formal and informal communication channels in a generalized autoregressive conditional heteroskedasticity model framework, we provide comprehensive evidence on the effects of U.S. monetary policy on 17 emerging equity market returns over the period 1998–2009. We find, first, that both monetary policy actions and communications have a significant impact on market returns. Second, target rate change surprises are an important driver of emerging market returns. However, informal communications—particularly when taking into account their higher frequency—have a larger (cumulative) influence on returns than do target rate surprises. Third, during the recent financial crisis, central bank communication played an even more pronounced role. Finally, American emerging markets react more to communications than do non‐American markets.     

Accounting Quality,Liquidity Risk,and Post‐Earnings‐Announcement Drift

Recent microstructure research finds that liquidity risk, in particular its information component, plays an important role in explaining the post‐earnings‐announcement drift (PEAD). We decompose liquidity risk into an accounting‐associated component and a nonaccounting‐associated component and examine their relative importance in explaining PEAD. Our research is motivated by recent findings that liquidity risk is a systematic risk and earnings quality is negatively associated with liquidity risk. We find that the accounting‐associated component is more strongly related to PEAD returns than is its nonaccounting‐associated counterpart. Further analyses reveal that the relation between accounting‐associated liquidity risk and PEAD returns is weaker for firms with greater analyst following. We also find that in a significant market downturn, the relation between accounting‐associated liquidity risk and PEAD returns becomes more pronounced. Our study is the first to document a liquidity risk‐based role of accounting quality in explaining the PEAD phenomenon. It parses out the PEAD risk premia associated with accounting versus nonaccounting sources and, by so doing, sheds light on the role of accounting quality in shaping the liquidity risk‐PEAD returns relation.     

Correlations in Returns and Volatilities in Pacific-Rim Stock Markets

Most studies on the correlations in stock returns and volatilities focus on the contemporaneous relationships and spillover effects in major stock markets such as the US and Japan. This paper adds to the literature by focusing on the dynamic relationship in the volatilities of the returns in the Pacific-Rim stock markets. The causality in variances test method of Cheung and Ng (1996), a multivariate GARCH model and VAR analyses are employed to model conditional volatilities and study the dynamic responses of volatilities to innovations in conditional variances. The results suggest that while the stock markets are correlated in returns and volatilities contemporaneously and with lags, idiosyncratic factors play important roles in national stock markets. In addition, the dynamic adjustment of the market return volatilities can take a much longer time than previously reported in some of the countries studied.     

全球主要石油市场间的信息溢出效应分析

本文选取了全球主要的石油市场组合,以均值、方差、分位数为代理变量,运用Granger因果检验方法研究了市场的信息溢出效应。在过滤掉了均值因果关系后,着重分析了不同市场组合的波动溢出效应,判断了市场间的信息流向。在波动溢出分析的基础上,又进一步检验了极端上涨和极端下跌时的风险溢出关系,提出了非对称风险溢出的概念,并将其应用于石油市场。实证结果表明信息大多从信息效率高的市场流向信息效率低的市场,上涨侧的信息传递速度比下跌侧更高,强风险溢出证据是普遍的。对波动溢出与风险溢出的比较表明,风险溢出继承了波动溢出的大多数特征,表明方差Granger因果关系的存在是风险溢出的主要原因,二者的差异表明高阶矩存在的因果关系也可以导致风险溢出。     

An Index‐Based Model for Determining the Investment Benchmark of Renewable Energy Projects in South Africa

In a globalised world, financial markets observe the optimal level of asset allocation and returns based on risk inherent in the economies. Whether public or private investors, they need to have an optimal return on their investment given the finite resources. In relatively new sectors like grid‐connected renewable energy, many investors face difficulty in assessing proper return, making them more averse to financing such projects, affecting transborder project development opportunities. In developing countries like South Africa, which has tremendous potential for renewable energy projects, an arbitrary choice of the required rate of return for project evaluations can negatively affect funding decisions. This paper explores an index‐based model to make fair estimates of the required equity benchmark internal rate of return (IRR) using financial markets observation for renewable energy projects in South Africa. The index‐based model is parsimonious and captures common macroeconomic factors. More specifically, it provides a simple and effective mechanism to calculate IRR for renewable energy projects given different gestation periods.     

Empirical Analyses of Extreme Value Models for the South African Mining Index

While the classical normality assumption is simple to implement, it is well known to underestimate the leptokurtic behaviour demonstrated in most financial data. After examining properties of the Johannesburg Stock Exchange Mining Index returns, we propose two extreme value models to fit its negative tail with a higher degree of accuracy. The generalised extreme value distribution (GEVD) is fitted using the block maxima approach, while the generalised Pareto distribution (GPD) is fitted using the peaks‐over‐threshold method. Numerical assessment of value‐at‐risk (VaR) estimates indicates that both GEVD and GPD increasingly outperform the normal distribution as we move further into the lower tail. In addition, GEVD produces lower estimates relative to that of the historical VaR, and GPD provides slightly more conservative estimates for adequate capitalisation.     

Credit Rating Agency and Equity Analysts’ Adjustments to GAAP Earnings

Moody's analysts and sell‐side equity analysts adjust GAAP earnings as part of their research. We show that adjusted earnings definitions of Moody's analysts are significantly lower than those of equity analysts when companies exhibit higher downside risk, as measured by volatility in idiosyncratic stock returns, volatility in negative market returns, poor earnings, and loss status. Relative to the adjusted earnings definitions of equity analysts, adjusted earnings definitions of Moody's analysts better predict future bankruptcies, yet they fare significantly worse in predicting future earnings and operating cash flows. These findings persist after controlling for optimism incentives of analysts, reporting incentives of companies, credit rating levels, and industry and year effects. Our findings suggest that credit rating agencies cater to their clients’ demand for a more conservative interpretation of company‐reported performance than what is offered by equity analysts.     

An intraday-return-based Value-at-Risk model driven by dynamic conditional score with censored generalized Pareto distribution

Encouraged by the literary fact that high-frequency data such as intraday returns contribute to estimating the tail risk of daily returns, we propose an intraday-return-based Value-at-Risk (VaR) model driven by dynamic conditional score with censored generalized Pareto distribution (hence, Censored GP-DCS-VaR model), which is a novel parametric VaR approach based on dynamic score-driven model and can incorporate intraday information into daily VaR forecast. This model helps present the dynamic evolution of intraday return distribution and well capture its tail feature. Applying bootstrap or a parametric method, we are allowed to form the daily return distribution in light of intraday data and thus can calculate VaR directly. Empirical analysis using the data of the Chinese stock market shows that our model gain an advantage in the risk estimation of extreme returns, proved by the comparison of out-of-sample forecasts between the Censored GP-DCS-VaR and the realized-GARCH-VaR.     

NO‐ARBITRAGE ONE‐FACTOR MODELS OF THE SOUTH AFRICAN TERM STRUCTURE OF INTEREST RATES

Short‐term interest rate processes determine the term structure of interest rates in an arbitrage‐free market and are central to the valuation of interest rate derivatives. We obtain parameter estimates and compare the empirical fit of alternative one‐factor continuous‐time processes for the South African short‐term interest rate (and hence of arbitrage‐free term structure models) using Gaussian estimation methods. We find support only for diffusions where the interest rate volatility is moderately sensitive to the level of the interest rate. Other common models with restrictions that either preclude this effect, or restrict it to be too high, do not fit the data. Differences in the specification of the drift function have no evident effect on model performance.     

LONG MEMORY IN THE VOLATILITY OF AN EMERGING FIXED‐INCOME MARKET: EVIDENCE FROM SOUTH AFRICA

This paper tests for long memory in volatility of fixed‐income returns; specifically, South Africa's local currency 10‐year government bond, given that the characterisation of stochastic long‐memory volatility is of interest and importance in portfolio and risk management. The long‐memory parameter is estimated using methods based on wavelets, which have gained prominence in recent years. Evidence of long memory in fixed‐income return volatility is conclusively demonstrated across a variety of volatility measures and wavelet forms. This finding suggests a pattern of time dependence, which may potentially be exploited to generate improved volatility forecasting performance especially over long horizons. This paper further extends the extant literature by comparing the predictive power of long‐memory forecasts with those obtained from a standard (short‐memory) generalised autoregressive conditional heteroskedasticity (GARCH) process. The results of this exercise suggest that the information content of long‐memory models does not lead to improved forecast accuracy. The GARCH(1,1) model is shown to provide the best forecasts across most horizons (i.e. daily, weekly and monthly). Forecast performance is further revealed to be sensitive to the choice of volatility proxy used. Finally, the derived volatility forecasts are generally very close, and in some cases, almost indistinguishable.     

Réactions du Marché d’Actions aux Chocs sur les Taux d’Intérêt de Court Terme,le Cas d’un Marché Emergent: la Bourse Régionale des Valeurs Mobilières (BRVM)

Abstract: Macroeconomics in general and interest rate policy in particular are believed, from a theoretical point of view, to act on stock market movements. This paper discusses this issue in the case of the West African stock market by studying the Regional Securities Exchange (la Bourse Régionale des Valeurs Mobilières (BRVM)) stock market reactions to interest rates innovations. Our results, based on VAR analysis, show that the BRVM stock market does not react immediately to short‐run interest rates’ innovations. Nevertheless, backed reactions occur in imminent periods: at the earliest in the second period and at the latest in the sixth period according to VAR models used. Long‐term effects of short‐term interest rates’ innovations on stock prices returns depend on models specified, on the kind of interest rate and on data frequency. Thus, the long‐term effect of central bank rent rates’ innovations is relatively more important than the one related to interbank rate's innovations. The previous situation of the market is, however, the main determinant of the change of stock prices. Our results show also that stock prices and short‐run interest rates have a similar reaction to both exchange rate and inflation rate's innovations.