South African stock return predictability in the context data mining: The role of financial variables and international stock returns

In this paper, we examine the predictive ability, both in-sample and the out-of-sample, for South African stock returns using a number of financial variables, based on monthly data with an in-sample period covering 1990:01 to 1996:12 and the out-of-sample period of 1997:01 to 2010:04. We use the t-statistic corresponding to the slope coefficient in a predictive regression model for in-sample predictions, while for the out-of-sample, the MSE-F and the ENC-NEW tests statistics with good power properties were utilised. To guard against data mining, a bootstrap procedure was employed for calculating the critical values of both the in-sample and out-of-sample test statistics. Furthermore, we use a procedure that combines in-sample general-to-specific model selection with out-of-sample tests of predictive ability to further analyse the predictive power of each financial variable. Our results show that, for the in-sample test statistic, only the stock returns for our major trading partners have predictive power at certain short and long run horizons. For the out-of-sample tests, the Treasury bill rate and the term spread together with the stock returns for our major trading partners show predictive power both at short and long run horizons. When accounting for data mining, the maximal out-of-sample test statistics become insignificant from 6-months onward suggesting that the evidence of the out-of-sample predictability at longer horizons is due to data mining. The general-to-specific model shows that valuation ratios contain very useful information that explains the behaviour of stock returns, despite their inability to predict stock return at any horizon. The model also highlights the role of multiple variables in predicting stock returns at medium- to long run horizons.     

Is Beta-t-EGARCH(1,1) superior to GARCH(1,1)?

Statistical performance, in-sample point forecast precision and out-of-sample density forecast precision of GARCH(1,1) and Beta-t-EGARCH(1,1) models are compared. We study the volatility of nine global industry indices for period from April 2006 to July 2010. Competing models are estimated for periods before, during and after the United States (US) financial crisis of 2008. The results provide evidence of the superior out-of-sample predictive performance of Beta-t-EGARCH compared to GARCH after the US financial crisis.     

Model stability and forecast performance of Beta-t-EGARCH

We show that the model stability of the recent QAR(1) plus Beta-t-EGARCH(1,1) is superior to that of the well-known ARMA(1,1) plus t-GARCH(1,1) because QAR plus Beta-t-EGARCH discounts extreme observations, while ARMA plus t-GARCH accentuates them. Model stability of QAR plus Beta-t-EGARCH is an elegant property; however, we show that the out-of-sample density forecast performance of ARMA plus t-GARCH is superior to that of QAR plus Beta-t-EGARCH. We study model stability and density forecast performance for a set of rolling data windows. We use data on the S&P 500 index for the period 1990–2015. For robustness analysis, we also study Monte Carlo simulations of asset returns for the stochastic volatility model.     

QARMA-Beta-t-EGARCH versus ARMA-GARCH: an application to S&P 500

Statistical performance and out-of-sample forecast precision of ARMA-GARCH and QARMA-Beta-t-EGARCH are compared. We study daily returns on the Standard and Poor’s 500 (S&P 500) index and a random sample of 50 stocks from the S&P 500 for period May 2006 to July 2010. Competing models are estimated for periods before and during the US financial crisis of 2008. Out-of-sample point and density forecasts are performed for periods during and after the US financial crisis. The results provide evidence of the superior in-sample statistical and out-of-sample predictive performance of QARMA-Beta-t-EGARCH.     

Momentum meets value investing in a small European market

In this paper, we investigate two prominent market anomalies documented in the finance literature – the momentum effect and value-growth effect. We conduct an out-of-sample test to the link between these two anomalies recurring to a sample of Portuguese stocks during the period 1988–2015. We find that the momentum of value and growth stocks is significantly different: growth stocks exhibit a much larger momentum than value stocks. A combined value and momentum strategy can generate statistically significant excess annual returns of 10.8%. These findings persist across several holding periods up to a year. Moreover, we show that macroeconomic variables fail to explain value and momentum of individual and combined returns. Collectively, our results contradict market efficiency at the weak form and pose a challenge to existing asset pricing theories.     

The Effect of Ex Ante Price on Momentum Profits

We find a strong effect of component stock prices (as of one year before the returns-ranking period) on the magnitude and duration of momentum. Relative strength portfolios formed of high-priced stocks earn statistically significant momentum profits for any holding period in the first three to four years. The effective annual return of the high-priced momentum portfolio for the first year is economically significant at 18.4%, after controlling for the capitalization, trading volume, and unconditional mean effects. This return is considerably higher than the 11.3% earned by low-priced momentum portfolios. Although the price level is correlated with capitalization and trading volume, the price effect is not a mere manifestation of the capitalization and volume effects, as it endures even when the other factors are controlled for. We discuss several implications of our results for the existing behavioral and risk-based explanations of momentum, as none of these models have an explicit role for the price effect.     

Long range dependence in an emerging stock market’s sectors: volatility modelling and VaR forecasting

This study evaluates the sector risk of the Qatar Stock Exchange (QSE), a recently upgraded emerging stock market, using value-at-risk models for the 7 January 2007–18 October 2015 period. After providing evidence for true long memory in volatility using the log-likelihood profile test of Qu and splitting the sample and dth differentiation tests of Shimotsu, we compare the FIGARCH, HYGARCH and FIAPARCH models under normal, Student-t and skewed-t innovation distributions based on in and out-of-sample VaR forecasts. The empirical results show that the skewed Student-t FIGARCH model generates the most accurate prediction of one-day-VaR forecasts. The policy implications for portfolio managers are also discussed.     

Markov regime-switching Beta-t-EGARCH

We suggest a Markov regime-switching (MS) Beta-t-EGARCH (exponential generalized autoregressive conditional heteroscedasticity) model for U.S. stock returns. We compare the in-sample statistical performance of the MS Beta-t-EGARCH model with that of the single-regime Beta-t-EGARCH model. For both models we consider leverage effects for conditional volatility. We use data from the Standard Poor’s 500 (S&P 500) index and also a random sample that includes 50 components of the S&P 500. We study the outlier-discounting property of the single-regime Beta-t-EGARCH and MS Beta-t-EGARCH models. For the S&P 500, we show that for the MS Beta-t-EGARCH model extreme observations are discounted more for the low-volatility regime than for the high-volatility regime. The conditions of consistency and asymptotic normality of the maximum likelihood estimator are satisfied for both the single-regime and MS Beta-t-EGARCH models. All likelihood-based in-sample statistical performance metrics suggest that the MS Beta-t-EGARCH model is superior to the single-regime Beta-t-EGARCH model. We present an application to the out-of-sample density forecast performance of both models. The results show that the density forecast performance of the MS Beta-t-EGARCH model is superior to that of the single-regime Beta-t-EGARCH model.     

Forecasting US real house price returns over 1831–2013: evidence from copula models

Given the existence of nonnormality and nonlinearity in the data generating process of real house price returns over the period of 1831–2013, this article compares the ability of various univariate copula models, relative to standard benchmarks (naive and autoregressive models) in forecasting real US house price over the annual out-of-sample period of 1874–2013, based on an in-sample of 1831–1873. Overall, our results provide overwhelming evidence in favour of the copula models (Normal, Student’s t, Clayton, Frank, Gumbel, Joe and Ali-Mikhail-Huq) relative to linear benchmarks, and especially for the Student’s t-copula, which outperforms all other models both in terms of in-sample and out-of-sample predictability results. Our results highlight the importance of accounting for nonnormality and nonlinearity in the data generating process of real house price returns for the US economy for nearly two centuries of data.     

WHY HAS TOP EXECUTIVE COMPENSATION INCREASED SO MUCH IN CHINA: A EXPLANATION OF PEER‐EFFECTS

Top executive compensation can be affected significantly by peer group pay. This paper investigates the impact of peer effects on the change in top executive compensation based on evidence from China. Empirical results show that if the top three executives' compensation was lower than the peer group median level in year t ? 1, the percentage change in the top three executives compensation in year t would be higher by 0.225%, and that the absolute level of pay would increase by 51 000 yuan. Furthermore, better performance, faster growth and state ownership increase the likelihood of peer effects, while corporate governance variables do not.     

Pairs trading strategies in a cointegration framework: back-tested on CFD and optimized by profit factor

Statistical arbitrage is based on pairs trading of mean-reverting returns. We used cointegration approach and ECM-DCC-GARCH to construct 98 pairs of 152 stocks of 3 currencies. Stocks trading is done by Contract for Difference (CFD), a financial derivative product which facilitates short selling and provides a leverage up to 25 times. To measure the performance of a leveraged strategy, we introduced the profit factor which is the annualized return rate per unit risk. And the historical risk is measured by maximum drawdown. We compared three main strategies: percentage, standard deviation of cointegration long-term residuals and Bollinger Bands (dynamic standard deviation), with and without double confirmation of short-term standard deviation modelled by ECM-DCC-GARCH. Each of the three main strategies is optimized by two optimizers: absolute profit and profit factor. The optimization period goes from 2012–01-01 to 2014–12-31, and validation period is from 2015–01-01 to 2016–06-01. Our results showed that the USD Bollinger Bands strategy without double confirmation and optimized by profit factor, outperformed other strategies and provided the highest annualized return rate per unit risk; 32% of our sample pairs ended up in loss, and 94% of which are explained by a cointegration break during the testing period.     

Intraweek and intraday trade anomalies: evidence from FOREX market

In this article, we search for the evidence of intraweek and intraday anomalies on the spot foreign exchange (FOREX) market. Having in mind the international scope of this market, empirical evidence against market efficiency (i.e. market anomalies) will have important consequences for the substantial number of FOREX investors all around the globe. We explore intraweek, intraday and interaction between days and hour trade anomalies on the FOREX market over the period of 10 years using hourly time-series data of Euro and US Dollar (EUR/USD) exchange rate on Swiss FOREX market from 1 January 2004 to 11 January 2014. We compare by analysis of variance test all pairs of mean returns on a daily, hourly and daily/hourly basis. t-Test is used to test whether intraday returns are significantly different from zero. We employ Tukey’s honestly significant difference test to explore which intraday pairs of hourly mean returns are significantly greater than zero. We find that intraday and interaction between day and hour anomalies are present in trading EUR/USD on the spot FOREX market over the period of 10 years. The best arbitrage opportunity is evidenced on Fridays, when selling USD and buying EUR at 00:00 and selling EUR and buying USD at 03:00 the same day.     

Why do popular mechanisms lack efficiency in random environments?

We consider the problem of randomly assigning n indivisible objects to n agents. Recent research introduced a promising mechanism, the probabilistic serial that has superior efficiency properties than the most common real-life mechanism random priority. On the other hand, mechanisms based on Gale's celebrated top trading cycles method have long dominated the indivisible goods literature (with the exception of the present context) thanks to their outstanding efficiency features. We present an equivalence result between the three kinds of mechanisms, that may help better understand why efficiency differences among popular mechanisms might arise in random environments. This result also suggests that the probabilistic serial and the random priority mechanisms can be viewed as two top trading cycles based mechanisms that essentially differ in the initial conditions of the market before trading starts.     

An empirical estimation for mean-reverting coal prices with long memory

In this paper we discuss the calibration issues of power models built on mean-reverting processes combined with long memory. The unknown parameters of fractional mean-reversion processes are estimated by a hybrid estimation method, which is built upon the marriage of the quadratic variation and the least squares. We perform a simulation study to test the efficiency of these estimators and to compare with the approach proposed by Høg (1999). Moreover, we apply our estimation procedure to some sample series of Chinese coal spot prices in real life situations. These results support the use of fractional mean-reversion processes in modeling Chinese coal prices.     

The Role of Industry Effects in Simultaneous Reversal and Momentum Patterns in One-Month Stock Returns

When stocks are ranked by returns in one month, the portfolio of loser stocks tends to outperform the portfolio of winner stocks in the subsequent month. Yet industry portfolios tend to display momentum. We develop a model of information diffusion among agents with constrained information processing ability that reconciles these well-documented phenomena. We test whether this model or the overreaction hypothesis is consistent with the data. Additionally, a trading strategy based on the model outperforms strategies based on overreaction and on industry momentum. The strategy produces abnormal returns while controlling for market–risk and the size, book value, January, momentum, and liquidity effects.     

The anatomy of returns from moving average trading rules in the Russian stock market

This paper examines the profitability of index trading strategies that are based on dual moving average crossover (DMAC) rules in the Russian stock market over the 2003–2012 period. It contributes to the existing technical analysis (TA) literature by comparing for the first time in emerging markets the relative performance of individual stocks’ trading portfolios with that of trading strategies for the index that consists of the same stocks (i.e., the most liquid stocks of the Moscow Exchange). The results show that the best trading strategies of the in-sample period can outperform buy-and-hold strategy during the subsequent out-of-sample period, although with low statistical significance. In addition, we document the benefits of using DMAC combinations that are much longer than those employed in previous TA literature. Moreover, the decomposition of the full-sample-period performance into separate bull- and bear-period performances shows that the outperformance of the best past index trading strategies over is mostly attributable to the fact that they managed to stay mostly out of the stock market during a dramatic crash caused by the global financial crisis.     

Dividend policy and earnings: a study of short- and long-term causality

This research examines whether earnings per share (EPS) and dividends per share (DPS) exhibit a short and long causality. The data employed in this study consist of quarterly EPS and DPS for 28 of the DJIA companies obtained from Bloomberg over a recent 10-year period. The companies under investigation all have EPS and DPS data available over the period studied. Dividends are generally paid out of earnings. The amount and timing of the dividend paid is a function of the respective company’s dividend policy. Therefore, the EPS_t can be expressed in terms of the DPS_t as follows: EPS_t = αDPS_t where α is a nonnegative constant. The equation suggests that there is a linear relationship between the EPS_t and the DPS_t. The results of this study indicate that bi-directional causality exists for some of the companies.     

Panel data analysis of multi-factor capital asset pricing models

In this article, we propose MFCAPM panel models with fixed effects and test theories associated with risk exposures and anomalies postulated by Fama and French, and we assess their out-of-sample predictive performances. Based on the portfolios formed by French, we construct 10 panel models, each consisting of 10 portfolios grouped by size deciles, and another 10 panels by value deciles. In the presence of cross-section dependence, the MFCAPM panel model is estimated by the feasible generalized least squares (FGLS) method for the sample period 1963(1)-2018(9). The results show that the market, firm-size and value risk exposures are significant and robust across three-, five- and six-factor panel models. Significant time-fixed effects indicate that there are several portfolios resilient to dot.com bubble peak in 2000, while some others resilient to GFC in 2007. We estimate the models for the in-sample period 1963(1)–1999(12) and generate the out-of-sample portfolio returns for the period 2000(1)–2018(9). We find that portfolio returns forecasts generated by the six-factor panel model are superior to other MFCAPM panel models, mostly due to the momentum factor (investor behaviour) explaining large return variations and volatility exposures. The findings have implications for investors, security traders and portfolio risk managers.     

A Measure Of The Trade‐Off Between Responsiveness and Nondictatorship for Arrovian Social Welfare Functions

If individuals are never indifferent between distinct alternatives then for any transitive‐valued social welfare function satisfying IIA, and any fraction t, either the set of pairs of alternatives that are socially ordered without consulting more than one individual's preferences comprises at least the fraction t of all pairs, or else the fraction of pairs that have their social ordering determined independently of everyone's s preferences exceeds, or is very close to, 1 ?t. The Pareto criterion is not imposed. (There is also a version of this result for the domain of preferences that admit indifference.)     

An alternative characterization of top trading cycles

This paper introduces two new characterizations of the top trading cycles algorithm. The key to our characterizations is a new condition, independence of irrelevant rankings (IIR). Intuitively, a mechanism satisfies IIR if whenever an agent’s ranking at an object is irrelevant to her assignment, then it is irrelevant to the assignment of all agents. We demonstrate that a mechanism is Pareto efficient, strategy-proof, IIR, and satisfies mutual best if and only if it is top trading cycles. This provides a new insight into what distinguishes top trading cycles from all other efficient and strategy-proof assignment mechanisms. We provide a second characterization in terms of weak Maskin monotonicity. A mechanism satisfies Pareto efficiency, weak Maskin monotonicity, IIR, and mutual best if and only if it is top trading cycles. This allows us to directly compare top trading cycles to known characterizations of the deferred acceptance algorithm in terms of weak Maskin monotonicity.