Robust monitoring machine:a machine learning solution for out‑of‑sample R_(2)‑hacking in return predictability monitoring

The out-of-sample R^(2) is designed to measure forecasting performance without look-ahead bias.However,researchers can hack this performance metric even without multiple tests by constructing a prediction model using the intuition derived from empirical properties that appear only in the test sample.Using ensemble machine learning techniques,we create a virtual environment that prevents researchers from peeking into the intuition in advance when performing out-of-sample prediction simulations.We apply this approach to robust monitoring,exploiting a dynamic shrink-age effect by switching between a proposed forecast and a benchmark.Considering stock return forecasting as an example,we show that the resulting robust monitoring forecast improves the average performance of the proposed forecast by 15%(in terms of mean-squared-error)and reduces the variance of its relative performance by 46%while avoiding the out-of-sample R^(2)-hacking problem.Our approach,as a final touch,can further enhance the performance and stability of forecasts from any models and methods.

Theoretical Explanation of Return Predictability Based on Stock Price Formulation

To find out which factors determine stock return and to give rational explanation of return predictability, according to the principle of stock price formulation, the trend of stock price is obtained by use of option pricing method. The trend of stock price is put into reconstructing CAPM （capital asset pricing model） beta; it is concluded that the firm-specific biases and the scale biases potentiaUy induce return predictabih＇ty. In addition, through the relation between the biases structure and the intrinsic value, an appropriate theoretic explanation is supplied for three-factor pricing model proposed by Fama and French.

Baidu index and predictability of Chinese stock returns

A number of studies have investigated the predictability of Chinese stock returns with economic variables.Given the newly emerged dataset from the Internet,this paper investigates whether the Baidu Index can be employed to predict Chinese stock returns.The empirical results show that 1)the Search Frequency of Baidu Index(SFBI)can predict next day’s price changes;2)the stock prices go up when individual investors pay less attention to the stocks and go down when individual investors pay more attention to the stocks;3)the trading strategy constructed by shorting on the most SFBI and longing on the least SFBI outperforms the corresponding market index returns without consideration of the transaction costs.These results complement the existing literature on the predictability of Chinese stock returns and have potential implications for asset pricing and risk management.

Loan loss provisions and return predictability:A dynamic perspective

This paper examines the impact of loan loss provisions(LLPs)on return predictability during 1994–2017.We find that on average,LLPs are negatively associated with one year ahead stock returns.This effect is particularly significant during the global financial crisis but much weaker during the Basel Ⅱ and Ⅲ periods.Consistent with these findings,a long–short trading strategy based on LLPs generates positive abnormal returns during the Basel Ⅱ and Ⅲ periods but negative abnormal returns during the financial crisis.Cross-sectional tests show that this effect is more pronounced among banks with greater information asymmetry.Decomposition of LLPs suggests that these findings are driven mainly by nondiscretionary LLPs.Overall,our results suggest that the relationship between LLPs and future stock returns is not linear but contingent on bank regulations and macroeconomic conditions.

Timing the market: the economic value of price extremes

By decomposing asset returns into potential maximum gain(PMG)and potential maximum loss(PML)with price extremes,this study empirically investigated the relationships between PMG and PML.We found significant asymmetry between PMG and PML.PML significantly contributed to forecasting PMG but not vice versa.We further explored the power of this asymmetry for predicting asset returns and found it could significantly improve asset return predictability in both in-sample and out-of-sample forecasting.Investors who incorporate this asymmetry into their investment decisions can get substantial utility gains.This asymmetry remains significant even when controlling for macroeconomic variables,technical indicators,market sentiment,and skewness.Moreover,this asymmetry was found to be quite general across different countries.

A statistical learning approach for stock selection in the Chinese stock market

Forecasting stock returns is extremely challenging in general,and this task becomes even more difficult given the turbulent nature of the Chinese stock market.We address the stock selection process as a statistical learning problem and build crosssectional forecast models to select individual stocks in the Shanghai Composite Index.Decile portfolios are formed according to rankings of the forecasted future cumulative returns.The equity market’s neutral portfolio-formed by buying the top decile portfolio and selling short the bottom decile portfolio-exhibits superior performance to,and a low correlation with,the Shanghai Composite Index.To make our strategy more useful to practitioners,we evaluate the proposed stock selection strategy’s performance by allowing only long positions,and by investing only in Ashare stocks to incorporate the restrictions in the Chinese stock market.The longonly strategies still generate robust and superior performance compared to the Shanghai Composite Index.A close examination of the coefficients of the features provides more insights into the changes in market dynamics from period to period.

DOES INVESTOR SENTIMENT PREDICT STOCK RETURNS? THE EVIDENCE FROM CHINESE STOCK MARKET

This paper examines the proxy variables of investor sentiment in Chinese stock market carefully, and tries to construct an investor sentiment index indirectly. We use cross correlation analysis to examine lead-lag relationship between the proxy variables and HS300 index. The results show that net added accounts （NAA）, SSE share turnover （TURN）, and closed-end fund discount （CEFD） are leading variables to stock market. The average first day return of IPOs （RIPO） and relative degree of active trading in equity market （RDAT） are contemporary variables, while number of IPOs （NIPO） is a lagging variable of stock market. Using the sentiment proxy variables with most possible leading order, and forward selection stepwise regression method, the empirical results on monthly stock returns reveal that three leading proxy variables can be used to form a sentiment index. And the out of sample tests prove that this sentiment index has good predictive power of Chinese stock market, and it is robust.

Timing Prediction Error Volatility and Dynamic Asset Allocation

We solve a portfolio selection,problem in which,return predictability,risk predictability and transaction cost are incorporated.In the problem,both expected return,prediction error volatility,and transaction cost are time-varying.Our optimal strategy suggests trading partially toward a dynamic aim portfolio,which is a weighted average of expected future tangency portfolio and is highly influenced by the common fluctuation of prediction error volatility(CPE).When CPE is high,the investor would invest less and trade less frequently to avoid risk and transaction cost.Moreover,the investor trades more closely to the aim portfolio with a more persistent CPE signal.We also conduct an empirical analysis based on the commodities futures in Chinese market.The results reveal that by timing prediction error volatility,our strategy outperforms alternative strategies.