Muhammad A. Cheema

Cross-Sectional and Time Series Momentum Returns and Market States

Recent evidence on momentum returns shows that the time-series (TS) strategy outperforms the cross-sectional (CS) strategy. We present new evidence that this happens only when the market continues in the same state, UP or DOWN. In fact, we find that the TS strategy underperforms the CS strategy when the market transitions to a different state. Our results also show that the difference in momentum returns between TS and CS strategies is related to both the net long and net short positions of the TS strategy.

Cross-sectional and time-series momentum returns and market dynamics: evidence from Japan

ABSTRACT We test the behavioural theories of overconfidence and underreaction on cross-sectional (CS) and time-series (TS) momentum returns in the Japanese stock markets. Both CS and TS momentum returns are large and significant when the market continues in the same state and turns into losses when the market transitions to another state, consistent with the overconfidence but not the underreaction model. We find that TS conditional momentum returns exceed conditional CS momentum returns because of its active position since TS takes a net long (short) position following UP (DN) markets while CS is a zero-cost strategy irrespective of the market state. Finally, we find no relation between idiosyncratic volatility (IV) and momentum returns which is not supportive of either the overconfidence or underreaction model but implies that IV is not a significant limit to arbitrage in Japan.

Cross-Sectional and Time-Series Momentum Returns and Market Dynamics: Evidence from Japan

We test the behavioural theories of overconfidence and underreaction on cross-sectional (CS) and times-series (TS) momentum returns in the Japanese stock markets. Both CS and TS momentum returns are large and significant when the market continues in the same state and turns into losses when the market transitions to another state, consistent with the overconfidence but not the underreaction model. We find that TS conditional momentum returns exceed conditional CS momentum returns because of its active position since TS takes a net long (short) position following UP (DN) markets while CS is a zero-cost strategy irrespective of the market state. Finally, we find no relation between idiosyncratic volatility and momentum returns which is not supportive of either the overconfidence or underreaction model but implies that idiosyncratic volatility is not a significant limit to arbitrage in Japan.

Cross-Sectional and Time-Series Momentum Returns and Market Dynamics: Are Islamic Stocks Different?

We search for differences in both unconditional and conditional momentum returns of Islamic and Non-Islamic stocks and test implications of competing behavioral theories that aim to explain momentum returns. Our results show that there is no significant difference in momentum returns between Islamic versus Non-Islamic stocks with respect to both cross-sectional (CS) and time-series (TS) momentum strategies even when we condition momentum returns on market dynamics, information uncertainty (IU), and idiosyncratic volatility (IV). We also find that the TS strategy outperforms (underperforms) the CS strategy in market continuations (transitions) consistent with the recent evidence in the U.S. market. Furthermore, we find that CS and TS strategies of both Islamic and Non-Islamic stocks are profitable only when the market continues in the same state consistent with overconfidence driving momentum returns of both Islamic and Non-Islamic stocks.

Momentum and Market Dynamics: Are Chinese B and H Shares Different?

Research literature shows weak momentum profitability in China due to the trading behaviour of local retail investors who trade in Chinese A-shares. Therefore, we suggest momentum strategy would be profitable in non-A (B and H) shares where foreign investors play an important role instead of local retail investors. Consistent with our suggestion, we find significant momentum returns in non-A shares that are similar in magnitude to the U.S. market. Furthermore, we show that momentum returns in non-A shares are large and significant when the market continues in the same state, consistent with the overconfidence model of Daniel, Hirshleifer, and Subrahmanyam (1998).

Investor Sentiment Dynamics, the Cross-Section of Stock Returns and the MAX Effect

Recent evidence shows that investor sentiment is a contrarian predictor of stock returns with speculative stocks earning lower (higher) future returns than safe stocks following high (low) sentiment states. We extend this argument by conditioning expected stock returns on sentiment dynamics and show that the mispricing of speculative and safe stocks worsens with sentiment continuations but is corrected with sentiment transitions, consistent with the view that the mispricing of these stocks is sentiment-driven. We show that the unconditional contrarian return predictability of sentiment, at least in the short-run, is due to the returns of stocks in sentiment transitions. Results show that ex post, sentiment is a momentum predictor if subsequent sentiment continues; and a contrarian predictor if subsequent sentiment transitions. We also show that the MAX effect can either be positive or negative contingent on sentiment dynamics. The absence of a negative MAX effect following Low sentiment states suggested by prior studies is due to the completely offsetting negative MAX effect when sentiment continues in a Low state and the positive MAX effect when sentiment transitions from a High to a Low state.

Is Investor Sentiment a Reliable Predictor in China

Recent evidence on the relationship between investor sentiment and subsequent monthly market returns in China shows that investor sentiment is a reliable momentum predictor since an increase (decrease) in investor sentiment leads to higher (lower) future returns. However, we suggest that momentum predictability of investor sentiment originates from the boom and bust period of 2006-2008 (the bubble period hereafter). The bubble period is characterized by several months of sustained optimism followed by several months of sustained pessimism, with the market consequently earning high (low) returns following high (low) sentiment months. Therefore, we find a strong positive association between investor sentiment and subsequent market returns during the bubble period. However, investor sentiment has a negligible impact on subsequent monthly market returns once we exclude the bubble period.

Momentum, Idiosyncratic Volatility and Market Dynamics: Evidence from China

Recent evidence on the relation between momentum and idiosyncratic volatility (IV) in the U.S. is mixed. We verify the relation between momentum and IV in China and find at best, no relation supporting the view that idiosyncratic risk is not a significant arbitrage cost for momentum returns. While the absence of a positive relation between momentum returns and IV rejects both the underreaction and the overconfidence and self-attribution stories of momentum, we find support for the overconfidence and self-attribution story from our results on market dynamics and momentum. Our results are robust when verified in other Asian markets. We also find support for the suggestion that cross-country differences in momentum returns could be the result of differences in market dynamics rather than differences in levels of individualism as suggested earlier in the literature.

Cross-Sectional and Time-Series Momentum Returns and Market States

Recent evidence on momentum returns shows that the time-series (TS) strategy outperforms the cross-sectional (CS) strategy. We present new evidence that this happens only when the market continues in the same state, UP or DOWN. In fact, we find that the TS strategy underperforms the CS strategy when the market transitions to a different state. Our results also show that the difference in momentum returns between TS and CS strategies is related to both the net long and net short positions of the TS strategy.