Narrow Results

The main purpose of this paper is to explore a high-frequency tactical asset allocation strategy. In particular, we investigate the profitability of momentum trading and contrarian investment strategies for equities listed on the Australian Stock Exchange (ASX). In these two strategies we take into consideration the short-selling restrictions imposed by the ASX on the stocks used. Within our sample portfolios we look at the relationship between stock returns and past trading volume for these equities. This research also investigates the seasonal aspects of contrarian portfolios and observes weekly, monthly and yearly effects. We report significant contrarian profits for the period investigated (from 2001 to 2006) and show that contrarian profit is a persistent feature for the strategies examined. We also document that contrarian portfolios earn returns as high as 6.54% per day for portfolios with no short-selling restrictions, and 4.71% in the restricted model. The results also support the view that volume traded affects stock returns, and show that market imperfections such as short-selling restrictions affect investors' returns.

Trend-chasing and Contrarian are well-documented empirical trading patterns that the literature generally attributes to behavioral biases. In contrast, we argue that both are rational portfolio rebalancing strategies in a dynamic asset allocation framework. Analyzing the interactions between strategies implemented in stocks and bonds, we find that a key parameter is the investor's level of relative-risk-aversion versus the market price of risk. Our mapping of preferences to trades fits remarkably well recent empirical findings of time-series momentum and reversal. Specifically, speculators trade like Trend-chasers throughout the momentum phase closing positions once the trend reverses, while hedgers trade like Contrarians. These trades seem to explain time-series momentum.

Cross-sectional characteristics of stocks such as market value, market-to-book ratio and accumulated past return can be applied to formulate equity portfolios in the stock-picking process to generate good profits in some markets, which relate to the well-known size, value and momentum or contrarian strategies in the literature. Alternatively, income measures in financial statements drive investors in stock markets to buy or sell in an intuitive way that can be also used in the stock-picking process to generate good profits in some markets. This study applies these types of information in the formation period to formulate long-short strategies and investigates both the returns and risk profiles in the holding period afterward and checks whether the measures can be used to generate good profits in the Taiwan markets for the period from January 1980 to June 2020. Given different lengths of the holding period and different equity segments, our empirical analysis shows that strategies filtered by the income measure of gross profitability outperform the counterparts filtered by the operating profitability. Moreover, while the momentum or contrarian effect is not, the size and value effects are helpful to improve the performance of long-short strategies filtered by gross profitability.

In this study, the determinants of profits to momentum/contrarian strategies were examined when applied to national stock market indexes. Using monthly stock market index data of 16 countries from December 1969 to December 2000, it is found that momentum strategies are profitable over horizons from 3 to 12 months, while contrarian strategies are profitable for long horizons such as 2 years or longer. However, the profit is statistically significant for only the 6-month horizon. The present decomposition analysis indicates that international momentum/contrarian profits are mainly due to the autocorrelations in these national market index returns, not to cross-serial correlations or to cross-sectional variation in their mean returns. Consistent with the trading strategy results, it is also found that most of the stock market indexes follow a mean-reverting process, implying positive autocorrelations in short-horizon returns and negative autocorrelations in long lags.