STYLE FACTOR TIMING: AN APPLICATION TO THE PORTFOLIO HOLDINGS OF US FUND MANAGERS
A new study investigates whether an investment strategy which rotates between different investment styles based on macroeconomic signals adds economic value using stock and fund level data. Mutual funds tend to follow an investment style or a blend of styles approach, including Value (buying underpriced stocks), Momentum (investing in stocks which have had high past returns), Size (targeting stocks with small market-capitalisation values) and Quality (purchasing stable, less volatile stocks). Recently, industry players have been focusing on timing the underlying style factors to drive portfolio returns using signals based on the market and macroeconomic environment. Such an approach aids managers in identifying which styles are favourable at a given point in time. Given the success of such timing strategies a new study by Camille Schmidt, Peter Gardner and Professor David Gallagher investigates a style rotation strategy and evaluates performance at the stock and fund levels from 1981 to 2011. The style rotation strategy involves investing in the portfolio of stocks characterised by the style factor which has the highest forecast return over the next period. Forecasts are developed based on 25 market and macroeconomic signals. The strategy involves shifting between long-short portfolios across four styles; Value which is based on the book-to-market ratio, Momentum which is based on stock returns over the prior 12 months, Quality as indicated by Return-on-Equity and Size which is based on market-capitalisation. At the stock level this strategy proves to be successful with positive outperformance on average identified over the 30 year sample period. Forecasts from the model are also tested on the portfolio holdings of a large sample containing 1,856 US active equity mutual funds. The stock level results show that rotating between styles translates into better performance. However investing in a portfolio of mutual funds i.e. a fund-of- funds does not generate outperformance over the sample period. Further investigations as to why the performance identified at the stock level does not translate to the fund level suggest the result is primarily due to the fact that the funds are characterised as long-only and thus unable to attain the long-short style factor returns. The research suggests that the use of long-short style factor portfolios in relation to long-only mutual funds in order to undertake tasks such as performance evaluation may not be appropriate.