While momentum as a concept is just about as basic as it gets, there are several factors that need to be carefully considered when formulating a momentum strategy. For inspiration on what strategy to look at, we've turned to a study on momentum in the UK market produced in February last year by Elroy Dimson, Paul Marsh and Mike Staunton of the London Business School in association with ABN Amro. The academics studied momentum strategies over a 108-year period, although the main thrust of the research covers the period between 1955 and 2007, which is what we've examined.
The first thing to decide when looking at momentum is the sort of shares you want to invest in. As short selling – borrowing shares to sell into the market in the hope they will fall – is generally regarded as a valuable tool in momentum investing, an index of large liquid shares, such as the FTSE 100, probably makes the best hunting ground.
Then there's the question of what proportion of the index's best- and worst-performing constituents should be used as the basis of a momentum portfolio. Studies tend to focus on the top- and bottom-performing 10 to 20 per cent of stocks in an index. The momentum strategy we've looked at takes the top and bottom 10 per cent of the FTSE 100 for our 'winners' and 'losers' momentum portfolios. This is the portion of the FTSE 100 that the London Business School study found enjoyed the best annualised returns. Remember, though, that the more concentrated a portfolio is, the more volatile it is going to be.
There is also the question of how money is to be allocated to momentum stock picks. A weighted approach can be used which decides a stock's prominence in a portfolio according to its market capitalisation. Alternatively, an equal weighing can be given to every stock, which tends to be the basis of most studies. In other words, every stock in the top- or bottom-performing 10 is considered to be of equal value at the outset. The London Business School research got better returns from an equal-weighted portfolio, which has the added benefit of being simpler to assemble.
The next question is what time period to use to assess the best- and worst-performing shares in an index from which the portfolio is to be assembled. In momentum jargon, this is known as a ranking period. Choosing the right ranking period is particularly important because momentum portfolios need to be sensitive enough to major changes in sentiment, such as the advent of a bear market, to get investors out of losing positions reasonably fast. But if the ranking period is too short, the portfolio is unlikely to identify major long-term money-making trends, and regular changes in holdings will produce excess trading costs that undermine positive returns.
Choosing winners and losers over a longer time period will mean the constituents of a portfolio will alter less. We’ve looked at a ranking period of one year, which the London Business School academics found got better returns than a portfolio based on the best performers of the past six months. We also tested a portfolio of stocks based on the best performers over a three-month period.
The time stocks are held for before the portfolio is rebalanced is also important in dictating trading costs. We've gone for a three-month holding period for stocks, which the London Business School found delivered the best returns. The effectiveness of the strategy began to deteriorate fairly sharply in the study for longer holding periods. Some strategies also use what is known as a 'skip' period, which is a period of time between the end of the ranking period and assembling the actual portfolio. The advantage of this is that it avoids the danger of buying into a stock that has made it into the top or bottom performers due to extreme performance just before the period end. These extreme spurts of performance are often quickly retraced and can therefore be dangerous. However, we've not used a skip period in our experiment for the sake of simplicity.
