Mark Dampier is entitled to believe, whatever the evidence to the contrary, that there is any reliable way of selecting superior active managers cost-effectively but he seems to have been misled on the issue of the importance or otherwise of asset allocation to the investment process (Money Marketing, January 11).
He refers in his column to academic studies which claim that “90 per cent of the return comes from asset allocation”. If he is referring to the studies by Brinson et al in 1986 and 1991, that is to repeat a common misreading of their conclusions.
The studies looked at the returns from a number of US pension schemes and the asset classes considered were US equities, US bonds and cash. Overseas assets and real estate were excluded, largely because the funds in question had very little exposure to them, so it is unsound to claim the studies had any relevance to the decision as to what exposure to have to particular geographic regions.
The studies’ conclusions were that over 90 per cent of the variation in returns, not of the actual returns, was attributable to the investment policy, that is, the split between cash, bonds and equities. Thus, it matters not whether the scheme managers were correct in their allocation decisions between those assets.
Clearly, the composition of each scheme would have been driven largely by its expected assets and liabilities, so thestudies were not assessing whether they were correct or not, just what was contributing to the variations in their returns. The assertion that these findings only work “if you get it right” is erroneous. If you get it “wrong”, it still has a big effect.
Subsequent research by Fama and French is actually more helpful in practical terms for investors. They found that around 95 per cent of the returns of a portfolio, positive or negative, could be explained by its exposure to five structural risk factors – credit risk and duration for debt and market, small capitalisation and value exposure for equity. Since investors can gain exposure to these risk factors via passive funds without the enormous costs, both explicit and hidden, of researching and paying for active management, they at least have the option not to have to try to predict the best future performers as long as they have a clear idea of their own objectives and their risk capacity.
One point on which we are in agreement is that “a combination of styles gives the best arrangement for any portfolio”. I am not aware of the “fashionable argument for diversification” but the reason that most portfolios should adopt diversification is that most of us are not blessed with sufficient foresight to predict reliably which asset classes, markets, styles or managers will outperform in the future. Those who are will probably be paying someone else to read this for them rather than doing so themselves.
Investment manager and branch principal,
Bloomsbury Financial Planning,