Turnover trauma?

The way risk and benchmarks are used by the investment industry is wrong and damaging to its clients. Different perceptions of risk bias the behaviour of the plan sponsor and the fund manager in important and different ways. On the one hand, plan sponsors are interested in maximizing the long-run return of a high-duration pool of assets. At the same time there are several accounting and institutional biases that can prevent them from allocating and assessing their investments with a suitably longterm horizon. On the other hand, the fund manager is subject to the greatest number of biases—and the most damaging ones.

A fund manager’s fiduciary duty to the client—maximizing long-term value—is often diluted by his perception of personal and business risk. This often leads to sub-optimal investment behaviour and significant agency risk. There are many fund managers operating with success and integrity. However, in aggregate, the industry seems to have moved too far from being a profession that serves its clients to a business serving its owners.

Is it enough time?
In defence of my own profession, it is not entirely the fault of the fund manager. I would argue that the periods in which we are measured and the process through which mandates are appointed and terminated is too short. James Montier, a global strategist with Dresdner Kleinwort Wasserstein, recently ran a simulation in which he created a universe of 100 ideal fund managers, with 3% annual outperformance and 6% tracking error, simulating their performance over 50 years. A third of these managers underperformed in any given year, and nearly half suffered three consecutive years of underperformance. I think I am on fairly safe ground in assuming that a significant number of them would have been fired for what is, by definition, a statistically insignificant period of underperformance in the middle of a very positive long-term trend.

The focus on short-term relative performance has several damaging consequences, the worst being the compression of investment time horizons. Given the random nature of short-run investment performance, avoiding noticeable underperformance in any quarter can only be achieved by taking very little risk relative to the benchmark. In an attempt to influence the next quarter’s numbers, we fund managers, who tend to be an overconfident bunch, are sucked in to making increasingly short-term decisions. Statistics are available on mutual fund turnover that I think are relevant to the institutional side as well. These show that turnover of stocks has risen from 20% in the mid-1960s to 112% in 2004 (i.e. from an average holding period of five years, to a holding period of 11 months).

This is intuitively strange. Can a fund manager receive enough relevant information about the companies in which he is investing to sell every single one of them in a year? It also imposes large costs upon the fund. Research suggests that the round-trip trading costs of 100% turnover, combined with the tendency to buy the wrong stocks, subtract around 1% from performance per year.

Long-term view
The most successful long-term investors of the last 50 years have two main characteristics: owning relatively few stocks and investing in them for very long periods of time. It is a surprise that, in a supposedly efficient market, such a widely observed and successful approach has not been copied to death. In fact, quite the opposite has happened, with ever increasing numbers in the investment community turning over their whole funds in a year. I would go as far as to call this speculation rather than investing and would encourage plan sponsors to take a longer-term view.

—Nick Thomas, investment manager, Baillie Gifford International LLC

For a PDF version of this article, click here.