This week in Outside the Box we take a gander at the always-insightful research of good friend James Montier, who poignantly addresses the pertinent topic of portfolio diversification and the pitfalls that ensue on account of benchmarking, wherein investors obsess over relative performance and their respective tracking error. James asks the question, why does the average US mutual fund hold 160 stocks, when diversification could be achieved with around 30-40 stocks. The answer in word, benchmarking.
James Montier posits that the average portfolio manager is focused upon short-term relative performance, paying scant attention to total portfolio risk, rather, the inclination of the average PM is to be primarily concerned with tracking error, that being stock specific or idiosyncratic risk. This misguided focus Montier suggests, leads the PM to manage very large portfolios in their attempt to control stock specific risk, holding nearly 4 times the number of stocks needed to meet diversification targets.
The solution you may ask? Montier suggests the utilization of Monte Carlo simulation to construct a universe of potential portfolios subject to construction rules that define your respective investing universe, thus permitting the measurement of skill to a comparable universe and impelling the manager to focus on absolute return performance.
John Mauldin, Editor
Outside the Box
The Relative Performance Derby And Other Evils Of Modern Investment
Global Equity Strategy, 31 May 2007
Why does the average US mutual fund hold 160 stocks? The usual answer is that this is required for diversification purposes. Now I am not a fan of the use of standard deviation and variance to describe risk (see Global Equity Strategy, 1 February 2007 for more on this). To me it simply doesn't capture…