Be wary of spreading your portfolio too thinly

For a portfolio manager, time and again studies and mathematical models show that maintaining a well-diversified portfolio of 25 to 30 stocks produces the most cost-effective way of reducing risk.

Investing in a higher number of securities will still yield further diversification benefits – but at a drastically smaller rate. Studies also show that exceeding the sweet spot of 30 holdings means the portfolio becomes too diversified and dilutes the potential for alpha.

Looking at the IA UK All Companies sector, I was surprised to see that very few of the fund managers follow recommendations from academics.

The average number of holdings for a fund in the sector stands at 52.5. Only 17 managers (out of 204 members in the sector) actually run a portfolio with an average number of holdings below or equal to 30.

What does that mean in terms of performance?

Looking on a five-year basis, at first it appears that “concentrated funds” did better, outperforming their benchmark by a higher margin than their “diversified” peers. Better information ratios tell us this result also stands even when adjusted for higher tracking error.

So concentrated funds are more “truly” active than their peers and have been rewarded for taking additional active risk. Interestingly, when you focus on risk, it appears these concentrated funds also did better than their peers in protecting capital.

Although the volatility of their returns is higher, the median beta relative to benchmark is lower, meaning better performance in down markets.

It’s also surprising to see that some managers in the UK equity universe run two portfolios under the same mandate, the only difference being portfolio construction. These managers offer two investment solutions: a “normal” fund and a “concentrated” or “best ideas” fund with a reduced number of holdings.

Comparing the performance between their funds over the last three and five years, the results again show the concentrated portfolios outperform their “normal” peers.

Other results are more mixed when comparing a single manager’s portfolios. “Concentrated” funds have a higher beta and failed to protect from the downside relative to their “normal” peers. Max drawdown is also higher.

But no matter which way you look at it, the academics are on to something. The evidence suggests it pays off to run a portfolio with 30 or fewer holdings.

Should you choose to enquire as to why managers of portfolios with 40 to 50 names reject the suggestions of the academics, you’ll find yourself at the barrel-end of William Wordsworth’s words: “The education of circumstance is superior to that of tuition.”

In more prosaic terms, they believe their experience of equity markets has taught them to run a more diversified portfolio. I believe this is a misjudgement – a simple attribution tool would show they would be better off running a more concentrated portfolio of their best ideas. To quote the best idea of another wordsmith, Oscar Wilde: “Experience is simply the name we give our mistakes.”

Charles Younes is research manager at FE