It’s widely accepted that your portfolio should be well-diversified to prevent a single disastrous decision from wiping out a large chunk of your wealth. At the same time, if a portfolio is too large, your chances of outperforming the market shrink and you might as well hold a cheap tracker fund.
These extremes are self-evident: the big question is the number of holdings that gives you the best trade-off between return and risk. Research by Miguel Antón of the University of Madrid, Randolph Cohen of Harvard Business School, and Christopher Polk of the London School of Economics gives us an interesting way to think about this. Their paper Best Ideas looks at how many of a fund manager’s highest-conviction stocks make a contribution to beating the market.
Based on holdings in US mutual funds between 1983 and 2018, they found that managers’ top picks outperformed the market by 2.8%-4.5% per year on average (depending on exactly how you measure risk-adjusted performance) – but this only extends to the top three-to-five highest-conviction ones. “The vast majority of the other stocks managers hold do not exhibit significant outperformance.”
Drag on returns
If investors only have a few stellar ideas at any one time, this implies that portfolios grow pretty rapidly to the point where no potential alpha gets added. Given that many funds run portfolios of 50 or even 100 stocks, it’s obvious that the vast majority can’t contribute much and the extra cost and time of looking after them is a drag on returns. However, cutting the portfolio down too much means that risk soars.
With five stocks, the cost associated with one going badly wrong is vast – and even if all ideas ultimately work out, volatility along the way will usually be staggering. For institutions, one answer could be to divide a portfolio up by regions and industries, and hire different managers to run ultra-concentrated five-stock portfolios in each area, suggests Cohen in an interview with Citywire. The result would be a diversified portfolio, but one that does not simply replicate the market.
Private investors can’t do the same and in any case our portfolios are often quite concentrated. Still, this kind of thinking may help. Does your portfolio consist of your very best ideas across several complementary sectors and strategies, or alternatively a smaller number of high-potential picks plus a cheap tracker to diversify away some of the risk? Or have you ended up with a large pool of similar stocks, all driven by the same factors? Do you hold a few distinctive focused funds?
Or have you got a grab-bag of 100-stock monsters that just add up to the market portfolio? In short, assume that you may only make a few good decisions, then try to do so in a way that gives the successes a chance to make a difference.