[[Start here]] → [[What works in stocks?|what works]] → [[diversify for resilience|diversity]] → 20-30 stocks --- In any single year, your losses on an individual stock are capped at 100%, but your upside is unlimited. It’s typical in a portfolio, even in bullish markets, to see some stocks falling 50%, but some more than doubling. If you don’t own enough stocks to ensure the selection of enough winners, your portfolio can significantly underperform. So this poses a question. How many stocks are enough to reduce the chance of underperformance to an acceptable level? A paper by Newbould and Poon[^1] dug into this question with some rather eye opening insights. They were pretty frustrated by much of the research into diversification that showed that [[owning just five stocks significantly reduces portfolio volatility]]. They felt these insights would guide investors into owning too few stocks. > *The results are based on the average of a large number of portfolios. The problem is that the individual investor holds only one portfolio. Unfortunately, any one portfolio need not be typical… At the eight to twenty stock portfolio sizes recommended, average return is very volatile.* They go on to simulate the performance ranges of portfolios of different sizes for small and large stocks. The range of results is illustrated in the below graphic. ![[newbould-poon-distribution-of-returns.png]] It’s easy to read the above chart. As the number of stocks increases (to the right), the range of return outcomes diminishes around the average return. It’s worth noting that the above graphic charts the lines at the “95%” limits. They believed that rather than imposing a “universal judgement on minimum portfolio sizes”, that investors should make their own decision based on their comfort level with the above ranges. Given that the majority of portfolios end up closer to the average than the extremes in the above chart, I’m more comfortable than Newbould and Poon with more concentrated portfolios. I’ve long believed that the “goldilocks range” for private investors is somewhere between 12 and 30 stocks. Highly experienced investors can invest at the lower end of the range (relying more on their stock analysis), while less experienced should invest at the higher end of the range (relying more on factors). I’ve run my own simulation on portfolios of high StockRank stocks in the UK[^2], similarly to the above, for a single, typical year. It shows the distribution much more clearly. While owning the worst sets of stocks at each portfolio size would be galling, the risks are hugely diminished by owning a 15-20 stock portfolio. More adventurous investors might feel there’s a sweet spot at 10-15 stocks. ![[distribution-of-portfolio-returns-at-different-sizes.png]] [^1]: [[Newbould & Poon - Portfolio Risk, Portfolio Performance and the Individual Investor]] [^2]: [[Croft - Investing with the StockRanks]]