The Case for a Concentrated Portfolio: Why 5 Stocks Beats 50

Modern portfolio theory, popularised by Harry Markowitz in 1952, recommends diversification as the only free lunch in investing. Most financial advisers translate this into holding 20-30 stocks minimum. Warren Buffett called this "di-worsification" — the dilution of good ideas with mediocre ones. The evidence is more nuanced than either camp admits.

The Mathematics of Diversification

Portfolio volatility decreases sharply up to about 12-15 holdings, then diminishes marginally with each additional stock. By 20 holdings, you've captured roughly 90% of the diversification benefit available from any number of stocks. Beyond this point, you're adding business risk exposure with minimal volatility reduction. Holding 100 stocks doesn't halve your risk compared to holding 50.

The Opportunity Cost of Over-Diversification

Buffett has said he would put 50% of his portfolio into his best idea if he were managing a small amount of capital. Munger has publicly held portfolios with 3-4 major positions. Seth Klarman's Baupost Group regularly holds 5-8 core positions. These are not reckless — they are the result of intensive research producing high-conviction ideas that the manager understands better than the market.

The opportunity cost argument is compelling: if you have identified a stock you genuinely believe is 30% undervalued with high confidence, why dilute your return by also holding 29 stocks you're less confident about? The expected value of concentration is higher if and only if your research edge is real.

The Case for Concentration in Net-Net Investing

Graham himself recommended baskets of 30+ net-nets precisely because individual net-nets have high failure rates — some genuinely do go to zero. The Graham net-net basket approach uses diversification to smooth the binary risk of individual picks. This is a specific, empirically validated use case for broader diversification.

However, if you are investing in quality businesses at discounts (not just statistical cheapness), concentration becomes more defensible. A portfolio of 8-12 thoroughly researched, diversely-sectored companies is achievable for a serious investor and has historically outperformed broader portfolios in the hands of skilled practitioners.

Our Recommendation

For most investors: 15-20 positions covering at least 4 sectors and 2 geographies. For active value investors with significant time for research: 10-15 concentrated positions. For net-net Graham investors: 20-30 positions given the higher individual company failure rate. Never hold fewer than 8 regardless of conviction — idiosyncratic risk is too high.

For educational purposes only. Portfolio construction depends on individual circumstances, time horizon, and risk tolerance. Consult a financial adviser.