Diversification is generally a prudent maneuver by all investors in some form given that so much of the future is inherently unknown. For longer-term, more passive investors, I don’t consider investing in stocks only to be much of a form of diversification. Stocks correlate so highly together that a bear market generally turns almost every stock red. Someone holding the NASDAQ through the dot-com bust – despite owning 100 different stocks – would’ve gotten a Great Depression-level drawdown of more than 80%. This means one would have to make over 5x just to get back to breakeven. This would take nearly 15 years. By comparison, a 40% tech and 60% long-term US Treasuries portfolio would have drawn-down just 24%. Likewise, during the financial crisis, someone holding the S&P 500 alone – seemingly diversified given it contains over 500 stocks – would have drawn-down 51% peak-to-trough. A 40% stocks and 60% LT Treasuries portfolio would have drawn down 15% and would have finished 2008 unchanged, versus nearly a 40% drop for a stocks-only portfolio. Diversification is thus not simply buying a bunch of stocks, but rather allocating funds across multiple asset classes with low correlations to each other. Remaining entirely invested in stocks will put one in the asset class that generates the highest returns (provided enough time), but the risk profile of the underlying portfolio, relative to something more balanced, increases exponentially and not linearly. In the graph below – showing the amount of risk that would theoretically be removed from a portfolio if one invested in X number of assets perfectly uncorrelated to each other – as you go more to the right on the curve (become more diversified) the benefits of diversifying diminish. However, as you go left (become more concentrated), the risk of the portfolio increases in steepening non-linear fashion. (click to enlarge) With that said, I wanted to comment less on other asset classes and more on how many stocks it likely takes to genuinely “diversify” the equities aspect of a portfolio. I did this by choosing the following 40 large-cap stocks below. All have been publicly traded since at least 2007, which is needed to understand a grasp of long-run volatility expectations given they all traded through at least one bad bear market: AAPLGOOGMSFTAMZNNVDAAMDMCDKOPEPC INTCMUTVZPCLNEXPEBACJPMAXPMA MRKBMYPFECELGCVXXOMPOTSOBAVRX SBUXGILDFWFCNFLXCSCOIBMWMTNKEGE I did this in five steps: (1) comparing the volatility of a portfolio of the first block (AAPL --> C) to the volatility of the S&P 500 (2) comparing the volatility of the first and second blocks (AAPL --> MA) to the S&P (3) comparing the first, second, and half the third block (AAPL --> CVX) to the S&P (4) comparing the first, second, and third blocks (AAPL --> VRX) to the S&P, and (5) all to the S&P Results First block (10 total) Annual volatility of 19.6% versus 13.8% for the S&P (42.0% higher) (click to enlarge) First and second blocks (20 total) Annual volatility of 19.5% versus 14.8% for the S&P (31.8% higher) Note: S&P vol changed from above due to some stocks only testing back to some part of 2006, unlike the stocks in the first block, which all have data going back to September 2004. First, second, and half the third block (25 total) Annual volatility of 17.6% versus 14.8% for the S&P (19.4% higher) First, second, and third blocks (30 total) Annual volatility of 16.6% versus 14.8% for the S&P (12.3% higher) All (40 total) Annual volatility of 16.3% versus 14.8% for the S&P (10.1% higher) Conclusion After about 30 individual stock holdings, there become diminishing returns to where the annual volatility reduction of the next stock becomes less than 10 bps. While I wouldn’t suggest that owning around 30 individual stocks is a diverse portfolio by any means – achieving such would necessitate ownership across several different asset classes (or some exposure to safe bonds at the very least) – it can come close to diversifying at least the equities portion of it.