How many stocks does an investor need to be truly diversified? In The Intelligent Investor, first published in 1949, Benjamin Graham suggested that you could minimize portfolio volatility with as few as 10 stocks, and that idea persisted for decades. By the 1980s it was still common for finance textbooks to suggest that once a portfolio contained 10 or 15 stocks, anything more was “superfluous diversification.”
Today, the conventional wisdom has not changed much—if you Google “how many stocks should I own,” most of the results suggest that about 20 or 30 is enough. Many investors have taken this to heart, building portfolios of a couple of dozen large-cap Canadian and U.S. stocks and considering themselves adequately diversified.
Unfortunately, this idea is at odds with the latest academic research. The fact is, a portfolio of just 20 or 30 stocks is far more risky than most investors believe.
The first clue came from a paper published in 1999 by Burton Malkiel, the Princeton economist and author of the classic book A Random Walk Down Wall Street. Malkiel presented evidence that the volatility of individual stocks increased during the 1980s and 1990s—even though the volatility of the overall market remained the same. Others have suggested that the price swings of individual companies are even greater today because of high-frequency trading and more institutional ownership.
For these reasons, investors who want to reduce volatility must now own more stocks than they did a few decades ago. The most compelling analysis was published in 2008 by Jim Davis, an academic and vice-president of research at Dimensional Fund Advisors. Davis came to the shocking conclusion that a fully diversified portfolio now requires not a few dozen stocks, but a few thousand.
Consider this basic question: why do we take risks with our investments? The answer, of course, is that we expect to be compensated with higher returns. In most cases, we are: to give an obvious example, stocks are riskier than bonds, but they have amply rewarded long-term investors for that added risk.
Yet not all risks carry the promise of higher returns. Driving a car is a risky activity that compensates you with improved quality of life, but removing your seatbelt adds to the risk while offering no additional benefit. Failing to buckle up is what an academic would call an uncompensated risk.
You can think of investing in similar terms. Owning a broadly diversified stock portfolio carries an expectation of reward commensurate with the risk. Concentrating a portfolio in a small number of companies adds more volatility and the risk of larger losses, yet it does not increase your expected long-term returns. On the contrary, Davis estimated that a 50-stock portfolio would actually have to produce excess returns of 1.2% annually to compensate the investor for the additional risk—and that is a tall order.
Back in Ben Graham’s day, building a portfolio of 10 carefully selected companies was the best that small investors could do. But as Davis points out, owning a broadly diversified portfolio today is actually cheaper and easier than building a concentrated portfolio of individual companies. Which prompted him to ask, “If the diversification benefits of a market-like portfolio are readily available at a reasonable cost, why should anyone own fewer stocks?”
Why, indeed. Failing to properly diversify your portfolio is saying no to a free lunch.