Skip to main content

Nutritious snacks? That’s a rare sight at the show. It’s usually biscotti, apple turnovers, chocolate and lots of cappuccino up for grabs. It’s no surprise the bowl of fruit is still full after several hours.Petrina Gentile/The Globe and Mail

It has become fashionable to compare every financial asset, asset manager or investor with an index of U.S. stocks. For example, many people compare hedge funds against the Standard & Poor's 500 index. We saw this recently as people discussed whether Donald Trump's business career has been more successful than an investment in the S&P 500. This isn't an apples-to-apples comparison. But actually, very few comparisons with the index are apples to apples. This practice grew out of an admirable impulse, but it has gone too far.

The idea of comparing returns to an index is loosely based on decades of financial research. Essentially, it comes from the logic of efficient markets. If markets are close to efficient, then very few investors will achieve above-market returns on a consistent basis without accepting some cost for doing so, such as higher risk or lower liquidity. In other words, an average of all assets – i.e., a value-weighted index such as the S&P 500 – gives you the best risk/return trade-off you can hope for.

But when we compare things with the index we go beyond that simple logic, and we end up making bad comparisons. Here are four reasons why, even in a world of efficient markets, we can expect portfolios or money managers to have returns that are different from the index:

Risk: Portfolios with less risk can be expected to earn lower returns on average. That's why Treasury bonds usually return less than stocks. People still buy Treasuries, because the low risk of Treasuries is compensation for the fact that they don't tend to go up as much as other assets. Leverage also creates risk, but can increase expected returns.

Now, here's the catch: We don't have a good model of risk. The most popular model, the capital asset pricing model, says risk can be captured by "beta" – roughly, returns that swing up along with the index. But this model doesn't work very well in practice, and there's no consensus on what should replace it. So if you see some manager or asset class doing worse than the S&P 500, it might be because it avoids risks that you don't even realize exist.

Liquidity: If you can't sell an asset without triggering a fire sale, then your asset isn't worth as much as it looks on paper. The ability to sell without moving the price is called "liquidity." There are many ways of measuring it and many factors that affect it, but it's clear that liquidity affects returns. Very illiquid assets should be compensated with higher average returns – and this is what we see happening. So if you see an asset manager whose portfolio consistently beats the index, you might want to ask how liquid his or her investments are. Many of Mr. Trump's assets are in real estate, which isn't very liquid. So we would expect Mr. Trump to earn a return that, on average, beats the index.

Correlation: When U.S. investors talk about "the index," we usually mean an index of U.S. stocks, such as the S&P 500. But although U.S. stocks might have been the main risk asset available to investors in 1915, they are only one small part of a huge universe of such assets in 2015. There are international stocks, high-yield bonds, real estate, commodities and much more. This means "the index" isn't the true index of the entire market. So if you're making an index comparison, at least try to compare similar asset classes. For example, Mr. Trump, a real estate investor, should probably be compared not with the S&P 500, but with an index of real estate investment trusts (REITs).

Mean reversion: We know the prices of stocks and other risky assets tend to bounce around more than their long-term fundamental values. These mispricings can take a decade to correct themselves. So if you see a portfolio of stocks that has beaten the index for five years, there is a better-than-average chance this outperformance will reverse itself during the next five.

Report an error