Gateway (GATEX) is the $8 billion behemoth of the long-short equity mutual fund category, and one of the biggest alternative mutual funds. I’ve long marveled at this fund’s size given its demonstrable lack of merit as a portfolio diversifier. Over the past 10 years the fund has behaved like an overpriced, underperforming 40% stock, 60% cash portfolio. Its R-squared over this period to the U.S. stock market index is 0.85.

Not only is its past performance damning, but little in the substance of the strategy suggests performance will radically change. Gateway owns a basket of stocks designed to track the S&P 500, with a slight dividend tilt. On this portfolio the managers sell calls on the S&P 500, capping the potential upside of the fund in exchange for a premium up front, and simultaneously buy puts, capping the potential downside of the fund at the cost of a premium up front. By implementing this “collar” strategy, the managers protect the portfolio from extreme ups and downs.

There is another way to soften volatility: Own less equities and more cash—which is pretty much what this fund achieves in a roundabout manner.

Portfolio theory says that an investment is only attractive to the extent that it improves the risk-adjusted return of a portfolio. That means three things matter for each asset: expected return, expected volatility, and expected correlation with other assets in the portfolio. The first two are intuitive, but many investors neglect the correlation piece. A low return, high volatility asset can be an excellent investment if it has a low enough correlation with the rest of the portfolio.

Consider an asset that’s expected to return 0% with stock-like volatility and a perfectly negative correlation to the stock market (meaning it moves in the opposite direction of the market without fail). Many investors, looking at the asset’s standalone returns and volatility, would be turned off. Someone fluent in portfolio theory would salivate. Assume the expected excess return of the stock market is 5%. If you own the stock market and the negatively correlated asset in equal measure, the portfolio’s expected excess return halves to 2.5% and its expected volatility drops to 0%. Apply some leverage to double the portfolio’s return and you end up with a 5% expected excess return with no volatility.

In practice, many investors do not assess assets from the portfolio perspective. They fixate on standalone return and volatility. Much of the time this is a harmless simplification. But it can go wrong when assessing alternatives, such as with Gateway. Judged by its Sharpe ratio and other risk-adjusted measures, Gateway looks like a reasonable investment. Judged by its ability to enhance a portfolio’s risk-adjusted return, it falls flat.

I don’t believe individual investors are responsible for Gateway’s size. If anything, institutional investors (particularly RIAs) are to blame. You would think that supposedly sophisticated investors would not fall into this trap. But they do. A large part of the blame belongs to committee-driven investment processes, which dominate institutional money management. When a committee is responsible for a portfolio, they often hire consultants. These consultants in turn promise to help members of the committee avoid getting fired or sued.

In this context, the consultants like to create model portfolios that have predefined allocations to investment types—X% in large growth, Y% in small-cap value, Z% in long-short equity, and so on—and then find suitable managers within those categories. When picking those managers, they tend to focus on return and volatility as well as performance relative to peers. If not done carefully, a fund like Gateway gets chosen, despite its utter lack of diversifying power.