My Fund Outperformed. My Return Still Lagged. Come Again?

A few weeks ago, I wrote about a paper that argued active funds hadn’t been as bad as some had claimed. I didn’t find the argument persuasive, but it was thought-provoking. In fact, it made me wonder: How often did the average dollar invested in active funds outgain the indexes to which those funds are compared?

That’s a different question than the one we normally ask: How many active funds beat their benchmark? In answering that question, we typically pay no attention to how much money each fund held or when those assets arrived. Rather, we take a fund’s total return—which assumes an initial lump-sum investment held to the end—and compare it with the index’s, repeating for all other funds and tallying up the number of winners.

But I wanted to take that a step further to consider the timing and magnitude of investors’ purchases and sales, as this would yield an estimate of how their average dollar performed, and then I could compare that with the index’s return. To that end, I compiled data for all share classes of all active US stock funds that existed as of May 31, 2016, and derived their total and estimated dollar-weighted returns over the subsequent 10 years.

Falling Short

There were more than 5,000 active fund share classes that began the 10-year period. About one in five of those funds bit the dust—through liquidation or merger into an unrelated fund type—over the ensuing 10 years. That left around 4,000 surviving funds, only 1,000 or so of which generated higher total returns than their index by decade’s end.

When I estimated the dollar-weighted returns of those outperformers, I found only 430 share classes where the dollar-weighted return topped the index. Meaning that even among these winning funds, the average dollar’s return lagged the index’s more than half the time. All told, only one in 12 funds that started the 10-year period survived to the end and earned both a higher total and dollar-weighted return than the index.

That’s not the whole story. Indeed, you could have a fund that lags on a total-return basis but where the average dollar earns more than the index thanks to deft timing by investors. Given that, here’s a breakdown of funds based on whether their total returns exceeded the index’s or not (horizontal axis) as well as whether their dollar-weighted returns topped the benchmark (vertical axis).

While there were some instances where laggards outperformed on a dollar-weighted basis, they were rare. Including them doesn’t change the picture.

Strike Two

What could change the picture, though, is a scenario where one of the funds that outperformed on a dollar-weighted basis happened to be huge in terms of net assets. Were that the case, it could mean the average dollar in the pool of assets invested across these outperforming funds did quite a bit better than I’m showing above.

Given that, I ran a second analysis, but this time, instead of treating each fund as a stand-alone entity, I aggregated them into a consolidated pool of capital, summing their net assets and flows. Then I estimated that pool’s internal rate of return over the 10-year period and compared it with a blended index that mirrored the composition of the funds in the pool.

As you’d expect, the outperforming funds had a higher aggregate total return than the blended index—they wouldn’t be “outperformers” if that weren’t the case, after all. But I estimate their aggregate dollar-weighted annual return was more than 3 percentage points less than that total return. As a result, the pool’s average dollar lagged the index. (For completeness, I’ve also shown the results for the funds whose total returns lagged the indexes.)

On this score, too, it appears the average dollar invested in winning funds didn’t earn as much as the benchmarks did.

Caveat

You could punch a hole in this argument—I’m comparing these active funds’ dollar-weighted returns to costless indexes’ time-weighted returns, a potential mismatch. What if instead I compared with the estimated dollar-weighted returns of actual passive funds? In that way, it’s apples-to-apples—the return of the average dollar in actives versus passives.

So, I reran the test above, but this time, instead of comparing the outperforming active funds’ with the costless indexes’ total returns, I compared them with the passive funds’ estimated dollar-weighted returns. To control for differences in the distribution of assets by style among active and passive funds, I compared the funds at the Morningstar Category level. Here’s how the numbers came out.

Viewed this way, things look up a bit for the outperforming active funds. The average dollar invested in those funds earned more than the average dollar in passive funds in four of the nine categories. For instance, active large-growth funds’ 20% per year dollar-weighted return topped the passive funds’ 18.9% annual return.

However, in aggregate, the average dollar invested in the passive funds earned a higher annual return (13.7%) than the average dollar in these active funds (11.7%). This is because large blend accounts for a disproportionate share of passive assets, and the dollar-weighted returns of passive large-blend funds exceeded those of active funds.

Final Thoughts

To be clear, when you’re estimating the average dollar’s return, as I have here, you’re really measuring two things: The manager’s performance (that is, the fund’s total return) and the effects of investors’ purchases and sales (which resolves to the fund’s dollar-weighted return). Thus, it’s not a pure litmus test of active management.

Nonetheless, it’s worthwhile to consider investors’ outcomes in dollar terms. Even when things went right and an active fund surpassed its benchmark, there was less than a 50/50 chance that the average dollar outperformed, too. Again, only around one in 12 funds survived and went on to notch higher total and dollar-weighted returns than the index.

Even when I compared with actual passive funds, the results were sobering—the average dollar invested in outperforming active stock funds still lagged the return of the average dollar in passives.

That doesn’t mean it’s hopeless. To an extent, investors can exert control over when they transact and how much they buy and sell at those times. The key—apart from identifying worthy active funds that boast attributes like a strong, repeatable process, a deep, talented roster of portfolio managers, and above all, low fees—is having a long time horizon, trading as infrequently as possible, and avoiding performance-chasing.

The last point is worth emphasizing: It appears that one of the reasons the average dollar in outperforming funds so often returned less than the index was investors’ propensity to chase returns. They tended to buy more (or at least sell less) after a spurt of outperformance and sell more following short-term underperformance.

To illustrate, here’s a time-lapse of the outperforming funds’ aggregate estimated rolling 12-month net flows. I’ve broken them down based on whether the fund had outperformed or underperformed over the preceding one-year period. These outperforming funds were in outflows over this 10-year period amid a shift toward passive investing, but it’s striking to see the difference in flows following a period of positive versus negative excess returns.

This is indicative of performance-chasing, with investors adding money following periods of stronger relative returns and the opposite amid lackluster returns. This chasing dents dollar-weighted returns, likely explaining why the average dollar gained so much less than the funds, and therefore should be avoided.

Switched On

Here are other things I’m reading and listening to:

• Our Model Portfolio landscape report
• Ben Carlson’s excellent new book, Risk & Reward
• Don Phillips on The Long View podcast
• Cuteness alert: “I Knew It, I Knew You”
• “Shenzou”--Gravity soundtrack

Don’t Be a Stranger

I love hearing from you. Have some feedback? An angle for an article? Email me at jeffrey.ptak@morningstar.com. If you’re so inclined, you can also follow me on Twitter/X at @syouth1, and I do some odds-and-ends writing on a Substack called Basis Pointing.