Mutual Fund Managers Get Paid For Picking Good Stocks Only When It's Worth It

Publish date:

Mutual funds are kind of weird in that they basically aren't allowed to get paid for performance, so they charge investors a flat percentage of assets under management, so for mutual fund managers looking dapper on CNBC is often more profitable than sitting in your office researching stocks. Still - and perhaps perversely - performance does seem to attract assets, so if you run a mutual fund company there is quite a bit of value in trying to get your portfolio managers to pick the right stocks. There are various ways to do that: you could, for instance, ask them politely to do a good job, or yell at them when they don't, or build them a treehouse. But, who are we kidding, basically there's money: if you give them more money for picking good stocks, and less money for picking bad stocks, then you will probably attract good stock-pickers and encourage them to pick good stocks.

I enjoyed this paper immoderately:

Using a unique hand-collected dataset on over 4,000 mutual funds, we find that about three-quarters of portfolio managers receive explicit performance-based incentives from the investment advisors. Our cross-sectional investigation suggests that portfolio manager compensation structures are broadly consistent with an optimal contracting equilibrium. In particular, explicit performance-based incentives are more prevalent in scenarios where this incentive mechanism is more valuable or alternative incentive mechanisms, such as labor market discipline, are less effective. Specifically, our results show that explicit performance-based incentives are more common when (i) the investment advisors are larger or have more complex business models, (ii) the fund returns are less volatile, (iii) the portfolio managers are not the stakeholders of the advisors, (iv) the funds are managed by a team rather than an individual, and (v) the funds are not outsourced to an external sub-advisory firm.

It's from June, and it's by Linlin Ma, Yuehua Tang and Juan-Pedro Gomez and, oh is it pleasing. Basically they formulate a bunch of hypotheses along the lines of "if I were an Econ 101 professor deciding whether or not to pay mutual fund portfolio managers based on their stock-picking acumen, what would I consider?," and then test those hypotheses empirically and find out that they all hold up.

If you think of a mutual fund company as a machine for turning stock-picking performance (good, bad or otherwise) into fee-paying assets under management, then the further a portfolio manager is from the engine driving that machine, then the harder it is for her to see the value from her efforts and the more you have to pay based on those efforts. A portfolio manager who runs a one-woman shop and keeps every penny of the 1%-of-AUM will have every incentive to pick stocks - not as well as possible, but as well as necessary to maximize AUM. One of six co-managers of one of 1,000 funds in a giant fund family can do much less by stock-picking to impact the company's earnings - the Vanguard or Fidelity name will attract assets however one fund performs - so she might as well just play Brickbreaker all day and collect her paycheck. But if you pay her explicitly based on her fund's performance, then she'll try to make it perform well - and if she makes it perform well, then the higher-ups can get more mileage out of plugging it on CNBC.

And so, in fact, "a one-standard deviation increase in the advisor assets under management is associated with an 11.2 percentage point increase in the probability of explicit performance-based incentives," while being a control person of your own mutual fund management company reduces the probability of getting paid for performance by 28.1%. There's more in the same vein, but this is my favorite part:

Performance-based incentive, however, comes at a cost as it can distort the efficiency of the risk sharing between the principal and the agent and imposes risk on the agent. Theory predicts a negative relation between risk and incentives. Given the trade-off between risk and incentives, we should observe less performance-based incentives in the compensation contract when the fund returns are more volatile. We formalize our second hypothesis as follows:
Hypothesis H2: The probability of observing performance-based incentive contracts decreases with the volatility of the fund performance.

Because you can't really pay people as a linear function of their peformance: if they make a lot of money you can pay them a lot of money, and if they make a little money you can pay them a little money, but if they lose a lot of money you can't, for the most part, make them give you back a lot of money. So "performance-based pay" really means giving people an option on their performance - an asymmetric instrument with more upside than downside - and that option's value increases with volatility.

Based on that, what would your hypothesis be about incentive pay versus fund volatility? Mine would have been: "fund managers who get explicit performance pay have more reason to create volatility, and so their funds should exhibit higher volatility." The authors guessed the other way: that fund companies are less likely to give managers an option on performance when that performance is likely to be volatile. When you manage a blue-chip income fund, your boss wants to pay you extra if you can eke out a few extra basis points. When you manage a micro-cap tech growth fund, there's no reason to encourage you to take even more risk.

That's a more optimistic story than my guess: they think that principals are better at optimizing incentives than agents are at gaming them. And it turns out that their guess is apparently better than mine: "a one-standard deviation increase in [trailing 1-year fund volatility] is associated with decreases in the probability of performance-based incentives by 4.1%."

That's kind of neat. The whole incentive-pay-increases-risk effect is well known, and I'm always interested when financial services companies do complicated things to try to reduce it. It's nice to see a whole industry that, at the margin, seems to be paying for performance without creating excess risk. Of course, it's the mutual fund industry, a business that after all is built around restrictions on performance-based pay. If you're looking for excessive risk-takers, there are plenty of other places to find them.

Portfolio Manager Compensation in the U.S. Mutual Fund Industry [Georgia State]