Rethinking How We Evaluate Fund Manager Performance

Includes: DIA, IWM, QQQ, SPY
by: Marc Gerstein

In a March 9, 2009 study, Eugene Fama and Kenneth French concluded that mutual fund managers, as a whole, stink (actually that’s my language, not theirs, but it does sum up their ultimate message). Although this particular study method had some new wrinkles, the grand conclusion is the same as what I’ve been hearing since I was a grad student in the late 1970s, and even back then, the topic was no longer new.

But consider the real world: Even though researchers have been blaring for decades about how fund managers do not add enough performance to justify the cost of active money management, the mutual fund industry continues to manage lots of money, and to the extent there have been shifts lately away from equity funds, much has involved tactical asset allocation (a recent switch from stocks to other assets perceived to be safer) rather than abandonment of active management, and in other cases, money has gone to other forms of active management (e.g., private advisors, hedge funds). ETFs present interesting issues to be discussed below, but for the moment, let’s continue to consider active managers.

Are investors really that dumb? Is it possible that so many continue to knowingly pay for advice or money management that is worthless at best and possibly damaging, and to do so over such a prolonged period? Maybe. Or maybe not. Is it possible that active management really is delivering something worthwhile. Could it be that deficiencies lie in the traditional academic approaches to measurement?

Frankly, I think it’s time for some fresh thinking, not just in terms of modeling, but even in terms of what skill sets are needed to tackle the job. Obviously, quantitative knowledge is necessary. But I think the finance community has been underestimating the human element (I’m talking about real interaction with real people, not utility functions) and the benefits that could result from looking a bit toward how colleagues in the sociology department conduct research.

Here are some thoughts.

Stock selection versus market timing

Life is good for the S&P 500. It’s just a mathematical construct. It doesn’t have to actually buy and sell, nor does it have to deal with stakeholders, clients or shareholders. (Let’s skip the transaction-cost issue. It’s been done to death. But the other consideration, stakeholders; that’s huge.)

Imagine the perfect fund manager; the genius who always knows exactly what to do in the market and when to do it and beats the daylights out of any imaginable benchmark when investing for himself.

Now, imagine that conditions are bearish, stock prices are depressed, the mob is running for cover, and the smart money is scooping up bargains, as our perfect manager is doing with his personal account. So we presume he will also be buying low for the mutual fund he manages. Wrong! He’s probably selling, perhaps aggressively. That’s because he needs to meet stakeholder redemptions and keep replenishing a cushion for future withdrawals.

Let’s now switch gears. Imagine we’re in an overheating market. Stock prices are at ridiculous levels. The dumb money buys, buys and buys some more. But those who know what they’re doing want to sell, to take advantage of Mr. Market’s unusual generosity. Unfortunately, however, our perfect fund manager, while he may be selling in his personal account, is buying like crazy for the fund he manages. That’s because he’s experiencing massive inflows and he needs to invest.

Must he really buy high? Isn’t he allowed to use his judgment to hold cash (or even sell) pending more favorable prices down the road? Based strictly on the prospectus, the answer is probably “yes.” But good luck to any fund manager who tries to act on that. I’ve been there – I managed a junk bond fund back in the 1980s Drexel era – and I did that very thing.

Shortly before junk bonds imploded, that market got very hot and spreads relative to better-quality offerings dropped to incredibly low levels, meaning investors were being paid next to nothing for the increasing credit risks being assumed. So I started declining the new IPOs and shifted the portfolio to about 90% treasuries and investment-grade corporate bonds, as my prospectus allowed me to do under appropriate market conditions. The result: brutal grilling sessions by the general counsel and the fund directors and a ton of hostility from shareholders (“If I want treasuries, I’ll go into a treasury fund!”). I survived in my job, probably because the crash came quickly and my relative performance rankings shot upward. Frankly, though, had the old junk market held out for another six months to a year, I’m not sure I would have been able to hang on.

Having gone through that, I completely understand and support any fund manager who buys high because he needs to be fully invested. They’re not completely oblivious. They may try to go to the higher end of the range of cash they feel they can hold, and they may do their best to rotate from the most overpriced stocks to less-overpriced alternatives, But that’s not the same as really acting full-out on one’s convictions, which might dictate a far more conservative strategy than is consistent with what is expected for the fund they manage.

Legally, the manager might win out, especially if the passage of time proves his conservatism to be well founded. If the market takes too long to drop, who knows. But in the court of public opinion, conviction is highly likely no matter what as stakeholders switch to other funds that invest the way they wish. So ultimately, if the manager isn’t axed by the lawyers, he’s likely to be nailed by the marketing folks. As to anyone who thinks managers should be above that sort of thing, well, all I can say is: Try it. (Perhaps this is why so many highly talented managers gravitated toward hedge funds, where they have greater freedom to formulate and implement strategies.)

In the real world, given the way active managers are forced to buy high and sell low, even when they’d rather not be doing that, I suggest that any manager who can come even moderately close to the S&P 500 is showing a heck of a lot of skill.

This, of course, is my personal opinion. Maybe it’s too heavily colored by my own experiences and wouldn’t hold up under a proper empirical study. That would be fine if it’s so. Hence I call on Fama, French, and all the rest to conduct new research into fund management wherein they control for stakeholder market timing and evaluate fund managers based only on those decisions the managers can actually control.

What’s the benchmark?

The Fama French study established a benchmark based on a four-factor approach that looks an extension of the capital asset pricing model (they use the market, the added size and value-growth factors for which they gained fame back in 1993, and also a momentum factor). Their market component is a value-weighted portfolio of all NYSE, AMEX and NASDAQ stocks.

That sort of thing is very interesting to finance researchers. As for myself, I have no idea what I’m supposed to do with it. If I were to read the paper, conclude that Fama and French are right on, that mutual fund managers can’t, on balance, outperform their benchmark, what am I to do after redeeming my funds and logging into my Ameritrade account to passively invest in that benchmark, which as it turns out, doesn’t actually exist as an investable alternative to mutual funds.

I’m hoping Fama and French will indulge me in my suggestion of the SPDR, the S&P 500 ETF, as a practical alternative for an investor who likes their conclusion and wishes to act on it. So for the rest of this discussion, I’ll assume the S&P 500 as the main benchmark.

Actually, I’ve hit a logical stumble here. Why should I presume to use the S&P 500? Why not the Russell 2000, or some other index?

When investors and commentators talk abut benchmarking equity performance, most jump right to the S&P 500 and don’t even bother explaining why. It’s a given. Seriously, had I continued on without raising this question, how many readers would have jumped on this? As it turns out, it’s usually those who propose other benchmarks who wind up having to explain.

I think it’s time to put an end to this situation.

It’s an important topic when it comes to evaluating active fund management. The academicians who’ve been making careers out of bashing fund managers presume that there’s a readily available alternative: buy the market, or invest in an index. I respectfully disagree.

If you want to index (which seems so easy nowadays given the advent of ETFs), you still have to decide which index, and frankly, I think that’s a pretty important, and active, decision that needs to be evaluated just the way one would evaluate stock picking. If someone says, don’t pick stocks, buy the S&P 500, my answer is: “Why the S&P 500? Why not the Russell 2000, or any of the other countless indexes that are just as readily available?”

Many might be tempted to brush aside much of what’s now available in the ETF-indexation world as being a bit too exotic or specialized. There’s much to debate here (I’m in the camp that embraces the new offerings). But I should think that by now, at least the Russell 1000, the Russell 2000 and the Wilshire 5000 have enough stature to make legitimate competitive claims versus the S&P 500 for benchmarking legitimacy. That alone makes the indexing choice quite active since there can be significant performance differences among just these three. And in this era of globalization, we probably should be expected to also justify the choice of a U.S. index as opposed to at least some of the major MSCI global offerings.

Bottom line: Passive investing is a fantasy. It doesn’t really exist.

You can choose among stocks. You can choose among indexes. Either way, there are choices that have to be made. And either way, the decision makers can be evaluated and held accountable for having made good or bad choices.

Assuming one wishes to ignore the global issue and look only at the U.S. market, I suppose one might construct some sort of argument to the effect that the S&P 500 is a broad representation that includes a huge percent of the total market capitalization of U.S. equities, especially, when adjusted for liquidity, hence that index is a reasonable proxy for the entire U.S. market. Therefore, according to this line of reasoning, the S&P 500 is tantamount to buying the market as a whole and is, hence, a truly passive decision against which active choices can be benchmarked. (That’s probably what motivated me to blurt out S&P 500 when I sought an alternative to the Fama French non-investable benchmark.

I’m sure there are plenty of smart people working at Russell or Wilshire who’d have something to say about that. But even without such support, I still can’t really accept the S&P 500 as the ultimate benchmark. And if someone were to create an ETF to track the more universal Fama French benchmark, I’d reject that, too as a standard for manager evaluation since we’d just wind up back at square one: why this thing, as opposed to the S&P 500, etc.

Let’s not get so wrapped up in theory that we forget why we even care whether active managers add value. We care because if we decide active managers are no good, we need to know what alternatives we should pursue. In this regard, I could care less about the ivory tower notions regarding the entire market (or as a former boss once put it, “God’s investment portfolio”). I just want to know how I can get better value if I’m told I shouldn’t be patronizing active managers. And in the modern ETF era, where there’s such variety of indexation choices, failure to address the if-an-index-then-which-index question would, in my opinion, render the research just as valueless as, active management is alleged to be.

Are you getting a headache from all this? I am. Let’s move on.

Re-designing the standard for evaluation

Let’s wipe the benchmarking slate completely clean and go back to the core question: Do active fund manager add value?

My initial hypothesis is “yes” because I find it hard to start with a presumption that so many people paying for so much active management for so many years can be so thoroughly duped. If proper evidence shows them to be suckers, then I’ll believe it. But as an initial hypothesis, I’ll presume they’re at least somewhat rational.

So how, then, are they measuring the value of active management? Clearly it’s not the S&P 500 or any other index because if it were, almost everybody would be in SPDRs or some other broad-based ETF. So there must be some other standard out there waiting to be articulated.

I propose we consider the idea that the real-world value of fund management is being measured by the difference between how the fund performs versus how the stakeholder (client, shareholder, whatever) is likely to perform if he invests on his own.

This can be a bit sloppy. People who read Seeking Alpha generally have strong understanding of the investment scene and probably could, on their own, do better than less sophisticated investors who might buy and sell based on random tips come their way. Hence a manager who seeks to attract clients from among Seeking Alpha readers would have to work harder and deliver more than would a the anger who serves only the mass market.

From an academic standpoint, where things are expressed as simple equations, this seems to be a nightmare scenario. We’d be looking at a different benchmark for each stakeholder. How, then could we ever clearly and objectively evaluate fund managers?

We might simply look at how our free market system answers the question. We presume that those who are more effective in attracting assets are delivering better value. I know this sounds weird. But reflect a moment: Many who are reading this believe in free markets for business, and feel outrage at any suggestion that someone somewhere knows better than the market. If we’re willing to trust market forces to price and allocate such things as oil, homes, banking services, automobiles, grain, etc., then why should we not trust market forces to efficiently manage the way investment management services are purchased, and presume this is happening?

Speaking for myself, I tend to wonder about market frictions. While I respect the above logic, I’d prefer to go further.

Another approach would require some new research to quantify how investors perform when they go it alone. This would be no small project, but I think it would be do-able and, perhaps, a more worthy use of academic research resources than a gazillion more studies rehashing the same fund-managers-stink projects we’ve been seeing for decades.

Executing this might involve a two-prong approach.

One path might borrow from the field of sociology: conduct ongoing surveys of various population samples regarding their investment preferences, habits and practices and use this as raw material from which we might create a series of investment do-it-yourself-indexes. We’d need different indexes to cover different samples group with different levels of income, investment knowledge, and so forth.

The other path might involve data gleaned directly from retail brokerage records (from firms that do not offer one-on-one investment advice). It would directly measure the results actually being achieved by those who really are going it alone. This obviously involves legal hurdles. Perhaps the records can be made suitably anonymous. Or, perhaps, we can again borrow a page from the sociology playbook: identify workable samples and seek permission from individuals to have their accounts confidentially tracked.

Measuring fund-manager performance (both risk and return) against do-it-yourself indexes would provide a true, useful, measure of the value they are really adding (above and beyond the costs of asset management). My guess is that most established professionals would outperform such indexes. The really interesting insights would most likely come about if we rank managers on the basis of alpha computed against benchmarks like these.

Such an approach would be consistent with how people behave. If I have a health problem, I pay for a doctor because I believe I’d get a better result than if I treated myself. If I’m not confident in my ability to choose investments, I’ll pay for advisory services to help me choose among stocks and/or indexes-ETFs. Whether or not I’m getting my money’s worth depends on whether the professional is really delivering an outcome better than what I could achieve on my own. That’s what we should be measuring.

Odds and ends

By the way, remember the initial section wherein I cited the difficulties comparing a manager to an index which isn’t forced to buy high and sell low. This concern would probably vanish if we were to use a do-it-yourself index, since the latter would likely reflect any high-priced buying and/or low-priced selling that actually occurs.

One more thing: What about managers who serve the most sophisticated investors? There, too, do-it-yourself indexes might be constructed. But here, I suspect we’d wind up with something that looks a heck of a lot like the S&P 500. But if we get there because we’re considering a segment of the investment community that chooses to invest this way, that’s fine. This is a heck of a lot different than saying we should measure all managers against the S&P 500 simply because we feel like it. Managers who serve such clients like this should be expected to beat this benchmark, assuming it truly represents what they’d do if they eschewed professional management. But even here, a do-it-yourself cousin to the S&P 500 might be preferable, just in case even upscale clients have buy-high-sell-low habits that need to be taken into account.

The future of manager evaluation

I hope someone somewhere in the academic community decides to undertake a project like this. And I really do think it needs to be a joint venture between the finance and sociology departments. I doubt either group could be effective going it alone.

I don’t know that the proposal offered here is the exact answer. But I do know that the answers being produced by business-as-usual financial research are stale.

It’s time to stop wagging fingers at those who continue to pay for active management and to instead, come up with some fresh studies that try to figure out what’s really going on out there.