Active vs. passive: how wise is a crowd? and is it really a zero-sum game anyway?

One of the arguments put forward in favor of passive management is that active management is a zero–sum game, i.e. that active investors in aggregate earn the same average return (before fees) as passive. So you may as well save time and money by relying on the wisdom of crowds, and just mimic the market portfolio. In this post, I discuss the logic behind that argument.

The wisdom of crowds

Let’s start by considering how wise the crowd of investors really is. Simply judging by how difficult it is to consistently outperform market benchmarks, this crowd does seem to be pretty wise. Some active investors do manage to win consistently, but nobody claims it’s easy. On the other hand, while most markets are pretty efficient most of the time, no market is completely efficient. All markets are subject to occasional bouts of the madness of crowds; economic and market bubbles (and subsequent crashes) being among the obvious examples.

Hence, investment markets never reach the stable equilibrium of classical economics, in which every security is fairly priced based on all available information. So the difficulty of active management does not stem from there being no mispricing to exploit. The zero–sum nature of the game does not mean there are no ways to win.

How close to efficient any given market is depends on many factors: market concentration, breadth of broker coverage, liquidity, the complexity of the market and so on. Part of what makes a crowd wise is the diversity and independence in the approaches taken and, in that light, the composition of the active manager community is a topic I discussed in a post last month.

So the strength of the wisdom of crowds argument varies. It varies across different markets and it varies across time.

And is it really a zero–sum game anyway?

Now let’s turn to the other part of the argument: that active management is a zero–sum game. Again, this argument is often largely true, but never completely so.

Here are a few of the limitations of the zero–sum claim:

1. Investors don’t all have the same goals

Few investment decisions are based purely on return considerations; most also involve some risk analysis and/or other factors. Occasionally, return considerations may not be a factor at all (for example, central bank action in currency markets is often driven by policy goals, not return expectations.) A previous post provides seven examples of how investor differences lead to different choices.

2. Some investments are more similar than others

Not only does it matter how similar/dissimilar investors are, it also matters how similar/dissimilar investment choices are. For example, consider the following exchange between an audience member at a Morningstar ETF Conference1 and Eugene Fama, who has argued for the efficiency of markets as strongly as anybody:

Audience member: “How do you even define the market portfolio?”

Fama: “It’s just the cap–weighted portfolio of all the traded stocks.”

Audience member: “But that’s just stocks.”

Fama: “I don’t fool around with bonds.”

So bonds and stocks are very distinct assets, offering significantly different expected returns and risks. Different types of investor will favor one or the other depending mainly on their own circumstances. Even Eugene Fama doesn’t rely on the wisdom of crowds to help with that decision.

What’s true of the stock/bond decision is also true within the broad bond market. Long–dated bonds, short–dated bonds, high quality corporate, Treasuries, high yield, overseas debt: these are all quite different from one another and serve different purposes. So the allocation across distinct types of fixed income investment should generally be driven by investor characteristics, not by market capitalization.

In contrast, if investment choices are less distinct (perhaps two bonds with similar yields and maturities issued by similar corporations, or two large cap U.S. equities in the same sector) then it seems less likely that an investor’s characteristics should drive the decision. This is why passive investing is more common within an asset class than it is across asset classes.

3. Passive investment requires a clearly–defined opportunity set

Further, we need a clearly–defined opportunity set or benchmark index against which to manage passively. A portfolio of non–Treasury bonds, for example, is so dependent on how you define the opportunity set, that it’s not really the wisdom of crowds that you need to worry about here but the wisdom of credit rating agencies.

And index weights should reflect the holdings of an appropriate group of investors in aggregate. While this is clearly not a problem for traditional listed markets, there is no straightforward measure of market capitalization for commodities, for example.

So the strength of the argument that active management is a zero–sum game also varies depending on the investor and the asset class in question.

The question is not going away

In short: even though the primary argument for a passive approach (that it is cheap) is a simple and strong one, other arguments—such as the zero–sum game argument—are more complex than they at first appear.

Quoted in “Morningstar Conversation: Eugene Fama”, Morningstar Magazine, December/January 2015.