For the most part, economists don’t give people advice on how to run their lives. Rather we tend to focus on explaining the behavior of consumers and businesses, usually assuming they are at least somewhat rational. One exception is when there is a “principal-agent problem”, the case where the people you hire (the agents) have interests that differ from you own interest.
Thus economists might advise someone to be a bit skeptical if one’s dentist recommends that you get a new crown. Is it actually needed, or is the dentist merely trying to pad his income?
There is one area where economists are especially likely to give advice–personal investments. The Efficient Market Hypothesis (EMH) suggests that it’s extremely hard for financial advisors to consistently beat the market. Because these professionals must be paid for their services, managed mutual funds tend to do worse, on average, than index funds. Thus almost all economists that I know recommend that average people invest in index funds.
Because of the EMH, the field of economics has its own distinct epistemology. We believe in the wisdom of markets. We believe that the optimal forecast of many economic variables is embedded in the consensus market forecast. AFAIK, other sciences don’t use this approach to ascertain what is true. Thus meteorologists don’t typically assume that a prediction market forecast of global temperatures in the year 2050 represents the optimal forecast, even were such a market to exist.
On the other hand, I do wonder if economists are being consistent in the way they apply concepts such as the principal-agent problem and the EMH. Doesn’t our criticism of managed mutual funds apply equally well to our own profession? Consider the following two approaches to policy:
1. Most economists seem to believe that it makes sense for our profession to do a lot of research on the macroeconomy, and then base our monetary policy on forecasts derived from computer models of the economy.
2. I believe that much of this research is wasteful, and that monetary policy should be guided by market forecasts of the relevant economic variables.
In order to see who’s right, let’s take the same analytical framework that makes economists so critical of the managed mutual fund industry and direct it toward our own field. We immediately see two problems. Just as with the financial industry, it is in the best interest of economists if society spends a lot of money financing research on predicting future macroeconomic outcomes. These are good jobs!
Second, the EMH suggests that the output of these investigations will be inferior to the consensus market forecast, and yet we usually argue that policymakers should rely on our computer models, not the consensus market forecast. Thus we seem to be dismissing the value of the EMH when it comes to our own profession, after using the EMH as a bludgeon to bash the financial services industry.
Of course one could argue that research by individual economists is a valuable input into the market forecast of inflation and GDP, but one could equally well argue that research by individual financial experts is a valuable input into the market pricing of assets.
And even if economic research should be subsidized because information has external benefits, that does not justify using a particular Fed model to set policy, rather than the market forecast.
Economists are also “agents”, and our self-interest is not the same as society’s self-interest. On the other hand, I’m also an economist, so why should you believe me? My self-interest might be to carve out a career as a contrarian.
I would respond as follows. I’m not trying to brainwash you; I’m merely pointing to some implications of ideas that many of you already know, especially those with some background in economics. Back in 1996 (in a defense of free trade), Paul Krugman gave four suggestions to people trying to become public intellectuals. This one struck home:
Adopt the stance of rebel: There is nothing that plays worse in our culture than seeming to be the stodgy defender of old ideas, no matter how true those ideas may be. Luckily, at this point the orthodoxy of the academic economists is very much a minority position among intellectuals in general; one can seem to be a courageous maverick, boldly challenging the powers that be, by reciting the contents of a standard textbook. It has worked for me!
That’s what I did in my new book, which comes out this summer. And that’s what I’m doing here. The principal-agent problem and the EMH are well-established ideas. And it is well known that economists are highly skeptical of managed mutual funds, and often recommend indexed funds. In this post, I’m merely pointing to the implication of applying this sort of analysis to my own profession. Don’t automatically believe what I say—think about whether it makes sense.
After all, my best interest doesn’t coincide with your best interest.
PS. You might argue that asset markets don’t exist for some key macro variables. But that’s no excuse; the Fed can create them.
PPS. Part 2 of my MMT critique is now out. Now there’s a theory that categorically rejects the EMH!!
READER COMMENTS
Thomas Hutcheson
Mar 4 2021 at 2:02pm
But how can the Fed know which instruments to move by how much in order to achieve it’s targets without a model of the economy?
Scott Sumner
Mar 5 2021 at 12:58pm
You need a model to decide what variable to target, but then they should just be guided by the markets on where to set the policy instrument. See my “guardrails” paper.
Jose Pablo
Mar 4 2021 at 8:37pm
You point out (but not elaborate on) a very interesting question:
Are not “the markets” and the “financial advisors” (at least the ones that do manage funds plus the ones whose advice is really followed), the very same thing?
Can “the whole market” be “following itself”?
Where does the “wisdom” of an “index fund” came from?
Scott Sumner
Mar 4 2021 at 10:53pm
You asked:
“Can “the whole market” be “following itself”?”
No, it cannot.
Andrew_FL
Mar 4 2021 at 9:51pm
The idea that central planners could solve the problem of central planning by artificially creating markets that wouldn’t exist, is precisely the argument of Lange and the Market Socialists-Economic Calculation is no problem for Socialism, if they can simply play at markets.
Scott Sumner
Mar 4 2021 at 10:52pm
I agree that central planning is a bad idea. But this post never even discussed central planning. Perhaps you commented after the wrong post by mistake?
Rajat
Mar 4 2021 at 11:49pm
I’m not sure what you mean in saying, “that does not justify using a particular Fed model to set policy”. I’m sure no central bank largely bases its policy on, say, a DSGE model. It might just be one input. Or do you mean something broader, like a standard New Keynesian model incorporating an interest rate, a Phillips Curve, and so on? If you mean that, then you would have to widen your question to why do central banks have so many applied macroeconomic researchers on staff, and have experienced researchers advise their decision-makers? But whereas equity analysts may be motivated by money or fame, it may be hard to motivate applied macroeconomic researchers by either money or even publication if their work is never used by policy decision-makers, except indirectly via market forecasts. Maybe that’s not a good enough reason, and if not, we could strip back the personnel working at our central banks dramatically.
Scott Sumner
Mar 5 2021 at 1:59am
Economists tend to be pretty contemptuous of the view that an investment manager can forecast the future price of Apple stock better than the market, so why do we suppose central bank officials can forecast inflation better than the market? Why not just use the market forecast?
Yes, society benefits from original research, both into stocks and into macro variables. But why the double standard?
Jose Pablo
Mar 5 2021 at 9:32am
The “market” forecast of the price of Apple is no other thing but the “weighted average” of the investment managers forecast of the price of Apple (the “weights” being the amount of money each investment manager manages).
Some “investment manager’s forecast” are going to be “better” (whatever that means) than the market forecast. Just not the same managers all the time.
Scott Sumner
Mar 5 2021 at 12:59pm
Exactly.
Phil H
Mar 5 2021 at 2:33am
Damn, that’s a really good post.
I was trying to explain stock exchanges to my 10 year old the other day. I told the story of selling shares in ocean trading ventures, then explained public disclosure… but perhaps I missed a trick. Given what you’ve said about the EMH there, do you think the biggest benefit of stock exchanges is that they make visible information about predictions of future profitability?
robc
Mar 5 2021 at 8:07am
I think you hit the nail on the head. In theory, the price tells you what the crowd thinks is the PV of the future dividend* flow is going to be. That is pretty powerful info, even if it obviously isn’t exactly right.
*counting things like a buyout as a lump sum dividend payout.
As a gedankenexperiment, I have wondered how much things would be different if each stock had a single auction per day. Different stocks would be staggered throughout the day on the exchange. Would that much information be lost versus constant sales?
Scott Sumner
Mar 5 2021 at 1:00pm
I suppose the biggest benefit is its role in allocating capital. But the information provided is a very valuable part of that process.
Jose Pablo
Mar 5 2021 at 6:35pm
The secondary market does not allocate capital. You can argue that the liquidity provided by the secondary market reduce the cost of raising capital on the “primary” market and IPOs do allocate capital. But IPOs are but a small part of the equity market.
Equity markets does not “tell” the expected value of the NPV of future dividends flow. At least, arguably, for the companies that does not pay any dividend and does not plan to do so (and still command a price).
If you look at shares of company A as “stores of value” (kind of “money”) equity markets are “just” places where you can exchange this “coins” for legal tender US$ notes (and vice versa).
Some of this “coins” provide a yield (in the form of dividend) and others don’t.
The “dollar exchange rate” of these coins does allow some companies to issue new coins and to pocket the procedures but only if/when they issue new shares. And private companies also raise capital.
Not so clear to me what all this information about the exchange rate of these “coins” (and all the related fuss) is socially useful for (apart from allowing fund managers to park Ferraris in their driveways).
Scott Sumner
Mar 8 2021 at 4:33pm
I don’t think that’s a very useful way to think about equity markets.
robc
Mar 9 2021 at 2:57pm
Don’t forget about liquidation value. That establishes a floor. If the price gets significantly below that, someone swoops in for a takeover, hostile or otherwise, and collects a “dividend”. That is why I included eventually buyout value as a dividend in my post above yours.
Also, especially with a cash-rich company, liquidation wouldn’t be necessary…someone could take over just enough shares to change out the Board and have the new board start issuing dividends. So there is really no company without plans to do so.
Todd Ramsey
Mar 5 2021 at 9:10am
A brilliant analogy, to which every academic defender of macro modeling should have to answer.
Scott Sumner
Mar 5 2021 at 1:01pm
Thanks Todd.
Gerardo
Mar 5 2021 at 10:22am
I teach EMH just like you say, I am telling students that given information is so efficiently integrated into asset prices, it is tough to get there first. If everyone bought index funds, the market would cease to be efficient.
I think a very good analogy would be our advice on voting. If you want to vote, great, vote. But it’s silly to think your vote matters. Unless, of course, no one else votes, in which voting would cease to be “irrational” and your vote would be decisive.
Scott Sumner
Mar 7 2021 at 2:24pm
I think voting is quite rational, even if there is just a small change it will affect the outcome. Voting boosts aggregate welfare.
Mark Brady
Mar 7 2021 at 3:28pm
Let’s assume that “[voting] boosts aggregate welfare.” How does that make it rational for an individual to vote?
Klaudia
Mar 5 2021 at 2:19pm
Great article! The principal-agent problem rings a bell beyond economics: “Can people be trusted?” What I want to point out is that it is important to question “intentions”, if individuals want to behave more rationally. I know I am not saying anything original, but I do think this concept is not enough acknowledged, which leads to mass manipulation.
Philo
Mar 7 2021 at 11:38am
Market forecasts for inflation, GDP, etc., one or two years in the future are superior to forecasts based on the usual computer models, in that the former are pretty consistently more accurate—-closer to what the actual values turn out to be. But the main reason for monetary policymakers to rely on the market forecasts is that they are *in themselves* much more important than the truth about what the values will eventually be. It is the market forecast for the future economy—-not the unknown eventual reality–that affects present decisions in the economic as a whole, and so this is what the Fed should use to guide policy: for present practical purposes, it is expectations rather than the actual future that matter.
Scott Sumner
Mar 7 2021 at 2:25pm
Yes, but future NGDP outcomes do affect future labor market conditions.
Federico Castillo
Mar 9 2021 at 8:56am
Without GDP, you get less of all. Economic policy should have one central goal – get productivity growing again.
Michael Rulle
Mar 14 2021 at 1:20pm
I never thought economists predictions were any better than investors predictions. But economics includes “actors” who can change the conditions which they forecast.
For example, when Powell targets AIT, it is based on a model, or a view of the world, which implicitly seems like a conditional forecast or prediction—i.e., if we do X, then we will get Y results. It is not merely a prediction, it is an action which claims cause and effect, conditional upon the model. That seems pretty complex.
When speaking of EMH, we first must define the market. I don’t think in our general discussions of EMH we do a very good job of this. My “default position” is the “market” consists of all publicly traded stocks and bonds, globally. But no one realistically invests in that definition of the market. It is almost impossible, as I am unaware of any fund that attempts to invest in this definition of the market.
So we tend to break the “market” up into smaller units—-which is fine. Once it is defined then we can discuss EMH more easily. There are circumstances where certain types of securities owners and traders have more information than that which is reflected in the prices of securities. JP Morgan and GS, for example, disclose their trading profits. They far outperform the markets . They have legal inside information. Some “hedge funds” (very very few) also have such information.
But investors have no such information. Even 95% of hedge funds. However, there are styles of trading which have very low correlations with long only investing, even when taking serial correlation into account. These “styles” are fairly persistent —-yet few have written about these in the context of EMH in a persuasive manner.
As far as long only is concerned, indexes (which need to be defined before hand) outperform those who try to “beat the market”
Comments are closed.