Scott Sumner  

How the EMH can reduce government spending, taxes, and inequality

I Want to Know How the Transpo... Bio of Paul Krugman...

Long time readers know that I am a huge fan of the Efficient Markets Hypothesis, mostly for pragmatic reasons. (As with all social science theories, I do not believe it is literally true---just approximately true.) Now it appears that the EMH might be even better than anyone had imagined.

In economics, there can be an "efficiency/equity trade-off". I believe that most economists overstate how often this trade-off occurs, but it can certainly arise in some situations. The nice thing about EMH-based policies is that they improve both efficiency and equity. Patrick Sullivan sent me to this WSJ article, which provides a good example:

Steve Edmundson has no co-workers, rarely takes meetings and often eats leftovers at his desk. With that dynamic workday, the investment chief for the Nevada Public Employees' Retirement System is out-earning pension funds that have hundreds on staff.

His daily trading strategy: Do as little as possible, usually nothing.

The Nevada system's stocks and bonds are all in low-cost funds that mimic indexes. Mr. Edmundson may make one change to the portfolio a year.

So far his approach seems successful, just as predicted by the EMH:

From his one-story office building in Carson City, Mr. Edmundson commands funds whose returns over one-year, three-year, five-year and 10-year periods ending June 30 bested the nation's largest public pension, the California Public Employees' Retirement System, or Calpers, and deeply-staffed plans of many other states.

Nevada's $35 billion plan is "dramatically smaller" than California's roughly $300 billion, notes Calpers spokeswoman Megan White. "That said, Nevada demonstrates the benefits of reducing the complexity, risk, and costs in a portfolio."

His success is being copied by others:

Now many public pension funds are embracing Nevada's do-nothing approach as they wrestle with dwindling cash and low interest rates. Calpers is severing ties with roughly half the firms handling its money. New York City this year slashed its hedge-fund commitments.

And it's paying off by reducing government spending in Nevada:

When Mr. Edmundson joined the Nevada plan in 2005 as an analyst, roughly 60% of its stocks were in indexes. He turned it even more passive after becoming chief investment officer in 2012. He fired 10 external managers, and, by 2015, all of its stock and bondholdings were in passively managed funds.

Its outside-management bill is about one-seventh the average public pension's, according to Nevada plan documents and Callan Associates, which tracks retirement-plan expenses.

If Nevada consumed a typical Wall Street diet, it would pay roughly $120 million in annual fees. In 2016, Nevada paid $18 million.

When thinking about that $102,000,000 annual savings, keep in mind that Nevada has less than 1% of the US population. We are talking serious money.

So that's the efficiency part---what about equity? A number of people on the left have pointed out that the growing importance of our finance sector has contributed to increasing inequality in America. As states switch from hedge funds to index funds, the finance sector will earn less money. This will allow states to reduce taxes, and/or boost spending on more worthy projects, such as infrastructure.

The WSJ portrays Steve Edmundson as a modest fellow:

He brings lunch in Tupperware. "Great days," he says, are when his wife makes lunch--a BLT or tuna-fish sandwich. Otherwise, it is leftover fish or salads. "I don't want to spend $10 a day for lunch." . . . He generally doesn't work outside 8 a.m.-to-5 p.m. hours. He commutes in a 2005 Honda Element with over 175,000 miles on it. His 2015 salary was $127,121.75, according to a Nevada Policy Research Institute database.

With no one else on his investment staff, Mr. Edmundson rarely uses his conference table and four extra chairs. He volunteered his office to pension-fund employees who work for accounting or benefit calculations.

Last month, a wall went up dividing the room. "I'm not going to complain about my office," he says. "It was too big."

America has holidays honoring moms, dads, veterans, workers, etc. How about a national holiday honoring the key role that "beta males" like Steve Edmundson have played in "Making America Great in the First Place".

Comments and Sharing

COMMENTS (12 to date)
bill writes:

Great post. Really great closing sentence.

Scott Sumner writes:

Bill, We beta males gotta stick together.

Luke writes:

Honest question that has always confused me. If markets are efficient, then why hasn't competition competed away active managers? Aren't the existence of active managers a market signal that markets aren't always efficient?

bill writes:

Hear, hear!!

Jason writes:

Luke, that's a good question. I would look at the incentive structures of active managers' customers. With pensions and endowments, there are long investment cycles, poor measures of manager skill, and principal-agent problems. Ultimately, someone else, e.g. taxpayers, will pay for poor decisions.

I found it remarkable that Mr. Edmundson fired 10 external managers and dramatically downsized. Most CIOs wouldn't find it in their interest to do so. Efficient? Probably, but it lowers the prestige of the position and eliminates possible scapegoats. I'd guess that Mr. Edmundson did so because of his frugal personality rather than institutional incentives.

Scott Sumner writes:

Luke, Markets can't be efficient if no one actively trades.

Jason, Good point.

BC writes:

There's a public goods and positive externality aspect of EMH, which I haven't seen many people ponder. The EMH essentially identifies active management as a public good. Active managers cannot exclude passive investors like Nevada from gaining the benefits of their securities research because that research gets embedded in securities prices. At the same time, Nevada's free riding on that research is non-rivalrous in that it doesn't lessen the active investors' benefits. (Active investors' benefits may be reduced by other active investors' activities but not by passive investors.) So, active managers' securities research provides about $102M/yr worth of external research benefits to Nevada. Scott's proposal that the Fed set up an NGDP futures market would be another example, where the Fed could non-rivalrously and non-excludably gain the benefits of active NGDP traders' NGDP forecasting research.

Usually, we expect that public goods will be underprovided by private entities. EMH, however, asserts that this public good will not be underprovided, and the empirical evidence seems to support that in many cases (liquid markets like US equities, for example). Yet, it seems to me that there are many more people that accept EMH, at least approximately, than there are people that, given EMH's empirical success, ponder the question of what kinds of other public goods might actually be sufficiently provided privately without government intervention.

Global security might be another well-provided public good. Local police is provided by local government as a public good, but there is no global government to provide global security. When the US provides security for its overseas interests, other countries can non-rivalrously free-ride on that security. We might expect that, as a public good, global security would be underprovided, and maybe it is. However, many people think that the US spends too much on its military, especially abroad.

A final example might be basic research. Although national governments provide research nationally, there is no global government to provide it globally. Why do countries spend money on basic research when they can just wait for other countries to do it and free ride on the results when they are published? In fact, the Efficient Research Hyphothesis suggests that it's pointless to do research because any findings that you might come up with are probably already reflected in the literature.

Might there be other public goods that we really don't need to spend taxpayer dollars on? For example, consider education. There is a public benefit to living in an educated society, but why do we assume that absent government subsidies everyone will opt for education beta, trying to free ride on everyone else's education? Why wouldn't the promise of education alpha, say the college education premium, be sufficient to get people to spend their own private resources on education, creating an educated society in the process?

The Original CC writes:

BC- good post. Some under-appreciated points in there.

Jose Romeu Robazzi writes:

Good active managers, the ones who deserve the name "hedge fund manager" in fact improve portfolios risk/reward ratios because they provide "true alpha". Ex-ante, investors make choices about who would be the "true alpha" in their portfolios, but ex-post, it can be found that that "true alpha" was a cost.
Mr. Edmundson may have been a little lucky here, he just is being tested on a period when beta risk provied good returns, and alpha risk, which he avoided, was just a cost.

But, there will be times when the opposite will happen. Don't get me wrong, I believe long term portfolios such as pension funds should have lots of beta risk, but, by not having any true alpha risk, there will be windows os time when they will show very poor performances ... Particularly when the Fed screw up...

Scott Sumner writes:

Jose, Studies show that in the long run, indexed funds outperform managed portfolios. And public pensions invest for the long run.

Jose Romeu Robazzi writes:

Yes, prof. Sumner, a pension fund manager does not care for medium term returns, but individual accounts do.

Also, risk is not very well understood, actually, there is no consensus at all about what is "risk", therefore, in order to evaluate returns, risk adjusted returns, on must have a proprietary measure of risk. Diversified portfolios may use standard deviation based measures, but, even thos may prove that long term performance evaluation is tricky.

A lot of these Studies don't have common risk measures. That is how everyone in the industry tricks the system, by choosing a measure of risk that favors their own portfolios or don't choose any at all and focus on returns only.

But just to mention something that is used widely, I know people who produce portfolios that show sharpe ratios above 1.2 for decades, that is sure much better than any passive portfolio , and those people have true alpha by any method of evaluation you may choose ...

Rick Bohan writes:

"I found it remarkable that Mr. Edmundson fired 10 external managers and dramatically downsized. Most CIOs wouldn't find it in their interest to do so. Efficient? Probably, but it lowers the prestige of the position and eliminates possible scapegoats. I'd guess that Mr. Edmundson did so because of his frugal personality rather than institutional incentives."

And therein lies the rub. The hypothesis of "efficient markets" works just long as the number of smart "managers" outnumbers the number of total goof balls. But the fact that we write blog posts highlighting one smart "manager" would seem to indicate that's not the case.

Comments for this entry have been closed
Return to top