One of the few boneheadedly obvious things financial theory tells us is that an asset is worth more if you can immediately convert it into cash. A CD maturing in a year is worth a discount to face value; an illiquid bond maturing in a decade is worth less; a minority shareholding in a small company is worth even less.

Fortunately for anyone who gets paid to explain this stuff for a living, the terminology is hopelessly confused. As one of my CFA workbooks pointed out, the two terms people use for the extra return investors demand in exchange for their money being locked up are “illiquidity premium” and “liquidity premium.” Nice.

This is especially confusing because there’s a certain category of investors who actually want illiquidity, and inasumch as investor preferences are expressed by the returns they accept, they’ll pay a premium for assets that are harder to sell. This flies in the face of financial theory, but the institutional logic behind it is sound — for the asset allocators, not the people they’re allocating on behalf of.

Achieving Superior Risk-Adjusted Returns by Being Bad at Adjusting for Risk

Private equity is a big business — PE funds raised around $700bn in 2018, and have about $2tr in capital available to invest. Historically, the industry has served a useful purpose: by raising managers’ equity stakes in a company, they incentivize better management, and since higher manager stakes correspond to bigger comp packages, they can generally attract talent. If you look at the early records of big PE shops, you see some incredible numbers — funds putting up 20%, 30%, even 40% annual returns. There’s basically no level of risk that can make such a return a bad deal.

Now, returns are still good, but they’re lower. PE as an asset class outperforms public equities, and has low measured volatility. However, the researchers at AQR have busted out their slide-rules to ask a simple question: how do you measure the risk of an asset that doesn’t trade very often?

Their answer is clever, and it’s worth reading the whole paper, but the gist is: you can construct a portfolio that gets “PE-like” returns by buying the public equities PE companies buy, and using the level of leverage they use. Your returns are similar, but your risk is higher.


Because if you own a private company, you don’t have to mark down its value just because the market dropped. In 2008, the S&P 500 dropped by about 38%. If you were long equities, you lost something like that. If you were levered long equities, you probably lost more. A portfolio that was 2:1 levered long the S&P would have lost three quarters of its value; a little more leverage and you’d be wiped out.

However, if you had an arrangement with your broker whereby your margin loan was based on your cost basis, not the current market price, you might be inclined to hang tight. Your broker would, too: as long as you can service the debt, the broker doesn’t take a writedown on the loan. Everyone wins! (Except whoever is supplying the broker with capital for those loans.)

A private company bought with a lot of debt is analogous to this hear-no-drawdown, see-no-drawdown, speak-no-drawdown arrangement. Economically, it’s a levered bet on equity markets. But accounting doesn’t reflect that; you simply don’t have to write down a private company just because comparable public stocks have dropped.

There’s sensible economic logic here. It’s like Ben Graham’s analogy, in which value investing is like taking advantage of a business partner with untreated manic-depression:

That’s a good description of how an investor should think; keeping your head when all about you are losing theirs is a good way to earn a superior return. But burying your head in the sand when everyone is panicking for justifiable reasons is not a great strategy, and it’s a terrible way to judge returns.

Is “Risk-Adjusted” The Right Approach?

I think so. Warren Buffett disagrees. Volatility is how much a given asset’s price bounces around, while risk is the probability that an investor at a given price will experience a permanent capital loss. Here’s Uncle Warren:

At any given point in time, that’s true: if a stock suddenly drops 50% for no reason, its volatility goes up, but it’s not more risky. It’s less risky! You have a valuation buffer.

But stocks don’t drop 50% overnight for no reason; they drop because somebody who wasn’t willing to sell at $x yesterday is willing to sell at $x*0.5 today. Usually, that’s because there’s news. Usually, the news is bad.

Suppose there’s a population of investors who treat sudden stock price drops as noise, and trade accordingly: when it’s up, they sell; when it’s down, they buy. Over time, if Warren Buffett is right, these traders make money.[1] If that happens, though, it has an effect on stock prices. Specifically, it dampens volatility, because the traders who take the opposite side of big moves have, over time, bigger bankrolls. That can’t go on forever: at some point, those investors will systematically underreact. If there’s news that makes a company worth 10% less overnight, it’s a market inefficiency if the stock drops 20%. But it’s also a market inefficiency if the stock only drops 1%.

The equilibrium point is the point at which fading big moves has equivalent returns to just buying and holding stocks. But for that to be true, the volatility can’t be pure noise; it has to be volatility that represents actual uncertainty. And uncertainty about a company’s financial prospects is just another way to state the probability of permanent capital loss.

Buffett’s intuition is straightforward, and his edge case makes sense — but we deal in aggregates, not edge cases, and in the aggregate it makes sense to treat volatility as a measure of risk.

(If you disagree, please quantify your disagreement in terms of the percentage of your net worth you’ve allocated to mean-reversion strategies. If Buffett is right, these strategies are free money. There’s some volatility there, but of course that’s not the same thing as risk, right?)

The Illiquidity Arbitrage

There’s one class of investors who have the means, motive, and opportunity to take advantage of the gap between the economic volatility of PE assets and the volatility of their reported returns: pension fund managers. State pension funds have:

  1. Multi-decade investment horizons.
  2. Substantial capital, such that they can invest when the minimum check size is seven or eight figures.
  3. An aggregate $1.35tr gap between the present value of their promises and the value of their assets.
Image for post
Image for post
Via Bloomberg, pension funding status by state. Red means load up on risk assets and floor it.

Over the last decade, pensions have shifted away from cash and bonds, and towards riskier assets. Equities are a bigger share of the shift than alternative assets like PE, but PE (and real estate, and hedge funds) are growing at a healthy clip. State Street has details.

One financial theory, which I learned in my CFA textbooks, is that the risk tolerance of a pension fund should rise only if it’s overfunded; an underfunded pension fund can’t afford to take big swings, because if they lose they’ll be even more drastically underfunded. That theory isn’t borne out by reality, though. You’d expect underfunded pensions to be getting very cautious, socking their money away in t-bills and such. Instead, it’s risk-on. This is because the compensation of a pension fund manager is basically a call option on the pension’s funded status; as it gets more underfunded, the option goes further out of the money, so raising volatility becomes preferable to raising absolute return.

Right now, pension funds’ assumed rate of return is about 7.4%, compared to 8.0% in 2002, according to NASRA. Or, to benchmark that to a low-risk long-term investment: in 2002, they were expected to earn about 3% more than 10-year treasuries. Now, they’re expected to earn 4.8% more. However, nobody is explicitly calling for pension funds to take more risk, and nobody (as far as I know) believes that pension fund managers, in the aggregate, are adding 180 basis points more alpha per year than they used to.

So the incremental returns need to come from taking risk that doesn’t look like risk, and that means buying assets that are infrequently marked-to-market. CalPERS walked through the logic here, albeit with more of a focus on the fact that 7% is a really high hurdle with today’s valuations, and that PE is the only asset class that routinely beats it.[2]

In effect, the poor measurement of illiquid assets’ risk is a stealth bailout to long-term investors with sticky capital.

Whither PE?

The private equity model is still a good model. I expect talented people to continue to join the industry, and for sophisticated, returns-seeking investors to keep allocating capital to it. But valuations are a brute fact: you simply can’t get the same returns buying high and selling higher that you can from buying low and selling high.

By getting the right talent, and taking advantage of the favorable optionality of corporate carve-outs, PE can continue to earn above-average returns. If you have a pension (or, more to the point, if you’re part of the tax base that will inevitably bail out pensions — pension recipients being politically better organized than taxpayers), it’s not the relative return you have to worry about. It’s the absolute return. Pension plan hurdle rates are sticky, and keeping the expected return rate high is a source of free money for politicians who won’t be around to face the consequences when the bill comes due.

The worst-case scenario, though, is not that a bunch of pension funds fail during the next market downturn. The risk is that they don’t: if avoiding writedowns pays better than avoiding risk, we’ll see even more pension assets shift towards these risky assets.

In a more realistic scenario, where pension funds try to reduce their PE exposure during a crisis, we could see something really interesting: PE funds have two trillion dollars of dry powder and they’re historically the buyers who step in when times get crazy. In this scenario, one of the distressed asset classes will be LP interests in PE funds. We could see a weird ouroborous where PE has good time-weighted returns because PE funds buy one anothers’ LP interests at a discount from distressed investors, but poor dollar-weighted returns because those same investors are exiting the asset class at the worst possible time.

Ultimately, pensions are a race: can investment managers generate superior returns faster than politicians can generate irresponsible promises? I know who I’m betting on.

Don’t miss the next story. Sign up for my occasional email newsletter. Or check the About/FAQ page for more.

[1] You could object to systematizing the strategy this way: what if the stocks that drop big were overpriced, and the stocks that rise a lot were underpriced, and the new price is generally fairer than the old? That solves half your problem: for stocks that drop, the increase in volatility is coupled with a reduction in risk; for stocks that rise, the increase in volatility is coupled with an increase in risk. So at best, you can argue that volatility might be uncorrelated with risk, but only if markets are efficient during big swings but inefficient during market lulls. And that’s hard to square with the fact that volatility persists over time: big swings are usually followed by smaller, but bigger-than-usual, swings, which imply that markets are unusually inefficient at processing the kinds of information that lead to higher volatility.

[2] This, by the way, is why plans to create PE-like returns from public markets will have a hard time raising money. They’re more accurately measuring what happens to the value of a levered portfolio of troubled companies when equity markets dive. If they achieve the same risk-adjusted returns (hard, since PE can recruit better executives — but doable, because they recruit those execs by charging higher fees) the reported sharpe ratios will be worse. It would be too cynical to say that PE sells long-term investors the right to avoid recognizing losses, but it would be irresponsible not to think that’s part of the appeal. Nobody has to explicitly think this way for it to be true.

I write about technology (more logos than techne) and economics. Newsletter:

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store