With the S&P 500 down approximately 20 percent year to date, virtually every investor is feeling the pain of unrealized losses. Unfortunately for most investors, bonds have not provided the ballast for portfolios that they typically do. Interest rates have risen 300 basis points or so in 2022, leaving core bond portfolios down roughly 20 percent as well.
There has been nowhere to hide. Well, except maybe private markets.
Endowments, foundations, and pension plans with large allocations to private markets are just about the only investors that have been able to report better numbers than the market. In fact, some are even positive on the year.
And yet, valuations associated with these private investments have come under criticism for being unrealistic or unrepresentative of underlying market dynamics, with one famous investor going so far as to label the lack of drawdowns “volatility laundering.”
With registered vehicles that invest directly into private markets up for the year — the Cliffwater Corporate Lending Fund has gained 5 percent, while StepStone Private Markets is up 6 percent — there does seem to be a disconnect.
But what if the private markets actually get it right? What if short-term volatility in public markets is a bug, not a feature?
Equity total returns can only come from two components: dividends and price changes, either up or down. There’s no other way to make money in stocks. The dividends are easy to observe, but price changes are a bit trickier. In fact, it’s almost impossible ex ante to predict what will happen, but ex post, those price changes can be attributed to two things as well: changes in earnings or changes in earnings multiples.
Price = price / earnings x earnings.
Mathematically, this is called a tautology.
Price changes only come from some combination of earnings and multiples going up or down.
Using data on the S&P 500 dating back to 1960 — compiled by NYU Stern School of Business professor Aswath Damodaran — we can quantify some of these effects more precisely in the rearview mirror. From 1960 to 2021, the S&P generated a total return of 10.6 percent per annum. During this time, the average dividend yield was just under 3 percent, and the annualized price change was 7.3 percent per annum. (There is an interaction effect, which means these don’t add up exactly to 10.6 percent, but that can largely be ignored for the sake of discussion here.)
But what drove that price change?
The trailing 12-month earnings multiples for the S&P expanded from 18.73 in 1960 to 23.09 in 2021, rising 0.3 percent per year. Multiple expansion certainly helped the total return of stocks.
However, the bulk of the returns came from earnings growing from $3.10 per share at the beginning of the period to $206.38 at the end of 2021 — an average annual growth rate of 7 percent. Ninety-five percent of the price return over this six-decade period can be attributed to earnings growth.
As Benjamin Graham once said, in the long run, the market is a weighing machine.
However, in the short run, changes in sentiment have far more impact on price returns than actual changes in earnings do. Using this same data on an annual basis, the correlation between the yearly price change and change in the multiple is 0.71; the correlation between price change and earnings change at the one-year time horizon is just 0.05.
In the short run, the market is a voting machine; declining sentiment causes market participants to put a lower multiple on earnings.
Many stock market investors will argue the market is forward looking; this is the result of the market rationally repricing lower expected future earnings by discounting the current multiple.
Unfortunately, the market doesn’t do a very good job of this. Since 1960, in the 28 years that multiples contracted year over year, the next year’s earnings growth averaged — wait for it — exactly 7 percent! And in 15 of those 28 years — 53 percent of the time, to be precise — subsequent earnings growth actually exceeded 7 percent.
The truth is the voting-machine mechanism of public markets isn’t terribly accurate — maybe not even as accurate as a coin toss! Which makes sense, because a copious amount of academic research indicates humans are prone to over- and underreaction (see here, here, and here, for instance).
The frictionless ease of transaction in public markets means it’s extremely easy to over- or underreact. You can panic sell a stock, and if you change your mind, it’s easy to buy it back. I highly doubt this efficiency equals more price accuracy in intra-year intervals, though. This is quite literally System 1 processing — quick, efficient, but often inaccurate.
On the other hand, in private markets, it is precisely the friction of transaction that prevents frivolous pricing. No one panic sells a house, for example. It costs far too much. The time, effort, and legal work that go into a transaction mean market participants don’t buy or sell unless they are truly convinced it’s worth it. Private markets force market participants to engage rational, System 2 thinking.
Put another way, private markets have no voting mechanism; they are only a weighing machine. There is no reason for multiples to oscillate wildly based upon emotional and inaccurate predictions about the future. In fact, most private-equity funds hold multiples fairly constant over the life of the investment.
If EBITDA in a hypothetical private company was $10 million when the company was purchased, and today it’s $15 million, most funds are still applying the acquisition multiple on those earnings. At 8x it was worth $80 million, and now it’s probably worth $120 million. Maybe a little more, maybe a little less, but randomly haircutting the valuation by 20 percent makes no sense.
With capital locked up, there is no need for the fund to access immediate liquidity. The option to be a seller when it’s advantageous accrues to the owner. Hence, there is also no need to “mark to market.” Further, most of the more vociferous critics of private-equity valuations don’t understand how fees and carry work in private-equity funds. Unlike hedge funds, fees are not paid on valuations and carry is not paid on unrealized gains. In private-equity funds, fees are paid on cost and carry is paid only on realizations.
Interim valuations don’t really matter!
Nor are valuations a proxy for terminal value. Good private-equity funds consistently hold businesses well below where they will ultimately sell; bad funds mark them up during fundraising and are often forced to write them down later. Frankly, with a little due diligence, it’s easy to tell the difference (even some academics have figured it out!).
Of course, there are funds that need to mark down because they, like most of the late-stage venture industry, have massively overpaid for bad businesses, and there are even some that have decent businesses that are just underperforming.
Absent a severe economic contraction — and with GDP rising 2.6 percent in the third quarter, that doesn’t appear to be occurring yet —plenty of private-equity funds will have positive returns this year because their companies are growing earnings, and the prospect for a sale at or near the acquisition multiple over the next two to three years is still highly likely.
If earnings do decline, we’ll mark the businesses down. In the meantime, you can keep your voting machine. Our weighing machine is working just fine, thanks.
Christopher M. Schelling is the founder and chief investment officer of 512 Alternatives, a boutique consulting firm dedicated to helping wealth managers, family offices, and small institutions understand and access alternative investments.