Swedroe: Key Market Value Metric Outdated

Larry Swedroe previews his Inside ETFs talk on why the Shiller CAPE 10 doesn’t measure the market’s value accurately.

Reviewed by: Drew Voros
Edited by: Drew Voros

Larry Swedroe is a principal and director of research for Buckingham, an independent member of the BAM ALLIANCE. He was among the first authors to publish a book explaining passive investing in layman's terms—"The Only Guide to a Winning Investment Strategy You'll Ever Need." His most recent book, "Think, Act and Invest Like Warren Buffett," was released in early 2013. He will be giving a talk at Inside ETFs, “Is the Market Fundamentally Overpriced?” in Florida on Jan. 24.

ETF.com: You’re giving a keynote speech at Inside ETFs in Florida on Jan. 24, in which you will be questioning how markets are valued; specifically, the CAPE 10 metric. Let’s talk about that.

Larry Swedroe: It's important to begin by defining this metric that has been used to state that the market is overvalued, which is the Shiller cyclically adjusted price earnings ratio, or referred to generally as the CAPE 10. It's a metric that many use to determine if the market is fairly valued, undervalued or overvalued.

For the last four years or so, many gurus, including people like Jeremy Grantham of GMO, John Hussman of Hussman Funds, and even Nobel Prize-winner Robert Shiller himself, who this metric is named after, have been sending warning signals this metric has been giving off.

The long-term average—and here, we're talking about an average that goes back to the first data in 1870—is an average of about 16.6, and currently, CAPE 10 is registering about 26.

Clearly, it's way above the historical mean, which leads many people to say that the market is overvalued. It would be fairly valued if it were around the mean; and it would be undervalued if it were below the mean. It's certainly good to look at long-term data, but you have to ask yourself, are there regime changes? Are you really looking at an apples-to-apples comparison?

My talk covers six issues that lead me to conclude that the market, at the very least, is not dramatically overvalued. And it may not even be overvalued at all, just more highly valued, which doesn't mean it's overvalued.

ETF.com: What do you mean by “regime change”?

Swedroe: You could think of a regime change going from a great depression environment to a more normal environment. But here, we're talking about regime changes that I think are very important. You have to remember, the data goes back to 1870, and you need 10 years of data for a data point, thus “CAPE 10.”

Using a 10-year adjusted average of the P/E ratios isn't really something new. It goes back to the 1930s when Graham and Dodd first suggested that a one-year earnings number—and therefore a current P/E ratio­—may be misleading in terms of valuations. One-year earnings are simply too volatile to be indicative of the long-term earnings prospects for a company.

In 1988, John Cochran wrote a paper with Robert Shiller, who went on to win the Nobel Prize in Economics. In it, they presented the evidence showing that a 10-year smooth average of earnings actually has good predictive value in terms of explaining future returns. It explains about 40% of the next 10 years' returns.

This CAPE 10 metric became popularized after that paper came out. So we have a 135-year average going back to 1780. But there have been regime changes that have occurred that may cause the data to be basically irrelevant.

First, for much of those 135 years, there was no Federal Reserve system. We didn't have some government entity that could help reduce the volatility in the economy. And if you reduce economic volatility, you should reduce equity volatility. If you reduce equity volatility, you're reducing the risk of owning stocks, which should drive up price/earnings ratios. You should be willing to pay more for a less risky asset.

Clearly, the Federal Reserve has, since the Great Depression, learned and been able to dampen the level of economic volatility. The level of real growth in the economy is much less volatile than it was 140 years ago and certainly 100 years ago. That would argue for a higher P/E ratio in the data before the establishment of the Federal Reserve.

The second big regime change was the establishment of the SEC, which clearly made for stronger investor protections. Of course, that came after the bubble burst in 1929, which brought much stronger rules, and should have made investors more comfortable investing in equities. Again, that could drive up P/E ratios that people would be willing to pay for the same dollar of earnings, because equities are less risky.

Thirdly, we now have much better accounting rules, with the establishment of the Financial Accounting Standards Board, which made earnings and balance sheets much more transparent than they were 140 years ago. These three key changes in the regimes have made equities a lot less risky.

If you look at the mean CAPE 10 since 1960, it's almost 20. If you look at it from 1970, it's again almost 20. And if you look at it from 1980, a 35-year period, it's over 21.

A more reasonable starting point, because of these regime changes, would be to use a mean of, say, 20, rather than a mean of 16.6. The market at that point would still look highly valued, but about half of the overvaluation, maybe 40% of it, would disappear.

ETF.com: Let's touch on some of the other issues you have related to CAPE 10.
There was a big change in 2001 in accounting rules, which I think were very good. The regulators increased transparency. What happened prior to 2001 was that if a company, say a high-tech company, made an acquisition and, paid, say, $100 million for it, it could write off the goodwill over a long term; I think it was 40 years.

However, in 2001, they put in a rule that said, “Every year we're going to require an impairment test be put on, and an immediate write-down to any impairment.” So if a year or two years later, after writing off, say, $2.5 million a year, you've got an asset on your books, of, say, $95 million that you're still going to have to write off over time. Then you decide the technology business is worthless.

Under the old rules, you would keep that asset on the balance sheet and keep writing it off over time. But under the new rules, you would have to take an immediate $95 million hit to earnings.

Then there were other new rules that applied not only to goodwill, but any intangible asset. That immediate impairment meant that earnings clearly were more accurate reflections of what was going on with the company. But if you're comparing P/E ratios across time, all of a sudden, earnings look a lot more expensive, or the P/E ratios look a lot higher today than under the old rules.

There was a study done that found if you were using the old rules, the P/E ratio would be about four points lower than they actual are. So if we take my starting point, for argument's sake, of 20, as more reflective of a true mean because of the regime changes we want to account for, and we add four, because of the accounting rule, now we would say the correct mean to look at would be 24, and we're at 26 or so. So maybe we're slightly overvalued.

Let’s now look at dividends. About 40 years ago, some two-thirds of all companies were paying dividends. By 1999, it fell to one-fifth. I think the figure is a bit higher today. If we looked at the S&P 500, the payout ratio prior to 1995 was about 52%; it's now down to about one-third.

The economic theory is very simple: Obviously, if you retain more earnings, you should be able to grow your future earnings using those assets. And that theory is supported by the data. Since 1995, earnings have grown much faster than they did before that.

The difference between a 52% payout and a 34% payout—which is the current payout ratio—is worth about one in the CAPE 10 P/E. So now we go from 20, add four for accounting rules, add one for dividends, and we are now up to 25.

Company balance sheets are also much stronger. More cash on a balance sheet (all else equal) and less debt on a balance sheet (all else equal) should make a company less risky. Are today's companies' balance sheets stronger or weaker than they have been on average?

It turns out today's companies are sitting on a bit more cash than they have historically. So that should move P/Es upward. Companies are sitting on a lot less debt as a percentage of their assets, about 15% less in their debt-to-equity ratios. More cash and less debt should make us willing to pay more.

And so, if we were already at 25 on CAPE 10, for argument's sake, stronger corporate balance sheets adds another one, and now we're right at where we are today and making the valuations look on an apples-to-apples basis certainly fair value.

ETF.com: Let's touch on two issues: the improvements in liquidity, and financial innovation.

Swedroe: We know liquidity is a risk factor. People require a higher expected return for a less liquid asset. For example, stocks are clearly less liquid than Treasury bills; higher trading costs because of the lesser liquidity is one reason you see stocks having an equity risk premium.

We have seen dramatic improvements in liquidity. First, we had the elimination of the fixed commission era. Commissions have come way down. Many trades are commission-free. Otherwise, it may cost you $10 or $20 to do a trade, where, previously, you could have paid several percent in commissions.

Bid/offer spreads have also come down dramatically; mainly because of the implementation of the decimalization rules. As trading costs have come down, and commissions have come down, that clearly should lead us to have a bit lower equity risk premium and therefore a bit higher P/E ratios.

Then there’s the financial innovations factor. For example, when I was growing up, there were no small-cap index funds if I wanted to invest in them cheaply. So if I wanted to invest in some small-cap stocks, I would have to buy them individually, pay big commissions, have big bid/offer spreads. Now I can buy Vanguard's mutual fund for Admiral Shares for a 0.08% expense ratio and invest in these illiquid assets at very low cost. That, too, should lead to higher P/E ratios.

There is a slew of very good reasons—we didn't touch on all of them—why valuations, I think, are fairly valued. They’re certainly not as dramatically overvalued as others have warned. But it doesn't mean we can't have a bear market.

The bear market would occur not because stocks were correcting this pricing of overvaluation, as they did in March 2000, but because we had some black swan event, maybe some series of terrorist attacks or whatever it might be, that would cause the equity risk premium to rise.

ETF.com: How important is it for an investor to worry about the valuation of the market?

Swedroe: First you should ignore all of the warnings from gurus about markets being undervalued or overvalued and trying to get you to act. They don't know. Having said that, it doesn't mean you should ignore valuations. You shouldn’t use them to try to time the market. What you need to understand is that high valuations mean expected returns are lower.

Most financial economists expect the U.S. broad market to have an expected real return in the 4-4.5% range. Add on roughly 2% inflation, and you're 6-6.5%. That's a lot lower than the 10% return that we've earned historically.

Today's investors are facing a perfect storm. They're going to live longer, which is good news, of course, but they're going to need a bigger pool of assets just to support their lifestyle for a longer period. The problem is that expected returns on both stocks and bonds are now much lower than they were historically. That means they're going to have to take that into account as they build their plans.

Contact Drew Voros at [email protected].

Drew Voros has nearly 30 years' experience in financial journalism. He was a longtime business editor for the Oakland Tribune and sister papers of the Bay Area News Group, and finance writer for the Hollywood trade publication Variety. Voros' past roles have also included editor-in-chief at etf.com and ETF Report.