Fellow asset allocator GMO is considering cases where mean reversion could happen over seven years, 15 years, and 25 years. GMO’s Jeremy Grantham recently said higher valuations are here to stay. To him, it seems that lower interest rates, coupled with a surge in corporate profits and profit margins, may be the primary driver of persistently higher price-to-earnings ratios, or an extension in the time frame for reversion to the mean.
We have read Grantham’s pieces. Many of the points he raises fall under what we’d consider a wide range of plausible outcomes. Yes, corporate profit margins could on average remain on the high side. Yes, interest rates may inch up relatively slowly so the mean reversion in P/E multiples may happen slowly. There are some elements we have already been incorporating into our U.S. equity analysis for some time. That said, on some aspects we arrive at a different conclusion. Here is a summary of a few key points and observations:
Our base case scenario currently estimates U.S. stocks will return about 2% in nominal terms five years out. (Note, our analysis horizon is different from GMO’s, so it’s not exactly an apples-to-apples comparison but close.) For some time now, our base case returns have been higher than what Grantham calls “traditional GMO Mean Reversion” returns, after converting their real return estimates to nominal. First, the implied S&P 500 margin assumption in our base case was and still is higher than historical averages, for reasons Grantham highlights in his piece. Second, in late 2013, seeing that deleveraging was progressing in a benign fashion, we reverted back to a valuation framework that emphasized the post-1985 P/E multiples environment rather than incorporating valuations all the way back to the 1930s, which we temporarily resorted to after the financial crisis.
In focusing on the post-1985 environment, we wanted to capture the effect of declining interest rates and the so-called Fed put, among other things. However, we decided to exclude the bubbly periods of 1999–2000 as well as the most pessimistic 2008–2009 period, when reported GAAP earnings cratered, leading to nonsensical P/E multiples. This gave us an average P/E multiple of 18x to 19x for the period. In our base case scenario we apply 17x to our estimate of normalized earnings, while in the optimistic case we utilize 20x to above-normal earnings. Our bearish scenario does take into account the margin environment that existed pre-1995, so here we take roughly post-1985 average margins and average nominal sales growth (the latter is still more than what we have achieved since the financial crisis so one can argue it’s not the most bearish). These three scenarios give us a reasonable range of potential outcomes that we use to manage our portfolios. We don’t believe we need to have perfect insight on how high P/Es or margins might get to. In our minds, the story for U.S. stocks looks much less alluring when we look at fundamentals other than net margins, combined with what we think will be a rising interest rate regime (whether slow or fast we can’t be sure, but it would mean a repricing of risk assets): just as stocks were bid up when rates were declining the past 30 years, they’d be bid down as rates rise.
But if margins have been outstanding, why have profitability metrics such as return on capital been steadily declining for U.S. companies? After all, in the long run that’s what should matter to shareholders—how efficiently are companies using their capital.
When we dig further, the key conclusion we come away with is that there’s been an excessive accumulation of capital that is not contributing adequately to corporate profitability. The question is, what are the companies doing with their capital and how come margins show they are doing a good job?
A good portion of margin improvement in this cycle, as Grantham also notes, has stemmed from low debt-servicing costs. Those are probably reversing, at what pace we don’t know. The second important reason he cites is leverage. We agree that’s been increasing (see the charts to the right), and we think it will be a problem at some point.
Most investors dismiss this leverage, asserting that debt-service costs are low or manageable and we have nothing to worry about for the next couple of years as companies have extended maturities and will not need to refinance. It makes us wonder: if almost everyone thinks that, and everyone thinks they will be the first to get out before companies are forced to contend with higher borrowing costs, then is it any different from other musical-chair games that have ended badly?
A more fundamental reason why returns on capital have been low and why leverage is a problem is companies have been using debt in value-destructive ways, such as buying back stock (which also gooses up net margins at the individual company level if net share count is reduced) and pursuing M&A.
Part of this behavior can be explained by how company managements are incentivized. And part can be attributed to the QE effect.
The median P/E for the S&P 500 Index has been rich for quite some time and currently stands at about 24x according to Ned Davis Research. When companies buy back stock at these rich valuations, they are (over a full cycle) destroying capital, which becomes most evident during the inevitable slowdown or recession. As one equity manager we respect points out, the implicit return on their capital of recent M&A transactions at these high multiples is in the mid-single digits, approximately. This return may make sense now when cost of capital is low thanks to QE. But as QE unwinds and rates rise, the cost of capital for these companies will increase, and in hindsight, these actions will prove to be value destructive. New management will come in, they will take corrective actions, such as spinning off what the previous management acquired, and we will see profitability measures go up again. But as this game or cycle is repeated, it is the equity shareholder that suffers the losses, which again gets realized or becomes apparent during bear markets or recessions, not when markets are reaching new highs. In a recent study, UBS analyzed 150 completed deals over the past five years. Factoring in deal premiums paid and returns or profitability generated following the deal announcement, they concluded only one-third of the deals generated returns that cleared the cost of capital requirements and hence could be considered value accretive.
Bottom line, the way we see it, since the financial crisis, U.S. companies in aggregate are generating poorer profitability and are more leveraged (i.e., more exposed to economic cycles). Grantham suggests maybe we should be more willing to pay higher multiples for earnings than before because that’s what seems to be the trend established in the past 20 years. For anyone who cares about capital preservation that would be a hard argument to fathom given the fundamentals. It’s also worth reiterating that for an asset that can decline 20% (easily) and much more (as we have seen in the past two cycles), low to mid-single-digit returns may be good enough to beat bonds, but they do not absolutely compensate for the downside risk. For someone who cares more about relative returns, they may choose to stay fully invested now. We at Litman Gregory tend to think more in absolute risk-and-return terms than relative terms.
—Rajat Jain, CFA
Take the complexity out of navigating the investment landscape for your clients. Sign up below to receive free Litman Gregory research.
This Website is specifically intended for and limited to financial planners, investment advisors, accountants, registered representatives of licensed securities firms, and other financial and investment professionals with a sophisticated and advanced level of knowledge about investments.Yes, I am an investment professional No, I am not an investment professional