In a recent Yahoo Finance article (h/t Nick Webb), Richard Bernstein attempts the latest rationalization of rising P/E multiples.
“There is an old investment rule-of-thumb called the Rule of 20 that uses combinations of headline inflation and the S&P 500 P/E to determine fair value,” Bernstein said. “Our valuation models are, of course, more elaborate than the simple Rule of 20, but based on a more rigorous analysis of inflation and P/E ratios, the current equity market appears, at worse, to be fairly valued. Investors forget that inflation was increasing leading up to the 2008 bear market. In fact, the CPI, which is a lagging indicator, peaked at 5.6% in July 2008. Today’s headline inflation is 1.0%.”
Bernstein’s suggestion is that the market “appears, at worst, to be fairly valued” when one does a “more rigorous analysis of inflation and P/E ratios”. So I’ve gone ahead and done a more rigorous analysis of inflation and P/E ratios.
What we find is that Bernstein is telling, well, a half truth, which is better than most market analysts these days. Specifically we find a strong inverse correlation between inflation (blue line) and P/E (red line) and this supports Bernstein’s proposition that periods of low inflation support higher P/E multiples. However, in 2000 we see P/E break through a 120 year ceiling on the long term P/E trend despite having far lower historic inflation values over that period than we find today.
This suggests that while inflation is playing some part of today’s near record valuations the P/E is still approximately 28% above where it should be after taking inflation into account. And so then we need to be a bit more rigorous than was Mr. Bernstein if really understanding today’s multiples is truly the goal.
In order to fully understand the rising P/E multiple phenomenon let’s first review some basics. What does P/E imply to the investor? Well in a very basic way, that if earnings growth = 0% then the P/E multiple is the number of years to breakeven on the investment. As growth increases the breakeven becomes something less than the P/E multiple. And so an investor, given some growth expectation, can decide what multiple he/she is willing to pay based on how quickly they need to breakeven and begin generating some positive return.
What Bernstein is suggesting is that P/E valuations are not just a function of growth but are also a function of cost, namely from his perspective, inflation. When costs are low, everything else equal, returns will be better is the thesis. And this is where his hypothesis falls short. First, everything else is not equal. Second, he leaves out a few costs.
He rightly incorporates a cost function to P/E but he mistakenly limits costs to inflation. In order to understand the excess 28% P/E premium beyond what can be explained by inflation we need to look to interest and opportunity costs. What we find is that these two costs too are historically low. And so one might assume that because the overall cost structure of investments has moved to record lows, we can pay a P/E premium for assets given the low costs should produce higher returns.
But, and this is a big ‘but’, we cannot leave the analysis there. Remember risk is a cost. So let’s look at risk. Currently we have record debt levels throughout both the public and private (enterprise and households) sectors and that equates to record leverage risk (don’t forget market cap is very fluid while debt obligation is very rigid – meaning those debt obligation to equity multiples may not be what they appear to be…). So we have elevated leverage risk throughout the system.
Risk is also a function of growth. Today we have historically low growth on both the macro and micro levels. We rarely think of growth this way but growth has an inverse relationship to risk. That is, when growth is robust risk is subdued as robust growth can offset bumps in the road. Low growth economic environments generally mean payout ratios are high (and thus excess cash is low) and so even small hiccups can have a material negative impact. In a sense we are walking very close to the edge.
And we have also created a tremendous amount of systemic risk in the overall system due to the record low opportunity costs (directly and indirectly due to Central Bank policy). As mentioned above low costs should imply higher returns. However, when opportunity costs are so low (meaning there are very few alternative investments generating required rates of return) that investments get further and further concentrated we begin to build a great deal of systemic risk.
Think about a large body of water that maintains a large ecosystem. As that water dries up and the pool gets smaller and smaller more and more creatures are being supported by that shrinking pool. Assets are the same. As the landscape for investment opportunities begin to disappear (the existence of negative rates imply broader risk adjusted positive return opportunities do not exist) more and more investors are forced into the same pool of assets that are still providing positive returns.
And this means a very large proportion of total asset value is totally dependent on this very small and shrinking pool of assets. In a market, such asset scarcity leads us to the record valuations (e.g. record P/E’s), via increasing equity demand but shrinking supply as corporations reduce outstanding shares through buybacks. The result is that upside becomes very limited while the downside potential becomes very robust.
It might be useful to have a look at a recent real world example. Mondelez (MDLZ) recently bid a 30x P/E for Hershey (HSY). Given HSY has an earnings growth of around 2% (that’s being generous), it will take 23 years for MDLZ to break even (even if they triple growth to 6% it’s still a 17 year breakeven). Now using the long term SP 500 average P/E of 15 and the long term earnings median growth rate of 11.5% we find the average breakeven is about 8 years.
And so we see the effect of a shrinking landscape of investment opportunities. Specifically returns are lower (23 yr break even vs 8 year break even), not higher as the low inflation and interest costs would predict. So while purchasing power parity (i.e. low inflation) solves some of the record valuation conundrum, 28% of the historical overvaluation appears to be a function of investment concentration and that means lower not higher expected returns. For a money manager to suggest that doesn’t create significant risk I believe is foolish, if not negligent.
This is all a direct result of the ignorant economic policies, both monetary and fiscal, that have been implemented over the past few decades. Deteriorating incomes, household balance sheets and breadwinner jobs and slowing population growth rates have resulted in severe demand deterioration. Expansion of consumer credit and government transfers have peaked and with no income growth there is then very little companies can do to expand the top line.
The data is very clear that revenues are declining across the spectrum and this is a definite canary. We’ve managed to expand EPS for several years by way of economic cannibalism (contracting operations and financial engineering) but this too is coming to an end. There is simply less and less fat to cut and capital to reallocate and as all firms cut what’s left it becomes a death spiral for demand as consumers are forced out of good paying jobs and into min wage, temp and part time work on a macro scale.
So let’s wrap this up. Is Bernstein correct about P/E levels and inflation? Well he’s half right. Sure low costs, everything else equal, will support higher valuations as they imply higher returns. However, the structural economic problems of stalled incomes, peaked debt and welfare make operational expansion i.e. sustainable growth extremely difficult, which has led to investment concentration in secondary equity markets. And that means the higher valuations simply represent higher risks.
The offshoot is that as such a large concentration of total asset value is dependent on the market, it becomes necessary to maintain the market at all costs. The market has become too systemically important to allow it to fail. And that means policymakers have changed the function of the market. The market left to its own devices is a consequence of the underlying economy. Today, however, the market is being used as a (false) portrayal of the underlying economy. It is intentionally using the logical fallacy of confusing cause and effect.
That is, the thermostat is no longer meant to reflect the temperature inside the house, its only use is to convince you the house is warm. That means policies are being targeted at manipulating the thermostat rather than keeping the furnace hot. The consequence is a spiraling of resource misallocation, furthering the structural breakdown of economic activity making it ever more important to keep the market looking strong. The market is no longer a market as we understand the term.