The media is flush with articles suggesting that the market is overpriced, that stocks are a terrible investment. Perhaps more than any other metric, the Shiller PE10 ratio is cited as firm evidence that while stocks may look cheap on current earnings, the cycle no doubt will swing back to recession, bringing margins and earnings back to earth.
While it is hard to argue that we are not near a cyclical peak in earnings, I believe that the Shiller PE10 is far too simplistic a yardstick to determine whether or not equities are cheap. Not only that, but it's highly impractical. Proponents advising buying when the PE10 Ratio is cheap, (below say 12 or 13x earnings) may be waiting literally for ten or twenty years before we get another crack at buying stocks!
Here, I explain partly why the PE10 ratio is seemingly high, as well as proposing an alternative means of looking at equity markets to determine cheapness. To me, a more forward looking strategy makes more sense for the long term investor, as well as one that incorporates forward interest rates.
First, what is the Shiller's PE10 Ratio? The math is simple: it's the price of the market, divided by inflation-adjusted earnings averaged over the past 10 years. In theory a long term average smoothes earnings out throughout the economic cycle. The chart below seems to offer compelling evidence that stocks today at 22.5x PE10 earnings are simply quite expensive. Those who rely on PE10 ratios would argue that today's P/E ratio of 13x (based only on 2012 earnings) is illusory given the coming recession and anticipated normalization of earnings and margins.
Here is the chart from Shiller's website:
click to enlarge
Given the faltering economic news lately (especially out of Europe, but also in the US and China), a valid argument can be made that we have reached peak cycle earnings, and hence the next shoe to drop is the economy. Recessions unquestionably lead to lower earnings; I do in fact agree that caution is warranted right now. But am I dumping all my stocks? No.
Conclusions of the PE10 Ratio Believer
An investor HAS to be skeptical that the PE10 is the end-all be-all to valuing the market. Yes, it clearly indicates that the past eras like the 1950s and the late 1970s were great times to buy stocks in general.
But what about the past 30 years? It completely ignores current and future interest rates and growth rates. I can only assume adherents to the PE10 would say that you should buy stocks below say 12 or 13x, like the early 1980s, and then sell them above say 20x (as pundits are advocating selling them now at 22.5x).
The problem with this investing theory is that I essentially haven't had one era of "cheap" equities in my entire adult life! Sure there was a dip to 15x in 2009, and if I had relaxed my 12x P/E multiple standards (by a full 20%!), then perhaps I might have thought about buying stocks. But again, I had one single month to purchase equities going back to essentially the mid 1980s if I follow this argument!
Put differently, if I missed that small window to buy in 2009, do I really have to wait perhaps a decade or two to be comfortable buying stocks again? What do I do with my capital in the mean time?
By the way, here is the math on owning 20 year treasuries. If I purchase 2% yielding 20 year Treasuries, and inflation is 3%, then in 20 years I will have guaranteed 20% cumulative real (inflation-adjusted) losses. That stinks! Is it really possible that real equity returns will be worse than that?
And lastly, do earnings in 2002 and 2003 really impact whether I buy stocks today? To me the PE10 incorporates too much weight on earnings 8, 9 and 10 years ago. Eli Lilly for example generated average EPS of $2.46 over the past 10 years, implying a PE10 ratio of 17.4x. I am not sure what that tells me to be honest. Does the fact that Lilly did $2.82 in EPS in 2002 have anything to do with what the stock will do in 2013? I'd argue no.
In my opinion, the Shiller PE10 at its core is mean reversion theory at its simplest. We need something better.
Explanation of Higher and Higher PE10 Ratios
A question to ask is, WHY has the PE10 ratio show such a consistently high level compared to current P/E ratios? That is, apart from the obvious tech bubble times in the late 1990s?
The answer is surprisingly simple. If the ratio is the average of the prior 10 years, then it goes without saying (or perhaps not), that profit margins must have been steadily rising. That is, throughout recessions in 1990, 2002, and 2008, margins may have dipped, but generally margins recovered, and then reached even higher peaks.
It's not hard to quickly outline several reasons for steadily rising profit margins since the early 1980s. Obviously productivity improvement is the number one reason, but here are several others either driving that or contributing directly to higher profitability:
- Declining clout of labor unions, causing a larger share of margin to owners of capital.
- Women's increasing share of the labor market, in light of the fact that female average wages are lower than men's. (Again, more margin to capital owners).
- Increasing corporate productivity, for obvious reasons driving margins higher.
- Globalization, meaning wider pools of cheaper labor as well as driving more businesses to more actively seek out greater efficiencies.
- Deregulation of industry, implying fewer compliance costs.
- Worldwide privatization of industry. As more and more nationally owned businesses enter the private world, inefficiencies by necessity are wrung out to remain competitive without taxpayer support.
- Decreasing cost of capital in the form of lower borrowing rates for corporations, particularly compared to the 1970s and 1980s.
- Higher unemployment. A more recent phenomenon, high unemployment rates simply equate to worsening bargaining power by new hires.
- Growth in sales overseas in lower tax jurisdictions. When a company as enormously profitable as Google for instance pays less than 15% tax rates, clearly globalization is pressuring tax rates, improving margins.
The chart above highlights partly why the PE10 ratio has continued to show an ever-increasing expensive market. Productivity gains are entirely benefiting the owners of capital. In essence, PE10 adherents have to believe that hourly compensation will revert to a high correlation to productivity gains, as was the case pre-1975. But with such high unemployment amidst our "new normal" economy, it seems a dubious prospect.
A Better Method
Since I have no intention of sitting on cash or in bonds for the next one or two decades waiting for that elusive golden entry point, I offer a better means of looking at the equity markets. Namely, I assume that profit levels are likely to remain at elevated levels.
This is admittedly a risky bet. Note though that I am not suggesting that profit margins won't fall however. In fact I fully expect that we will experience a recession sometime in the next one to three years (or months?!). That could easily tank profits by 15%, lowering margins as well.
But neither would I suggest that until the S&P falls by 50%, I wouldn't touch stocks. In fact, that kind of a drop to me is highly unlikely. Even in 2008, easily the scariest investment era that I have ever seen, the S&P fell 38%. Amongst the financial illuminati of the world, it seemed that we had not only a terrible recession on our hands, but a global financial meltdown of epic proportion. Stock reacted accordingly.
I simply think no central banker in his right mind ever allows another liquidity crisis to unfold again, the way the US Federal Reserve (helped by the Bank of England) let Lehman Brothers file for bankruptcy. Every single major big banking near-failure since Lehman in September 2008 has been thrown ample capital, whether it's Spain's Bankia or TARP money to the US banking system. Central bankers want no part in shutting down the global financial system.
A recession in the not-to-distant future is inevitable. A recession combined with a liquidity crisis? I doubt it. In fact, I would argue that the market still lives in fear of this possible deadly combination. Every hiccup in Greece or Spain is obsessed over by the media, with investors continuing to avoid stocks and purchase ridiculously overpriced bonds. The market already is discounted to some extent for another 2008 breakdown.
I mean, in 2007 the S&P traded at 1468 on average, and generated only $83 in EPS (almost 18x earnings). Today we are at lower 1400 levels, with the S&P likely to generate around $104-105 in EPS. That is a 13.4x multiples, 24% lower than in 2007.
In any case, for this alternative method to work, you have to believe that the nine profit margin improvement trends (listed above) continue to some extent. While it is arguable that say government regulation and higher taxation will crimp margins in the future, I also think that unemployment rates will remain high, globalization will continue, technological change will continue to foster productivity gains, interest rates will stay low, and privatization of industries worldwide will continue.
These seem like reasonable economic assessments to make. And if they do continue, profit margins likely aren't going to fall back to the 4-5% levels witnessed in the 1970s and early 1980s. It took extremely high interest rates coupled with more highly levered companies in that era to push margins so low. Not to mention an oil crisis or two. The world is far different than it was in 1980.
Here is a chart of margins going back to the 1940s.
All we really need is for profit margins to remain in the same ballpark as we have witnessed over the last decade. Let's call it 7-9% as opposed to the long term average of 6%. Today's near 11% level seems extraordinary, and does warrant caution however. I do want to point out that during the late 1990s margins contracted, mostly because unemployment rates were so low. As it became exceedingly difficult to find good hires, wage pressure escalated. Today, wage pressure today is nil, and unless we see some real traction in hiring, likely to remain nil.
Using 8% as Normalized Margins
For arguments sake, let's assume a more normalized profit margin through another cycle is around 8%. While S&P EPS today looks like $104, implying an 8% margin (down from almost 11%) on 2013 revenue implies something like $82 in EPS. That means today the market is trading at 17x next year's normalized earnings. This is not only forward looking but more reasonable than a 22.5x PE10 multiple. In fact, a 5.8% Free Cash Flow yield compared to 1.7% Treasuries may not seem that terrible. (I simply inverted the 17x PE multiple to get the 5.8% FCF yield).
Looking out 5 years, assuming meager 1% real growth and 3% inflation, (less than the 2% real growth we are seeing today), means that EPS in 2017 could approach $127 on the S&P500. Put differently, this is a pretty terrible GDP real growth scenario of only 1% per year, and uses margins 30% lower than where we are today.
Where does it make sense to add exposure if this turns out to be the case? I would suggest at 8-10x a normalized 2017 figure, which is 1020 to 1270 on the S&P500. The middle of that is around 18% lower than today. Essentially, if I have a 5 year time frame, and believe that margins may fall, but average at least 8%, then I am happy to own the market at around 1100-1200 if we do head into a recession.
At today's 1400 prices on the S&P500, this conservative case only gets you 3% in return per year. That is not too good and suggests to me that today's market does seem a little pricey. But 3% is still a better return than bonds! Not to mention that if I can select better performing stocks than the market, then I can do much better than bonds today. Needless to say I may be lightening a stock here and there, but not dumping them by any stretch.
By the way, if I can buy the S&P at 1200, then this low growth case provides a healthy, nominal gain of 6.6% per year for equities.
Upside Case, 11% Margins
Assuming that the trends in productivity continue, then 11% margins in 2017 may actually be achievable. Assuming only 1% real growth at 11% margins implies $174 in S&P earnings per share in five years.
If the S&P can get to $174 in EPS in 2017, then if stocks traded at 13x 2017 EPS by the time we get there, then we could reach 2262 on the S&P, 61% higher than today! That would net you 10% per year in nominal market gains, far better than 1.7% bonds.
Its hard to ignore the optionality that the market may do quite well, even through a recession over the next five years.
Quick Note on Fiscal Deficits
US fiscal deficits are a huge wildcard in the equation. Government spending in excess of revenue generally equates to higher levels of GDP than would otherwise be attained. With $1TT deficits, GDP essentially is over-inflated by $1TT per year. Cutting deficits by raising taxes and curtailing spending likely would be a big drag on profitability. Think of lost sales, and lost economies of scale at the corporate level.
If the US Government attempts to spend $200BB less per year (to cut our $1TT deficit to zero in five years), then no doubt that would impact margins. In simplistic terms, $200BB cuts every year for 5 years, divided by a $15TT economy is 1.0% in margin impact per year tax adjusted. Likely real GDP growth of 2% per year may offset this, but it's certainly going to be a drag on profits and margins going forward. Its also why I assumed only 1% real growth through 2017.
At the end of the day however, normalizing margins at 8% seems quite a reasonable 5 year drop even given expected fiscal deficit reduction. But I wouldn't ignore the bull case either.
There are any number of conclusions you can obviously make here depending on your view of the sustainability of high profit margins. More than anything I think investors should rethink any oversimplistic measure like the PE10. In fact, apart from taxation and government interference, higher margins are likely here to stay. I wouldn't argue that 11% is sustainable, but if 8% is sustainable, then at 1200ish on the S&P, I get pretty excited, especially when compared to negative real returning Treasury bonds.