Cyclically Non-Adjusted EarningsEric Cinnamond
Last week I played tennis with a mortgage broker I’ve known for over twenty years. In addition to being a good friend and tennis player, he often provides me with a timely and accurate assessment of the Florida housing market. Since we hadn’t talked in several months, I was eager to hear what he had to say.
As I drove up to the tennis court, I watched my friend step out of a shiny new $100,000+ luxury vehicle. I knew immediately an update of the Florida housing market wasn’t necessary! I greeted him by saying, “If this isn’t a sign of the top, I don’t know what is!” We both laughed. He proceeded to show me all the different features and gadgets on his new ride. It was impressive.
While not as shiny and new as my friend’s luxury car, the current profit cycle has also been impressive. Approaching its 10-year anniversary, today’s profit cycle has shown considerable persistence and strength since its 2009 beginnings. As the current expansion marches on, I’ve been thinking more and more about business cycles and how I incorporate cyclicality into my valuation process.
Historically, I’ve used a normalized cash flow assumption when calculating the value of most businesses. My preference for normalizing is based on my belief profit margins and earnings for most businesses are cyclical (albeit to differing degrees). In effect, normalizing enables me to avoid valuing businesses on peak or trough cash flows, in addition to the large valuation errors that can accompany extrapolation.
Normalizing was essential in helping me navigate through the tech and housing bubbles. It allowed me to avoid becoming too optimistic during the booms and too pessimistic during the busts. The results of normalizing during the current cycle have been less conclusive. While normalizing was beneficial during the earlier years of the current cycle, it has appeared unnecessary and even counterproductive in later years.
The extended duration of the current profit cycle – currently twice as long as the average cycle – has challenged the value and usefulness of normalizing. This cycle’s length is also affecting normalized valuation metrics, such as the Shiller PE. As a reminder, the Shiller PE uses a 10-year earnings average in its attempt to smooth or cyclically adjust earnings. As discussed in the post “Normalizing Earnings and Real Rates,” assuming the U.S. economy does not enter a recession in the near future, the Shiller PE’s 10-year earnings average will soon consist of all economic boom and no bust. As the depressed earnings of 2008 and 2009 roll out of its calculation, average earnings will increase and the Shiller PE will likely appear less expensive.
The 10-year anniversary of the current profit cycle raises an important question. How useful is a cyclically adjusted PE (CAPE) that only includes a period of profit expansion? While I don’t know Robert Shiller personally, I suspect when he developed his CAPE, he didn’t envision a 10-year profit cycle without a recession. I know I didn’t!
The longevity of the current profit cycle is also influencing how investors perceive the cyclical nature of operating businesses and the economy. Based on equity valuations, investors appear to be dismissing the risks associated with recession and profit reversion. In effect, the longer the current cycle persists, the greater the temptation to abandon normalizing and embrace extrapolation.
There are many examples of extrapolation in today’s market and business environment. From an individual perspective, look no further than my friend’s new luxury vehicle (after asking his permission to write about our conversation, he reminded me that he also recently bought “another” house!). Would he have made such an extravagant purchase if he believed a decline in Florida real estate was forthcoming? Probably not. Instead of normalizing over an entire cycle, I suspect he extrapolated current business trends well into the future. While I’m not certain if there has ever been a real estate boom in Florida without a bust, I suppose it’s possible. That said, if it were me, I would have normalized and went with the Toyota.
In addition to individuals, there are many examples of investors assuming extrapolation risk. The most obvious is the use of recent earnings results and forecasts to calculate the value of equities. One of the most popular methods of equity valuation is to simply apply an earnings multiple to next year’s earnings estimate, with many estimates assuming current trends will persist without interruption. In effect, earnings predictions are based on the past several quarters of operating results (extrapolation), not full-cycle margin dispersion and scenario analysis (normalizing).
A less obvious, but interesting example of investor extrapolation appeared when the corporate tax rate declined from 35% to 21%. The response to the lower tax rate was overwhelmingly positive as earnings estimates and equity prices rose considerably. Instead of normalizing historical tax rates (according to Trading Economics the average tax rate from 1909 through 2018 was 33%), investors appear to be extrapolating the lower rate well into the future. This is despite the fact that future tax rates should be expected to fluctuate along with shifts in fiscal policy and political leadership.
Business leaders also appear to be extrapolating recent profit trends as many companies aggressively pursue acquisitions, buy back stock, and issue debt. Corporate decisions based on extrapolation are often responsible for the eventual industry or economic bust. The energy industry is a good example. When oil was trading over $100 a barrel, energy companies extrapolated boom-time operating results and expanded aggressively – the future appeared certain. Instead of certainty, the energy industry experienced supply increases, price declines, and bankruptcies.
As analysts and portfolio managers, it is not our job to know the future with certainty, but to properly incorporate uncertainty in our valuations and investment processes. While current equity valuations imply a high degree of certainty in future profits, abundant examples of extrapolation and boom-time decision making support my belief that human behavior and profit cycles remain cyclical. As such, I believe normalizing, not extrapolating, remains the preferred method of business valuation for absolute return investors seeking attractive full-cycle returns. The future will be different from today – guaranteed. So why not value businesses accordingly?