Bearish calls based on mean-reverting profit margins may be missing important points.
I generally avoid using forward earnings in valuing equities. I have therefore not spent much time writing about profit margins and whether or not they are about to mean-revert, one of the main points of the equity bear population.
Also, I tend to avoid forecasting future equity levels, preferring to concentrate on evaluating fair market values and monitoring the risk/reward equation along with economic momentum in order to assess whether equities should be favoured or not.
This blog began in early 2009 and rapidly advocated buying equities on the basis of very attractive valuations based on a detailed analysis of P/E ratios over the previous 80 years (S&P 500 P/E Ratio at Troughs: A Detailed Analysis of the Past 80 Years). During the following 4 years, using unbiased economic analysis and the objective Rule of 20 valuation tool, I modulated my general bullishness along with the fluctuations in the risk/reward ratio provided by the Rule of 20’s factual analysis (black line in chart below). Managing risk is what this blog is all about.
In DRIVING BLIND, I recently highlighted the facts that S&P 500 trailing earnings peaked one year ago and that revenue growth has decelerated to +1.3% in Q1’13 with most indicators pointing to more weakness in coming quarters. This suggests that maintaining profit margins will be a real challenge for corporate managers, unlike the past several years when rising revenues combined with sharply decelerating costs boosted margins to all-time highs even though world economies remained pretty sluggish.
(Gavyn Davies in FT)
Wondering what is the downside risk in profits, I recently started to look into profit margins and the reasons for their admittedly elevated level. Among the many straight line forecasters and the several astute and patient mean-reverters, few seem to have done a thorough analysis to support their thesis, other than posting charts like the one above and ridiculing people who say “this time may be different”. I wonder what such people were saying back in 1995. While margins have been pretty volatile lately, the two recent lows never went back near previous lows while new highs were recorded in each of the last 3 cycles. Patience is an essential investment virtue, but there is a limit. As Keynes said, in the long run, we’re all dead.
The relationship with wages is most interesting and obviously revealing of some secular trends. The slow declining long term trend in the share of wages observed since the 1960s clearly accelerated during the last 10 years. The easy explanation is that high unemployment allowed corporate America to squeeze labor as the caption on the chart above suggests.
This may be true of the last 4 years but productivity gains have been accelerating for much longer. The huge advances in technology since 1995 coupled with accelerating globalization have strongly contributed to the more recent productivity gains. Previously, these gains eventually transpired into lower prices, keeping margins pretty constant over time. The fact that profit margins have kept rising since 1995 may have more to do with the changing complexion of the economy than with the evil side of capitalistic corporate America. In fact, Gavyn Davies’ next chart shows that the same phenomenon is observed in all advanced economies.
Raj Yerasi, a money manager, wrote a guest post for Greenbackd to show how rising foreign earnings have helped boost American corporate profit margins. Importantly, he explained how the quirks of the national accounts can distort the reality:
To understand this, it is important to note that current analyses do not directly measure profit margins per se (meaning, profits divided by revenues). Rather, they measure corporate profits as a percentage of GDP, which captures not total revenues but the total value addition of corporations (along with other components). While there are multiple potential data issues in comparing profits to GDP, it nonetheless stands to reason that profits as a percentage of GDP should generally correlate with profit margins.
However, one big source of error is that the most widely known NIPA corporate profits data series, which the analyses referenced above appear to be using, represents profits generated by corporations that are considered US residents. As such, this data series includes profits generated by US companies’ international operations (e.g. Coca-Cola India, Coca-Cola China) and excludes profits generated by foreign companies’ US operations (e.g. Toyota USA). GDP, meanwhile, captures all economic activity within US borders, whether undertaken by US companies or foreign companies, and it excludes any economic activity abroad. It should be clear that one cannot compare these two metrics, since the corporate profits data series introduces profits generated by other economies and excludes profits generated by the US economy.
US companies’ profits from abroad have grown tremendously over the last 10 years, much more so than foreign companies’ profits from US operations:
This skews the calculated profits level upwards and by an increasing amount over time, making profit levels today look exceedingly elevated.
To do the analysis correctly, we need to use data that are more apples-to-apples. Fortunately, the NIPAs do include a data series of corporate profits that simultaneously excludes US companies’ profits from abroad and includes foreign companies´ profits from US operations, called “domestic industries” profits. Comparing these profits to GDP, profit levels still appear elevated but now not as much as when using the prior “national” profits data series:
He then tackles another important factor:
It is also worth noting that effective corporate tax rates are lower today than in the past. Per the NIPA data, tax rates have decreased from about 45 percent in the 80s to 40 percent in the 90s to 30 percent in recent years. Using pre-tax “domestic industries” profits as a percentage of GDP, profit levels today may be closer to 20 percent elevated relative to historical norms. (…)
National accounts are one thing but, in reality, we invest in equity markets. The composition of the S&P 500 Index is different than that of the corporate component of the national accounts. Importantly, the industrial complexion of the Index fluctuates over time. To the extent that certain industries have very different structural operating margins and effective tax rates than other sectors, changes in industry weightings within the index will impact the total.
This chart from the American Petroleum Institute (note the typo, they meant 2012) shows how net margins vary by industry. Pharmaceuticals, tobacco and technology enjoy much fatter margins than most other sectors.
Part of the reason is found in the respective effective tax rates, likely the effect of R&D expensing and foreign earnings:
Bespoke Investment plots historical sector weightings of the S&P 500 Index. The 15% increase in Technology stocks’ share of the Index over the last 20 years, reflective of the huge advances in technology referred to above, no doubt has had an impact on total Index margins. The strong rise in importance enjoyed by Tech and Heath Care companies, combined with the decline in Industrials’ weight, phenomena observed in all advanced economies and supported by normal evolutionary and demographic trends, surely explain much of the secular growth in Index margins.
Another factor for currently elevated margins is historically low interest rates which have substantially reduced financing costs in recent years, boosting margins more so than in previous cycles. Obviously, this will succumb to higher rates later in the cycle.
In brief, arguing that margins are historically high and assuming they will necessarily mean revert appear to be too simplistic analysis, at least until we begin to see a real trend toward tax rates normalization across the world.
I don’t have the resources nor the time to try to thoroughly explain why margins have increased over time. It would be useful if large organisations like GMO and Hussman Funds did and shared thorough research on this non-trivial matter.
The following charts from CPMS Morningstar plot a proxy for net margins for various market sectors since 1993. Current margins are in effect high and have been rolling over. Note how margins and trends can vary significantly over time. Also, cyclicality varies meaningfully between sectors. Importantly, note the volatility at a high level for the CPMS average during the 1990s.
Forecasting aggregate margins is a perilous activity, especially if done without a good understanding of the root causes for trend changes. Making investment decisions on the basis of such forecasts can be very frustrating.