In recent months, I have been exchanging views and data with the CIO of a highly respectable investment advisory firm who is very concerned about valuations and future equity returns. His last email attached a weekly commentary by John Hussman and his own comments:
He’s so right that investors look back and if they see a series of very high returns they feel good and increasingly bullish.
He’s also right that in the long run valuations have the best correlation with future returns than any other metric/factor. The higher the valuations, the lower the future returns.
Pundits use the fact that the market has kept rising despite being increasingly overvalued as proof positive that valuations don’t matter. The reality—as repeatedly demonstrated by history–is that the higher they go and the longer the inflated phase lasts, the more likely a brutal shakeout becomes.
All true statements, generically speaking. But Hussman is very present-day specific in his assessment of equity markets:
We are at the far edge of a monumental mesa here, but speculators are ignoring the cliff, assuming that they are on a permanently high plateau. The unfortunate aspect of these mesas and valleys is that they encourage backward-looking investors to believe that projected returns based on “old valuation measures” are no longer relevant, precisely when valuations are most informative about future returns.
(…) we do expect S&P 500 total returns to average a loss of just over -2% annually over the coming decade, or about -4% excluding dividends, which would place the S&P 500 about 35% below current levels in 10 years (the decade following 1999 was worse).
(…) there is ultimately little way to mitigate the obscene valuations we observe at present without a 50-60% collapse in the S&P 500 Index over the completion of the current cycle. (…)
The real issue is that earnings-based valuation measures dramatically understate the actual extent of overvaluation here. That’s likely to be a problem not just for investment returns over a 2-3 year horizon, but for investment returns (at least measured from current valuations) on horizons that extend well beyond the coming decade.
Hussman gives a lot of weight to the Price/Sales ratio (he uses MarketCap/GVA) and asserts that the current P/S level is unjustifiably high, even more so when considering that profit margins are at unsustainable levels and will collapse, “like they always do” when the main reason for their elevated level, “low real unit labor costs”, eventually reverses.
I use P/S regularly when assessing individual stocks, always relating it to the trends in each company’s profit margins in order to validate current valuations. It is generally easy to identify the causes of rising or declining margins at individual company levels and assess their sustainability.
Index profit margins, let alone economy-wide profit margins, are more difficult to analyse, in part due to shifts in the structure of the index or of the overall economy. I have written extensively on possible reasons behind the rise in profit margins (CETERIS NON PARIBUS, Certain Uncertainties, HARD HAT ZONE), including “increased globalization”, “increasing corporate power” and “monopolistic behaviour” (reasons Jeremy Grantham says will keep margins from mean reverting).
Low cyclical labor costs are not as important as Hussman suggests since a sectorial analysis of profit margins clearly shows that only two sectors realized clear and meaningful margins improvements since 1997: Energy and Consumer Staples. All other sectors in the S&P 500 Index experienced flat or lower margins. I concluded in “This Time Seems Very, Very Different.” Really?:
The rise in S&P 500 margins since 1997 is not the result of a widespread margin boost, nor of lower interest rates, but rather the result of increased concentration in the Consumer Staples industries and the five-fold jump in oil prices between 1997 and 2014. No other sector has shown meaningful uptrends in margins during that period (…)
In all, without crystal balling, there are enough facts to discredit the notion that higher margins are here to stay. Time will tell but 8 years into this recovery, with rock bottom interest rates, very slow wage growth and collapsed energy prices, there have been no new highs in corporate margins in recent years.
The chart below plots the U.S. non-financial P&L statement from the national income and products accounts (NIPA) back to 1947. It reveals the significant break that U.S. corporations got from slow labor costs growth since 2000. Labor costs as a percent of gross value added peaked at 66% in 2001 and troughed at 57% between 2010 and 2014. Pretax margins (blue line) bottomed at 6.0% in 2001 and rose by the full 9% savings to 15.5% in 2014. Net after tax margins (black) reached 11.8% in 2014, up from 4.4% in 2001.
Hussman’s assertion that low wage costs are the primary reason for higher corporate margins is thus technically right economy-wide. We can readily see how pretax and net margins currently seem to be negatively impacted by higher labor cost ratios.
Labor costs have risen faster than revenues since 2015 but it had more to do with very slow revenue growth than accelerating wage costs. Total U.S. business sales actually declined in 2015 and most of 2016, along with producer prices, giving the impression that profit margins were being squeezed by rising labor costs.
In reality, business sales (and PPI) were dragged down by falling oil prices:
In effect, the recent decline in NIPA profit margins is all because of falling oil prices which dragged total business sales lower. Sales have recovered in 2016 along with oil prices and given that sales are currently rising 5-6% YoY, we should witness somewhat better margins in the early part of 2017. However, oil prices have lately turned down YoY (-6.1% in June) which will no doubt pull down the growth rate in the PPI and business sales as lower oil costs permeate throughout the economy, threatening total corporate margins into the second half of 2017 and 2018, even if labor costs don’t accelerate.
As the BEA explained in a 2001 paper, the long-term trends of NIPA profits measures and the S&P profits measures are roughly similar, but the year-to-year changes sometimes differ substantially. This is due to significant differences in coverage, industry representation, and accounting principles. S&P 500 Index earnings are roughly 50% of total U.S. corporate profits (a ratio that can vary meaningfully from year to year). The industry composition of the S&P 500 index does not reflect the structure of the
overall economy. In addition, the composition of the S&P 500 index is based on market values, so strong market sectors—such as technology—may have a much higher weight in the S&P profits measures than in the NIPA profits measures.
NIPA data reflecting profit margins were very similar to S&P and Thomson Reuters data up to 2008 but have been diverging since. S&P margins (including Financials) followed NIPA downward in 2015 but both S&P and TR measures have turned up in 2016.
NIPA profits from “Petroleum and coal products” have completely evaporated in 2016 while S&P 500 Energy earnings returned positive in Q3’16 and reached 36% of their 2014 average in Q1’17. The NIPA industry breakdown is not available yet for Q1’17 but the 2016 quarterly trend offers little hope for an early turnaround in 2017. Obviously, the S&P 500 Energy sector is very different than that in the overall economy which includes hundreds of small producers very negatively impacted by low oil prices.
Meanwhile, the relative importance of Energy in the S&P 500 Index has cratered from 14% in 2009 to 10.2% in 2013 and to its current 6.5%. At the other extreme, IT, Financials and Health Care now account for 54% of the S&P 500 Index, from 47.8% in 2013 and 42% in 2009.
Q1’17 earnings rose 19.9% YoY for Financials, 7.3% for Health Care and 19.8% for IT (per TR). Their respective revenue growth rate was 8.8% for Financials, 6.0% for Health Care and 7.2% for IT, indicating substantial improvements to the profit margins for these very heavy S&P weights. Three other sectors improved their margins in Q1’17 but their combined weight is only 15.3%.
Expectations for Q2 are for rising margins on total EPS up 7.9% on a 4.6% revenue growth rate. Excluding Energy, growth rates drop to +3.7% for revenues and 4.8% for earnings. Interestingly, Financials and IT are providing the bulk of the growth and margin improvement. Ex-Energy, 7 of the remaining 10 sectors are expected to see margins compressions in Q2.
Looking at the next 6-12 months, the recent trend in oil prices should be a strong restraining factor on PPI and business sales growth as seen above. NIPA margins should thus continue to trend down in this environment of subdued final demand and tightening labor resources.
Will S&P 500 companies margins continue to trend differently? Analysts think so as these Ed Yardeni charts show:
But trends in S&P 500 revenues correlate very well with trends in Business Sales and analysts could be overoptimistic if oil prices and PPI inflation remain weak.
It is interesting to see how profit margins have recently diverged between large, mid and small cap stocks:
S&P 600 company margins have dropped from 6.0% in 2013 to 4.9% with all 11 sectors but one (Utilities) seeing lower margins over the period. Big drops occurred in Energy, of course, but also in Industrials, Consumer Discretionary and Staples and Health Care.
The erosion in small caps margins continued in 2016-17 while margins of mid and large caps improved. The S&P 600 Index is 35% Consumer Discretionary and Industrials and only 14.5% IT compared with 23% and 22% respectively for the S&P 500 Index. Smaller companies are essentially domestic and generally more labor intensive than larger ones.
In effect, trends in S&P 600 margins are currently similar to the trends in the overall economy as measured by NIPA data. In 2006, NIPA margins started to weaken in Q3, two quarters before small caps margins and 4 quarters before large caps margins. This also tends to be the pattern in market breadth whereby end-of-cycles normally become apparent in deteriorating breadth in small cap stocks first, then mid caps and finally large caps.
This cycle, NIPA margins peaked in the 4th quarter 2014 and small caps 2 quarters later. Yet, large cap margins keep hitting new records. More puzzling, Lowry’s Research observes that “the OCO [Operating Company Only] Small Cap Adv-Dec Line, which is usually the first to show signs of deteriorating breadth, reached a new bull market high on July 3rd and is leading both the OCO Large Cap and OCO Mid Cap Adv-Dec Lines, which are only marginally below their bull market highs, set on June 19th and June 13th, respectively.” Investors are obviously not seeing the of this cycle just yet.
The reality is that margins just don’t fall from their own weight. The big drops really only occur during recessions when sales actually decline like in 2008-09 (business sales –20.4% at worst YoY point), and in 2001 (-5.0%). One would think that a meaningful slowdown in sales growth rate could surprise businesspeople and hurt margins but history does not support this notion. Since 1992, there have been 5 non-recessionary periods when business sales growth slowed markedly and only 2 led to a significant drop in NIPA margins: 1997-1998 (margins dropped 1.7% to 7.9%) and 2014-15 (-2.3% to 9.5%).
Businesses have thus proven pretty agile amid fluctuating economic conditions. NIPA margins hit a historical low at 4.4% in 2001. In 2009 they bottomed at 7.3% and since then, they have narrowly fluctuated between 9.1% and 11.8% during a rather challenging and economically highly volatile eight-year period.
Looking forward, a number of developments and observations need to be considered:
- Labor costs as a percent of GVA have been below 60% throughout this recovery, and actually below the lowest points since 1947. Could there be something structural (e.g. changing demographics, robotics) that will keep labor costs lower than historically going forward?
- Non-labor costs peaked at 30% of GVA in 2009 but have been shaved back below 28% even during a slow revenue growth environment. Cost control remains a priority for corporate officers.
- As a result, net margins have spent five consecutive years (2010 to 2015) above 10%. Since 1947, net margins have been above 10% only 13% of the time and generally only for very short periods prior to recessions.
- Net NIPA margins were sustained above 10% for 21 consecutive quarters between 1962 and 1967, a period during which real GDP growth accelerated sharply to a near 6.0% average and total inflation remained below 2.0%. It is only after inflation accelerated above 3.0%, on its way to 6.5% and a serious Fed tightening that margins started to shrink, reaching their low in Q3’70 near the end of the Fed-induced recession. Equity markets peaked in November 1968 at 18.9x actual P/E and 23.4 on the Rule of 20 P/E. The mesa cliff was 32% deep but equities only tumbled after inflation really accelerated and reached 4.5%.
- S&P 500 margins have been below NIPA margins since 2008. Yet, only 2 sectors have realized lower margins post 2008-09 than before: Energy and Financials. There may also be the increased costs of increased regulations since the financial crisis which have impacted larger corporations more than smaller ones. That said, public companies profit margins have recently exceeded their pre-2008 levels, whichever aggregator one uses and whether one considers operating or GAAP earnings.
As said before, margins just don’t fall from their own weight.
- There have been only 3 recessions in the 35 years since 1982, one every 12 years on average. Between 1947 and 1982, there had been 8 recessions, one every 4.2 years.
- Inflation was a recurring problem for central banks prior to the Volcker era (1982). This made central bankers continually trying to tame the price cycles.
- Inflation (headline CPI) has averaged 2.5% since 1982 compared to 4.6% between 1947 and 1982 and 7.4% between 1968 and 1982. Central banks have recently been more preoccupied with potential deflation, keeping a strong bias towards economic stimulation rather than contraction.
- Yet, inflation remains very muted.
When economic, inflation and interest rate cycles are much farther apart, calling the next down cycle becomes much more risky: time may not be on your side and good old mean reversion may take much longer than before to materialize. Big bears beware, the market can stay rational, though expensive, longer than you can keep your jobs…
All this to say, beware of the price to sales ratio applied to equities in general as a critical valuation tool:
- The recent drop in NIPA margins is mainly due to declining economy-wide revenues on lower oil prices and there is little evidence at this time that a generalized cost squeeze is underway, particularly when only considering S&P 500 companies.
- S&P 500 margins have been going back up during 2016 and look set to rise further in 2017.
- Accordingly, rising P/S ratio reflect rising margins and, barring a sudden spurt in inflation, the risk of a cyclical drop in margins from a Fed-induced recession remains low at this time.
- Similarly, we must acknowledge that labor costs ratios are surprisingly low and resilient so far in the cycle and that secular factors could well be at play and mitigate the next cyclical swing in margins.
- Hussman’s view that the current “obscene” valuations on revenues “dramatically understate the actual extent of earnings-based overvaluation” totally disregards the above factors and unnecessarily dramatizes the extent of the risk in equity markets.
The CPMS/Morningstar chart below plots the S&P 500 P/S and net profit margins since 1983. The current reading suggests that P/S is not “obscene” when measured against margins, unlike in 1998-2000 and in 1987.
P/E ratios of 19.7 times trailing earnings are undeniably elevated:
But the Rule of 20 P/E, which takes inflation into account, is currently only 7.6% above the 20 “fair value” in an environment of rising earnings and stubbornly low inflation.
The Rule of 20 says that fair value for the S&P 500 Index is [(20 minus inflation) X trailing EPS]. The chart below plots the two series. When the blue line is above the red fair value line, equities are above fair value and vice versa. The chart above simply plots the spreads to fair value, showing a very consistent 15 to 25 range over a very long period. This is a much more dependable valuation tool: it is based on earnings, the actual reason one buys equities; it is also based on trailing earnings and inflation, implying no forecast; and it has shown a very consistent pattern throughout the years and many and varied economic cycles.
Trends in the red fair value line are dictated by trends in trailing EPS and inflation. There is a clear fit between the trends in equities and the Rule of 20 Fair Value. Notice how the Fair Value line recently perked up as trailing EPS have kept rising and inflation recently trended lower. As a result, the measured overvaluation has dropped from 12% last February to the current 7%.
This next chart combines the two charts above with colors indicating varying risk levels (upside potential vs downside risk on Fair Value). This market is no bargain but it is not at “obscene” valuation levels based on known trailing data.
An environment where the economy keeps chugging along at 2.0-2.5% (3.5-4.0% nominal) and labor costs rise about 2.5% and inflation remains subdued would likely keep earnings and the Rule of 20 Fair Value improving, much like in 1962-66.
Here’s a close up view of recent trends:
The recent uptrend in the Rule of 20 Fair Value (yellow line above) has reduced the downside risk and is providing a stronger backwind to equities, potentially overcoming the high valuations, the economic uncertainties and the rather volatile political environment in the USA and the world, especially in the context of a fairly dovish Fed.
So when economist and strategist David Rosenberg, however famous and smart he can be, says “this is not an earnings-driven market, it is a momentum, liquidity, and multiple-driven market, pure and simple”, he is purely and simply wrong. Rosenberg uses NIPA profits to make his claim but equity markets don’t trade on NIPA profits, they trade on S&P 500 Index earnings and these troughed on a trailing 12-month basis exactly one year ago and, since then, are up 18.6% per S&P data (+21.4% GAAP EPS) and 7.9% per Thomson Reuters data.
The reality is that, even after such a long recovery cycle, there are no major imbalances that require really painful medicine from central banks. No economic bubbles, no financial bubbles to speak of (other than cryptocurrencies), no inflation bubble. Plain dull world, apart from the amazing entertainment from the White House. Mild growth, mild inflation, mild interest rates, mild cost pressures…
Perhaps mild markets as a result. Mild is not bad, especially when inflation and interest rates are so low.
Another money manager included this from John Hussman in its June 2017 letter to clients telling them “It is worthy of your careful consideration”:
Investors should recognize that in data since 1940 and prior to 2008, U.S. interest rates were at or below present levels about 15% of the time. During those periods, the average level of the Shiller cyclically adjusted P/E was about -50% below present levels, and the average ratios of MarketCap/GVA, MarketCap/GDP and Tobin’s Q (market capitalization to replacement cost of corporate assets) were all about -60% below present levels.
That’s roughly the same distance that current market valuations are from post-war pre-bubble norms, even regardless of the level of interest rates. Put simply, investors have vastly overstated the argument that low interest rates “justify” extreme market valuations. Indeed, the correlation between the two is weak, nonexistent, or goes entirely the wrong way in most periods of U.S. history outside of the inflation disinflation cycle from 1970 to 1998.
I called upon my old pal I. Bernobul to verify Hussman’s stats. As you will see, this money manager should also be careful before using Hussman’s stats in its client letters. Some clients could very well decide to carefully reconsider their manager…
Here’s the reality:
- 10Y Treasuries were effectively below 2.5% about 15% of the months between 1940 and 2007.
- What Hussman failed to disclose is that
- Virtually ALL of these months were prior to 1951.
- The Fed pegged U.S. interest rates between 1942 and 1951, well into the Korean war (1950-53).
- 57% of the months were during WWII.
- 27% were between 1946 and 1948 when inflation skyrocketed to 19.7%.
In effect, world politics, economies and financial markets did not function anywhere near normality through most of the period that Hussman uses to support his point that extremely low interest rates are no reason for higher valuations. Amazingly, Hussman tries to link the truly abnormal 1940-50 period with the current environment, arguing that “hey, we’ve been there before and, see, valuations were never so “extreme””, without even alluding to the highly unusual circumstances of 1940-50 where virtually all the relevant data points were located. He then presents a supportive chart which starts in …1950, avoiding to display the reality of the previous decade.
Yet, at one point in his piece, he declares “remember that Fed policy matters little once investors become averse to risk”. But between 1940 and 1950, investors had to deal with a second world war, huge swings in inflation, pegged interest rates with hugely negative rates most of the period, a severe recession in 1948-49 and the Korean war to top it all out. If you were not risk averse in those years, it was only because you were totally unconscious.
Meanwhile, the Rule of 20, using inflation rather than manipulated interest rates, performed very well as a barometer for equity valuations even though inflation was also manipulated through occasional price controls during the war. The Rule of 20 showed equities as seriously overvalued in March 1946 (Rule of 20 P/E at 22.3); the S&P 500 peaked in May 1946 and dropped 25% by February 1948 (Rule of 20 at 17.7). The Rule of 20 remained very undervalued up to March 1961 while the S&P 500 Index quintupled.
Equities are no bargain at this time and they could very well correct anytime. But they are nowhere displaying “extreme overvalued, overbought, overbullish syndromes” as Hussman keeps repeating. Bearish investment management organizations would be wise to find another mouthpiece.