Gluskin-Sheff economist and hardcore sceptic of the rally David Rosenberg is out with a “special report” on the V-Shaped Recovery, and in his case the “V” stands for valuation, because:
…every basis point of this 60% rally in the U.S. equity market from the lows has been due to an
unprecedented expansion in P/E ratios. In fact, by some measures, the S&P 500 is already trading at valuation levels that would ordinarily be consistent with an economic expansion that is five-years old as opposed to a recovery that, at best, is in its infancy stages.
There has been plenty of debate over whether equities are overvalued or not, and certainly we would assume that many investors know where we stand on the topic. Let’s look at the facts now that the September data are in. On an operating (“scrubbed”) basis, the trailing P/E multiple on the S&P 500 has expanded a massive 10 points from the March lows, to stand at 27.6x. Historically, when the economy is taking the turn away from contraction towards expansion, which indeed was the case in Q3, the trailing P/E multiple is 15x or half what it is today (and that 15x is also calculated off depressed earnings level of prior recessions – we have more on the historical comparisons below). While we will not belabour the point, when all the write-downs are included, the trailing P/E on “reported” earnings just widened to its highest levels in recorded history of nearly 140x (see chart below), which is three times the levels prevailing during the height of the tech bubble.
It is interesting to hear market bulls talk about how distorted it is to be using trailing multiples that include ‘recession earnings’ (even though using ‘forward’ earnings means relying on consensus forecasts on the future and these are rarely, if ever, correct). It is also interesting that the last time the multiple was this high was back in March 2002, again after a huge countertrend rally that deployed ‘recession earnings’ from the 2001 downturn. If memory serves us correctly, this was right around the time that the bear market rally started to roll over and in fact, six months later, the S&P 500 was hitting new lows and 34% lower than it was when the multiple had expanded to … today’s level!
But the point is well taken that if in fact we are at an inflection point, moving out of recession and into expansion, looking strictly at multiples on depressed trailing earnings could be misleading. So let’s take a look at what valuations looked like at previous turning points in the cycle. For example, when therecession ended in November 2001, the trailing multiple was 29.3x, muchhigher than today indeed, which may be a reason why the market did not bottomfor nearly another year. It was too expensive.
At the end of the recession in March 1991, the trailing P/E multiple was 17.2x. In November 1982, it was 11.0x and in July 1980 it was 8.3x. When the recession ended in March 1975, the P/E ratio was 9.9x, and in November 1970, it was 17.0x. Now there is no doubt that the bulls would argue that the
multiples troughed at unusually low levels because inflation was extremely high. Point taken. But when we go back to the low-inflation periods of February 1961, the multiple at the recession trough was 20.5x, not 27.6x. At the April 1958 recession-end-point, the P/E was 14.8x, and in May 1954, it was 11.1x. The P/E was not 27.6x.
Here are some charts to emphasise his point. Here’s where PEs should be at the end of the recessions.
And obviously, if we went by pure trailing earnings (with all those massive losses) then PE ratios are insane.
But then, even if you look at forward earnings (which are always wrong) PEs are high.