The Market Isn’t Cheap
We frequently run across assertions that the US stock market is allegedly “cheap,” because the trailing P/E of the S&P 500 Index has not yet reached the dizzy heights of 1929 or 2000 (of course, quite often the “forward P/E” is cited rather than the trailing P/E. We believe this to be a worthless indicator, as it relies on overoptimistic analyst estimates that are continually revised lower as time passes).
Apart from the fact that the valuation peaks of the two biggest stock market manias in history hardly represent a useful yardstick for determining whether the market is attractively valued or not, these assertions are focused on an index that is capitalization weighted and the average P/E of which is greatly influenced by a small number of momentum stocks. In 2000, the extreme valuations accorded to technology stocks ended up producing a trailing P/E for the S&P 500 in cloud cuckoo land – however, the market as a whole was actually noticeably cheaper than it is today.
We have already mentioned a few times in the past that the market has never sported a higher valuation in terms price/sales as well as in terms of the median P/E. Dartmouth University Professor Kenneth R. French has calculated the US stock market’s median P/E. We were previously unable to find a chart for it, but have now come across a chart that was published by Bloomberg a little while ago, which we reproduce below. Note that the median valuation is even higher at the moment, as the calculation is apparently only performed once a year in the summer months. So the chart depicts the situation as of July 2014.
Only companies reporting a profit are included, so it would be more precise to state that this is the median P/E of the market segment with positive earnings. For comparison purposes, the S&P’s P/E ratio is shown as well. What is interesting about this is that the median P/E is currently actually higher than that of the S&P 500 on a trailing basis – something that has never happened before.
No Place to Hide
This has actually an important implication; it means that when the next bear market begins, there will be only very few places to hide within the market. For instance, when the S&P 500 reached the nosebleed valuations of the years 1999-2000, a great many “value stocks” had been immersed in a bear market for between two to three years already and were trading at extremely low valuations relative to growth stocks. Hence, it was possible for fund managers to rotate from overvalued tech stocks into a broad range of undervalued stocks in other sectors, such as for e.g. commodity producers, homebuilders and gold miners. We can therefore expect the next bear market to be an equal opportunity massacre (gold and coal stocks appear to offer some potential for rotation, but the number of relatively cheap stocks is overall far smaller than in 2000, and these two sectors are too small for a large-scale redeployment of funds).
Note as an aside here that whether and when the market will go up or down does only tangentially depend on its valuation. For instance, the market can sport a high valuation at the bottom of a recession due to a temporary cyclical decline in profits and nevertheless be a good buy. This is why it usually makes sense to look at the Shiller-P/E as well, which tends to smooth such cyclical fluctuations out. Also, a market that is clearly overvalued after a long cyclical bull market and the associated expansion in earnings multiples, can easily become even more overvalued. So we are not making any assertions here with respect to the timing of the eventual trend change. Our main points are only those made above, namely that the market is expensive and there will only be very few places to hide once the trend turns down.
S&P P/E ratio vs. US median stock P/E ratio since 1951. The market has never been more overvalued in terms of the latter.
In the context of valuations, it remains to be noted that earnings per share of many of the big cap companies in the S&P 500 have been manipulated higher with the help of share buybacks. The problem with this is that it has masked the rather mediocre underlying business performance and endangers future returns, as capital investment has been neglected and many companies have loaded up on debt to fund these buybacks.
Corporate profits relative to GDP, as well as profit margins are well above their previous cyclical highs as well (see John Hussman’s discussion of the topic here). Since these tend to be mean reverting, the market is in reality even more overvalued than it appears. Note that this is true even if overseas profits are excluded from the calculation, i.e., if one looks merely at the net profit margins of US domestic corporate business:
Profit margins are extremely stretched historically even when excluding corporate earnings made overseas.
Lastly, the market is also very overvalued in terms of Tobin’s Q ratio (which compares the valuation of corporate assets in the stock market relative to their replacement value). There are several ways of calculating Tonin’s Q, resp. different ways of looking at it, as Doug Short explains here. However, regardless of which method is employed, the market has only been more overvalued in terms of the Q ratio in the period between late 1998 and the year 2000 peak. Even the 1929 top is positively dwarfed by today’s valuation. Currently the Q ratio stands 80% above its arithmetic mean and 96% above its geometric mean:
Tobin’s Q. In this version of the ratio, intellectual property is excluded in the calculation on the grounds that one company’s competitive advantage due to intellectual capital is another’s disadvantage, i.e., IP is considered a zero-sum game.
There are certainly valid reasons bulls can cite as to why they expect the market to continue to rise. The most important one is undoubtedly that loose monetary policies across the world continue to pump up the money supply at a frenetic pace (as we have recently pointed out, money supply growth in the euro area has recently accelerated to almost 10% p.a., while in the US, it remains close to a very hefty 8% p.a.).
However, the notion that the market is cheap certainly makes no sense. The opposite is actually the case, and the wide dispersion of overvaluation makes the market potentially even more dangerous than it was during previous asset bubbles.