Wolf Richter sounds the alarm about the S&P 500's current price earnings ratio:

This chart shows those adjusted earnings per share for all S&P 500 companies (black line) and the S&P 500 index (blue line). I marked July 2012 and March 2014 (via FactSet, click to enlarge):

Given that there has been zero earnings growth over the past three years, even under the most optimistic “adjusted earnings” scenario, and only about 2% per year on average over the past five years, the S&P 500 companies are not high-growth companies. On average, they’re stagnating companies with stagnating earnings. And the price-earnings ratio for stagnating companies should be low. In 2012 it was around 15.5. Now, as of July 7, it is nearly 26.

In other words, earnings didn’t expand. The only thing that expanded was the multiple of those earnings to the share prices – the P/E ratio. Such periods of multiple expansion are common. They’re part of the stock market’s boom and bust cycle. And they’re invariably followed by periods of multiple contraction.

Before we go any further, let's acknowledge that Wolf would be absolutely right if earnings were the fundamental driver of stock prices.

Unfortunately, they're not. Instead, dividends serve that function (here's the math and several years of forecasting results to back that claim up).

From that perspective, the seeming inexplicable levitation of stock prices is completely understandable during this period of time. When Wolf looks at stagnating companies in the years between 2012 and 2017, there is perhaps no better example of the kind of performance he describes than oil producers, where beginning in mid-2014, with the crash of global oil prices, their earnings crashed right along with their revenues.

By Wolf's reasoning, there would be no reason for the stock prices of these companies to continue to remain elevated at the levels they reached while they were riding high on the profits from high priced oil. But the boards of directors and management teams at most of the largest of these firms took a risk - they acted forcefully to cut their costs at rates faster than their revenue was falling. And in doing so, they bet that they could ride out the collapse of oil prices.

So even as their earnings crashed, they were able to maintain the cash flow needed to continue making their dividend payments at the same levels they were before the bottom dropped out of their revenues. Their stock prices responded by... not changing very much, keeping more in step with their dividends and not acting anything like what happened to their earnings, which by our theory of how stock prices work, is exactly what they should have done.

But because their earnings crashed, their P/E ratios soared, which is why many capable market observers like Wolf are concerned (emphasis ours).

How long can this period of multiple expansion go on? That’s what everyone wants to know. Projections include “forever.” But “forever” doesn't exist in the stock market. The next segment of the cycle is a multiple contraction.

The 10-year average P/E ratio, using once again the inflated “adjusted earnings,” not earnings under GAAP, is 16.7, according to FactSet. This includes two big stock market bubbles, the one leading up to the Financial Crisis, and the current one, but it includes only one crash. This imbalance skews the results. Two complete cycles would bring the average substantially below 16.7.

Nevertheless, even getting back to a P/E ratio of 16.7 for the S&P 500, when the current PE ratio is 25.6, would signify either miraculously skyrocketing earnings or a sharp contraction of the market. The first option is a near impossibility. And the second option?

Markets overshoot, which is what reversion to the mean entails: the average isn't going to be the floor! And that’s why this type of unsustainably high earnings-multiple is like a tsunami siren where the arrival time of the tsunami remains unknown – and that’s why it is ignored until it’s too late.

Under our theory of how stock prices work, the kind of multiple contraction that Wolf anticipates would happen if, and only if, the financial situation of the most distressed industries of the S&P 500 were to deteriorate to the point where they could no longer avoid dividend cuts. And then, we could see a situation similar to what we saw back in late 2008 and early 2009, provided that the affected companies would sequentially follow each other in taking that action.

Outside of that happening, any major or sudden changes in stock prices would be described as a kind of quantum random walk, where in addition to their expectations for future dividends, the volatility in stock prices is affected by how far forward in time investors are looking.

In all this, you'll notice that we didn't put any weight on the value of P/E ratios for telling us anything actionable or worthwhile about the state of the stock market. We do have some thoughts on how to make the ratio a better indicator of the market's relative valuation, but we're afraid that you'll have to stop thinking horizontally about it.

More on that soon!