Noise Vs Signal: Stock Market Drivers, Short and Long-Term

I found myself, on two separate occasions today, telling clients that while I won't say that this is the most headline-driven market we've ever had to manage through, I will say, that, indeed, it is as headline-driven as any market we've ever had to manage through.

The following from Bloomberg's "Macro Man" Cameron Crise's latest podcast pretty much says it all, at this juncture:

emphasis mine

"... the well known Ben Graham saying, "in the short run the market is a voting machine, whereas in the long run it's a weighing machine."  Put another way, while economic and earnings fundamentals represent an investible signal over long periods of time, in the short run they get drowned out by a cacophony of noise. That's all the more so in today's market."

"If we look at the cumulative change in the S&P since 1954, 84% of that can be explained by a rise in index earnings (the weighted corporate earnings of its members in the aggregate), with the remaining 16% being explained by expansion in the p/e (price to earnings) multiple -- something like 9 in 1954 to 24 today.  Unsurprisingly,  earnings explain most of the long-term appreciation of the market.

If we look at the dispersion of index changes across various geographies over the last 20 years, we find a pretty solid positive correlation between those relative changes and relative changes in index level earnings growth.

The short-term however is something very different. Again, using days since 1954, I looked at the monthly change in the S&P and disaggregated the attribution in monthly changes in the p/e ratio on the one hand, and monthly changes in trailing earnings on the other. And what I found is that something like 98% of the monthly change in the S&P is a reflection of the change in the p/e ratio, which again we're calling, for the most part, the noise of the market voting machine. The signal of the market weighing machine of earnings really only explains 2% of the monthly price action."

"I ran the same analysis looking at the year-on-year change in the S&P versus the year-on-year change in the multiple and the year-on-year change in trailing earnings. And what I found is that the multiple (the noise), even on a year-on-year basis, on a median level since 1954, explains 68% of the change in the market.

So even over a 12-month horizon noise can dominate signal. I guess that shouldn't come as a huge surprise because even if we look at some of the more egregious bubbles in the market's memory, those often take years to build up and inflate, reflecting the noise inherent in market sentiment."

So what's our takeaway?  

As I continue to express in our video updates, while we remain open to all possibilities, per my illustrations, sustainable bull markets typically don't emerge from p/e, and other, valuation ratios resembling the present... Which of course -- in a noise-driven market -- in no way means that we won't see fresh all time highs in stock prices before the market finally succumbs to present fundamental reality. Or, on the other hand, before present fundamental reality rises to meet the present level of stock prices.

Of course we'll continue to explore the possibilities, while assessing the probabilities, here and in future videos -- and invest accordingly.

Stay tuned.