In the latest installment of my Five Good Questions series (AdvisorOne version here), LPL strategist Jeff Kleintop offers a bullish near-term perspective on the equities market. “We continue to subscribe to the forecast we have held all year for modest single-digit gains for the stock market. We believe the gap between feelings and facts got too wide with pessimistic investors failing to notice solid economic and profit reports and progress dealing with the challenges in Europe.” There’s nothing wrong with such a view on its face, even though I am decidedly less sanguine – reasonable minds can differ and forecasting is always fraught with peril and error. Moreover, near-term forecasting is especially difficult.
But something he says later troubles me a lot.
I pointed out what I call “leading investment indicators” (PE10, dividend yields, Q, market cap-to-GDP, interest rates) that are very negative and asked how these measures impact his views. Jeff replied as follows: “We believe that history has made it clear that the most consistently accurate predictor of long-term stock market returns is the S&P 500 price-to-earnings ratio (P/E).” That reply makes sense coming from a bull because PE is fairly low today – suggesting healthy returns going forward (in Jeff’s words, “[t]he level of the P/E and the annualized return on stocks over the next 10 years have a very close relationship”).
That answer alone gives me pause because even though Jeff claims that P/E “has a nearly perfect track record of forecasting long-term performance,” solid academic research supports the view that P/E isn’t as good a predictor of future returns as PE10. As my friend Wade Pfau pointed out to me, that’s largely because earnings can be so volatile. Even so, PE10 can only explain about 30% of the movements in the subsequent 10-year real returns on the S&P 500. Even the best predictors are subject to randomness (noise). There is nothing like certainty in the markets. Note these charts below (from Wade).
I’m not terribly troubled by Jeff’s use of P/E per se (“the trailing PE has a nearly perfect track record of forecasting stock market performance over the next 10 years”) in that P/E does explain about 27% of the subsequent real 10-year total returns – that’s nothing like the claimed “nearly perfect,” but it’s noteworthy nonetheless. I’m much more troubled by Jeff’s outright rejection of PE10 (“the trailing PE has a nearly perfect track record of forecasting stock market performance over the next 10 years – the 10-year PE does not”) because the data does not support it.
As noted above, Jeff’s claim overstates the impact of P/E, but it’s also dead wrong on PE10, as shown above and as Wade demonstrates in a recent paper. Perhaps more significantly, the way Jeff explains PE10 makes it sound like he isn’t using the proper definition. PE10 is not a 10-year moving average of P/E values. Rather, it is the current stock price divided by the trailing 10-year average of real earnings. As Jeff defines it, “PE10” would indeed be an unlikely predictor of future returns. I’m also troubled that Jeff relies solely upon P/E and neglects other indicators (such as DY), perhaps because they do not support his desired result.
In another of Wade’s papers, he discusses PE10 and DY10 as reasonable predictors of future returns, but note how he defines them. DY10 is the 10-year moving average of dividend yields, but PE10 is current price divided by the 10-year moving average of real earnings. As Wade argues in his paper:
“The dividend yield (DY) is aggregate dividends divided by the stock price. I find that a 10-year moving average for the dividend yield (DY10) provides a better model fit. This can be justified as a way to obtain the underlying trend in dividend payments after removing the cyclical trend in stock prices. Unlike EY10 [this is just 100 divided by PE10], the 10-year moving average for DY is not the average of previous dividends over current price, but rather the average dividend yield. Because EY10 already includes the current price, another variable with current price is not needed.”
Using this definition, PE10 is a reasonable predictor of future returns (it’s the best we have) and, today, does not provide much solace for the bulls.1 This definitional difference may simply reflect a difference of opinion. However (and perhaps I’m being too cynical for my own good here), as a noteworthy new blog points out:
“You may have been under the assumption that analysts, economists and strategists are compensated to the extent they are correct in their forecasts. This is only true to the extent that accurate predictions generate trade commission. In the end, analysts, strategists and economists are paid directly in accordance to how much trading commission can be generated from their reports. Importantly, bearish news is not conducive to trade generation. Even in the rare instance where bearish news does instigate sell ticket commission, unlike buys a sell does not imply repeat business. Stocks that are bought need to be sold, with commission on both sides. The cash realized by sells may never return to the trade desk again.”
As every equities-jockey is well aware, it’s hard to make money being bearish. Thus LPL has an interest in being bullish irrespective of the data. Jeff’s forecast needs to be viewed in that light.
1 I hasten to emphasize that PE10 does not provide much solace for the bulls over the longer-term. Finding a near-term trading strategy using valuations has proven vexingly difficult. As Keynes put it, the market can remain irrational longer than you can remain solvent.