It took me a long time to come to the understanding that the economy and the stock market are only second cousins, that they’re not twin brothers as most people tend to believe.
Since being taught this and shown the definitive evidence, I’ve devoured papers and studies on the subject and have seen this truth borne out from a variety of different research methodologies. I’ve also freed up much of my time that used to be spent reading global macro newsletter BS. While many of these newsletters contain several interesting observations and statistics, almost every one of them then goes on to form the wrong conclusions or an uninvestable premise.
Economic forecasts are not terribly helpful for the purposes of asset allocation because they do not (cannot!) allow for the fact that human market participants will not necessarily react as we expect they will. Sentiment and psychology and politics and natural disasters cannot be accurately and mathematically represented in any forecasting model and everybody knows it.
Paul Lim casts a spotlight on this perennial disconnect this weekend with a column at the New York Times that cites a key Vanguard Group report.
It should be noted that Vanguard does have a particular bias – they are the anti-active management shop and the idea that markets and asset classes can be timed is anathema within their vaunted Valley Forge castle keep and mead halls.
But! But their take on the predictive power of various economic gauges and market valuation metrics is not to be ignored. If you only learn one thing today, let it be this (emphasis daddy’s):
Roger Aliaga-Díaz, senior economist at the Vanguard Group, says investors shouldn’t be surprised about the seeming disconnect between basic economic variables and stock market performance. He and his colleagues at Vanguard recently studied equities’ returns going back to 1926, looking specifically at the predictive power of important variables. Those include market price-to-earnings ratios, growth in gross domestic product and corporate profits, consensus forecasts for gross domestic product and earnings growth, past stock market returns, dividend yields, interest rates on 10-year Treasury securities, and government debt as a percentage of G.D.P.
Their conclusion was that none of these factors — which investors often cite when explaining their moves — come remotely close to forecasting accurately how stocks will perform in the coming year. “One-year forecasts of the market are practically meaningless,” Mr. Aliaga-Díaz says.
Even over a 10-year time horizon, considered by many investors to be long term, only P/E ratios had a meaningful predictive quality. Since 1926, those ratios, based on 10 years of averaged earnings — a gauge popularized by the Yale economist Robert J. Shiller — explained roughly 43 percent of stocks’ performance over the following decade.
Of course, “that means about 55 percent of the ups and downs in the market can’t be explained by valuations,” Mr. Aliaga-Díaz says. What about economic fundamentals like G.D.P. and corporate earnings growth? Over the course of a decade, those factors had even less predictive power over future returns.
No long-term efficacy and no short-term efficacy. And yet we obsess over these things on a daily basis, sometimes making huge changes to our portfolios and even our business plans as a result.
At my shop, we watch weekly charts and base decisions on monthly charts. We look at trends and rolling multi-month moving averages in economic reports as opposed to single-variant data points. I’m not saying this approach is always right or is certain to do better than any other approach. What I am saying is that it keeps us sane and it enables us to ignore so much of what distracts so many others – which I believe will have more positive effects than negative effects over the course of a career.
How can it not?