The First Flake

Story Highlights:

• Perhaps guided by instinct, investors tend to rely heavily on monthly government statistics.
• Most economic indicators are not only flawed, myopic, and backward-looking, but they're far from all-encompassing.
• Monthly unemployment data (or any other monthly government statistic or survey) is not predictive for stocks and never will be.

______________________________________________________________________

There's something foreboding in the autumn air—a trace of bittersweet chimney smoke, a subtle chill riding amid fading summer smells. We nervously lift our noses and sniff winter's waking, wondering when the winds will bring snow and for how long. Instinctual creatures that we are, we do the same thing when investing. Just observe the flurry of headlines greeting today's economic data.

This morning many assumed the latest stock prices reflected recently released employment data, and many warned of winter's onset. After all, markets were down Thursday and continued to fall unchecked Friday morning. Yet after a big morning drop, the S&P 500 recovered to finish the day up +0.4%.

How can this be? How were markets able to shake off the morning's dire news so blithely? Simply, monthly government statistics don't predict market direction. It's as true for long time periods as it is for daily price moves. We've dedicated plenty of space in these pages counseling not to rely too heavily on government numbers (especially the monthly ones). They're founded on many debatable assumptions and the preliminary reports often depend more on estimates than actual data. Hence, we've seen many a revision down the road.

Additionally, so-called economic indicators are backward-looking. It's faulty to believe last month's behavior predicts next month's returns. Or tomorrow's or even this morning's (clearly!). Employment is one of the most backward-looking of all. Job cuts or additions are typically the last thing to trickle through the system, whether the economy's moving forward or faltering.

Most economic indicators are not only flawed, myopic, and backward-looking, but they're far from all-encompassing. The premise is employment on its own can tell us what to expect for spending or general economic activity. But if employment (a single indicator) determines general economic activity, how can we explain personal income and GDP growth over the same period employment's been falling?

Some might argue the reason stocks recovered Friday was because the jobs data had been priced in already. Maybe so. But that's just more evidence such indicators are useless to stock investors. The employment data either had no bearing whatsoever on markets or was priced in before it was ever released.

We considered dedicating this commentary to why the unemployment uptick to 6.1% isn't so bad. (And it's not—just a bit above its long-term average, even counting a couple decades of low unemployment.) But then we asked ourselves the critical question, "Is this number at all predictive for stocks?" And there's just no way around it: Monthly unemployment data (or any other monthly government statistic or survey) are not predictive for stocks and never will be. They're more like sniffing for winter than finding the first flake.