More or less than last year?

The Holiday Sales Forecasting Fools

Barry Ritholtz is a Bloomberg View columnist. He founded Ritholtz Wealth Management and was chief executive and director of equity research at FusionIQ, a quantitative research firm. He blogs at the Big Picture and is the author of “Bailout Nation: How Greed and Easy Money Corrupted Wall Street and Shook the World Economy.”
Read More.
a | A

Forget August. The real silly season is upon us.

As evidence I present the forecast of a 13 percent increase in holiday sales made by Forrester Research, ascitedin the New York Times this morning. Color me skeptical.

As we noted about this time last year, these sales projections are as much of an annual holiday ritual as Thanksgiving and overeating. Every year, numerous trade groups, research firms and other assorted media shills release their holiday sales forecasts. The data set of failures is so large that even if anyone gets it right, it should be chalked to random chance. That is what happens when you are so wrong for so long and don't bother to think about why you are wrong.

I am glad that Forrester has the ability to forecast a 13 percent jump in this year’s holiday shopping. To reach that conclusion, it only needed to figure out:

• How much income people will earn between now and the holidays, minus their actual expenses, plus their availability of credit.

• The psychology of how much Americans will actually be spending money on gifts.

• The weather.

• Any other random events -- ISIS, Ebola, Taylor Swift -- that may interfere with holiday cheer.

• Knowing how much savings Americans have accumulated during the year.

I am only kidding about that last one, because these are Americans we are talking about, and theirsavings rate could substitute as a slide at a water park.

Forrester's forecast is but one of the many acts of idiocy we have coming up in November. We will also be regaled with the silliness of the National Retail Federation’s Black Friday surveys, an exercise in innumeracy, flawed statistical methodology and foolishness. ShopperTrak will count shopping-mall foot traffic, and from that it will somehow extrapolate holiday spending.

All of the above have track records that make it clear Nostradamus won’t be getting kicked out of the Forecasting Hall of Fame anytime soon.

Why are these retail sales predictions so bad? Start with the simple reality that the future is inherently unknowable. Add to that a series of very, very bad methodologies. Experience has repeatedly taught us that surveys where people forecast their own future spending are inaccurate.

If we can't even forecast our own behavior -- which a wealth of academic literature and real-world experience confirms -- then how can we turn that sort of survey into a worthwhile forecast of national economic activity? Let me save you the trouble of answering that: We can’t.

Add to the mix a news media that is hungry for content of any sort, regardless of how inaccurate or meaningless it might be. The net results are lots of useless or even misleading filler on retail sales between now and the new year.

There arepockets of skepticismfinally appearing. I am not the only one who has noticed that these predictions are consistently wrong.

The best approach I have seen is the credit-card company reports based on actual customer usage. Visa, MasterCard and American Express all release various reports that have at the center of their methodologies the actual spending of Americans -- at least their credit-card spending.

Even these have some issues, but they are relatively minor compared with the aforementioned nonsense. Internet shopping is almost all done by credit card, and must be seen as a shift in spending habits, not an increase in spending. There also is the issue of cash-crimped shoppers who might use their credit cards more but actually spend less. Other anomalies might mean that the models developed by the credit-cards companies are imperfect. And they are, at best, coincidental or even lagging indicators of actual holiday spending.

Regardless, they are based on real spending by actual shoppers who are buying goods. That is a lot more than I can say for the rest of the surveys. I won’t say these other approaches are worthless; at the very least, they are all good for a laugh when the actual retail data is released in January. By then almost everyone will have forgotten about the silly measures of surveys and foot traffic.

Except for this column. Be sure to check back then to see how well the usual suspects did this holiday season.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

To contact the author on this story:
Barry L Ritholtz at britholtz3@bloomberg.net

To contact the editor on this story:
James Greiff at jgreiff@bloomberg.net