Go to the Globe and Mail homepage

Jump to main navigationJump to main content

(Jupiterimages/(C) 2006 Mehmet G.Emin)
(Jupiterimages/(C) 2006 Mehmet G.Emin)

market lab

Recent economic data not so surprising, research shows Add to ...

A week removed from the Philly Fed That Shook the Markets, I have to wonder which was the bigger stunner: How wildly below expectations the regional manufacturing-activity index was, or how violently investors reacted to it?

Yes, the index was a massive disappointment to the downside. But to take a surprise in the Philly Fed as evidence that the U.S. economy is much weaker than expected would be to give the Philly Fed too much credit.

Research by Goldman Sachs shows that the index is among the least predictable of all U.S. economic indicators. Furthermore, the notion that this was “another in a string of disappointing economic data” simply isn’t accurate. Economic indicators haven’t been unusually surprising at all lately, Goldman Sachs found.


Goldman Sachs economist Zach Pandl examined the accuracy of consensus forecasts for a wide range of economic indicators. He compared the average error in each forecast (the difference between the forecast and the actual result) to a “naive benchmark” – which simply took the previous actual result in each indicator and used that as the “forecast” for the next result. The idea was to see how much the consensus forecasts improved upon, essentially, wild guesses.

( In this chart, the further the number is below 1.0, the more superior the consensus forecast was to the naive benchmark.)

Mr. Pandl found that for some key indicators for the market – such as factory orders, consumer price index and retail sales – the consensus forecasts added a great deal of reliable advance information. However, the various regional and national manufacturing activity surveys – including the Philly Fed, the Empire State index and the Institute for Supply Management’s manufacturing index – are among the least reliable.

That’s eye opening, especially given the market’s persistent focus in recent years on the manufacturing surveys. They may be useful indicators of the economic trend, but traders who try to anticipate them are playing a dangerous game.


Goldman’s “MAP” tool – a composite measure of surprises in a range of economic indicators, weighted for each indicator’s relative importance – also shows that despite perceptions to the contrary, recent economic indicators, overall, haven’t been wildly at odds with expectations.

“Recent surprises have been slightly larger than normal, but are not exceptionally large relative to history. Moreover, forecast errors have declined significantly since the recession,” Mr. Pandl said.

This suggests that the market has actually had a decent handle on the economic conditions as they have deteriorated; investors may be increasingly disappointed, but they have little call to be surprised. The reaction to Philly Fed was certainly an excuse for an overdue correction – but maybe it wasn’t really the trigger.

Report Typo/Error

Follow on Twitter: @ParkinsonGlobe

Next story




Most popular videos »

More from The Globe and Mail

Most popular