Tuesday, April 13, 2010

The Economic Nostradamus

All forecasts are wrong, some are useful. Despite the uncertainty surrounding forecasts, news media remains obsessed with the ranking of forecasts and forecasters. These rankings obscure the fine point that forecasts are, not can be, wrong and long-term forecasts are more wrong than the sort-term forecasts.

Understanding Economic Forecasts (David F. Hendry and Neil R. Ericsson editors, MIT Press, 2001) lists numerous accounts of how guesswork may also beat sophisticated econometric forecasts!

Below is another ranking of forecasters that has Dr. Sherry Cooper’s team at Nesbitt Burns mentioned as well.

EconAlpha's Staples is man of mystery, and a winner

Consistency lands him on top of forecasting mountain

By Rex Nutting, MarketWatch

WASHINGTON (MarketWatch) -- One of the best economic forecasters is a guy who's never quoted in the newspapers and never yells on CNBC. He just gets the numbers right -- at least more right than anyone else.

You've probably never heard of Spencer Staples, the owner, top economist, and chief bottle washer for a relatively new economic research firm, EconAlpha. Staples is the winner of the March Forecaster of the Month award from MarketWatch, beating out 43 other forecasters.

Indicators Forecast Actual
ism 56.90% 56.50%
nonfarm payrolls -79,000 -36,000
trade gap -$39.9 bln -$37.3bln
retail sales -0.30% 0.30%
housing starts 562,000 575,000
industrial production -0.30% 0.10%
consumer price index 0.00% 0.00%
durable-goods orders 0.90% 0.50%
new home sales 310,000 308,000
consumer confidence 51 52.5

Not only did Staples win the monthly award, he's also now the top forecaster in our contest over the past 12 months, knocking 2009's Forecasters of the Year Nigel Gault and Brian Bethune of IHS Global Insight out of the top spot.

Staples won the March award based on his predictions on 10 top economic indicators released in the month, including the Institute for Supply Management index, the consumer price index and new-home sales. On six of the numbers, his predictions were among the 10 most accurate.

Staples earned his undergraduate degree at Colby College in Maine, and got his career started at Bear Stearns. He went to the United Kingdom 20 years ago and worked for Mizuho Corporate Bank and an investment management company. Along the way, he earned an MBA from the Henley Business School.

He started his own firm in 2008, hoping to provide institutional investors with more consistent forecasts of the high-frequency economic data, the numbers that are released every week or every month, the numbers that tend to move financial markets the most.

Staples says forecasting the high-frequency data takes fundamentally different skills than forecasting over a longer period of time. Long-term forecasters try to eliminate the statistical noise caused by one-time aberrations in the data, such as weather, the timing of holidays and the like. They want to know the trend.

On the other hand, high-frequency forecasters embrace the noise.

Staples takes this further than do many of his colleagues. He flatly refuses to talk about the longer-term trends.

"My strength is in the high-frequency area," Staples said in a phone interview.

Staples says he approaches forecasting as an information-management problem. With so much information now available about every aspect of the economy, the trick is to know what to trust and what to ignore, he said. Asked for an example, he laughed, saying he didn't want to give away his secrets.

Indeed, Staples provides us with less explanation of his forecasts than anyone else in our contest.

Instead of writing hundreds or even thousands of words to explain his forecasts and what the implications might be, as many of his peers do, Staples simply emails us the bare essentials, like this: "February Non-Farm Payroll is expected to be -79K, with the Unemployment Rate at 9.8%." Everything else is "intentionally left blank." It's for the eyes of his clients only.

Staples is convinced that his clients can make money off forecasts that are consistently accurate.

"I think there is alpha imbedded in some forecasts," he said. Normally, investors think of short-term deviations from the trend as volatility, or beta. But Staples suggests that, for some market participants, a good forecast of those deviations can be like a regular dividend check: It's alpha.

This past year, Staples' forecasts have been plenty consistent. No one can get all the numbers right, but he's been more consistent than most. He's finished in the top 15 in our contest in nine of the past 12 months, more times than anyone else.

The runners up in the March contest were Peter D'Antonio of Citigroup Global Markets, Sherry Cooper's team at BMO Capital Markets, Michelle Girard and Omair Sharif of RBS Securities, and tied for fifth: Avery Shenfeld and Meny Grauman of CIBC World Markets and Ethan Harris's team at Bank of America Merrill Lynch.

The median forecasts that MarketWatch publishes each week in the Economic Calendar come from the forecasts of the 10 economists who've scored the highest in our contest over the past 12 months, as well as the forecasts of the most recent winner.

Over the past year, the top economists are, in order: Staples of EconAlpha; Bethune and Gault at IHS Global Insight; Maury Harris of UBS; the RBS team formerly headed by Stephen Stanley; Ethan Harris of Merrill Lynch; D'Antonio of Citi; David Resler of Nomura Securities; John Silvia's team at Wells Fargo Securities; Lou Crandall of Wrightson ICAP; and independent economic consultant Brian Jones.

No comments:

Post a Comment