Advertisement

July 16, 2004

News & Opinion: Beating the Business Cycle - Our Method

By: Dylan Schleicher @ 4:07 PM – Filed under: Current Events & Public Affairs

The Challenge


Economic forecasting deserves its bad reputation in predicting recessions and recoveries. As a recent 63-country IMF study concluded, "The record of failure to predict recessions is virtually unblemished."
The reason for this failure is simple. Most economists forecast by extrapolating economic trends. While this methodology has its merits, predicting turning points is not one of them. Why is that so?
One key reason is that these forecasting models assume the recent past to be a good guide to the near future (see chart below). Most of the time this is a good assumption. But as we approach economic turning points, by definition, the pattern changes. The gap between such forecasts and reality balloons, resulting in large forecast errors.
bizcyc1.gif

A Better Way


There is a better way. By using ECRIs approach, which focuses on the right cyclical indicators, it is possible to tell clearly and objectively when a turn in the cycle lies ahead.
Unlike econometric models that project from past trends, these cyclical indicators are specifically designed to predict future changes in the direction of the economy. They turn before the economy does. The focus is on the timing of a change in direction (see chart below).
bizcyc2.gif
But even the best leading indicators sometimes give conflicting signals. The key is to glean from their collective wisdom the signal that the economy is headed for a turn. This is why ECRI develops composite indexes of leading indicators using state-of-the-art proprietary techniques.
However, even a single composite index is not enough for the purpose. Therefore, ECRI uses an array of 14 such specialized indexes to monitor the U.S. economy alone, and over a hundred proprietary indexes to monitor the global economy.
This brings up another point. There is a lot of data out there. A data vendor we use maintains a database of 4,000,000 time series, but most of them are useless for forecasting cycles. We use maybe 400 series to monitor the U.S. and most major international economies thats 0.01% of the data available. Still 400 series is a lot, and many times some individual indicators are rising while others are falling. Leading indexes like the Weekly Leading Index (WLI) sum up all the on the one hand this is rising, and on the other hand this is falling and provide a simple direction that we should expect the economy to follow in coming months.

About Dylan Schleicher


Dylan Schleicher has been a part of the 800-CEO-READ claque since 2003. Even though he's stayed on at the company, he has not stayed put. After beginning in shipping & receiving, he joined customer service and accounting before moving into his current, highly elliptical orbit of duties overseeing the ChangeThis and In the Books websites, the company's annual review of books and in-house design. He lives with his wife and two children in the Washington Heights neighborhood on Milwaukee's West Side.