ECRI Testimonials

Investech Research: April 2005

"In our opinion, ECRI is one of the extremely few reliable economic research firms in the country. And we have met the principals of the firm personally, including the founder, the late Geoffrey H. Moore, who was Alan Greenspan's professor at NYU. ECRI’s warnings prior to the 2001 and 1990 recessions stand out from the crowded economic field, where most participants follow only trends (or each other) in making forecasts."

The Economist: January 2005

Divining the Future

Embarrassingly, conventional economic forecasts have rarely correctly predicted a recession.

Some economists argue that recessions are typically caused by unpredictable shocks, such as the events of September 11th 2001 or a sudden rise in oil prices. But that excuse will not do, at least for 2001: official figures show that the recession started months before the September attacks. The trouble is not the unpredictability of events, but that the economic models on which forecasts are based are not up to the job. Despite containing hundreds of equations, models are notoriously bad at predicting recessions because they tend to extrapolate the past, for most of which the economy has been expanding. Yet recessions stem from abrupt changes in the behaviour of firms and consumers.

Scrutinising indicators that have given advance warning of downturns in the past, such as share prices, order books or the shape of the yield curve, seems more helpful than relying on models. There is no single magic measure; each leading indicator has failed to predict a turning point at one time or another. But leading indicators are unlikely all to give false signals at the same time, so a composite index that combines several economic and financial measures stands a good chance of spotting changes in the economic weather.

In a survey in March 2001, 95% of American economists said there would not be a recession. One of the few exceptions was the Economic Cycle Research Institute (ECRI), an independent research firm, which that same month correctly forecast, on the basis of its leading economic indicators, that a recession was unavoidable. ECRI was set up by the late Geoffrey Moore, an early pioneer of business-cycle research who in the 1960s developed the American government's first leading indicator (which is now published by the Conference Board).

Prophet maximisation

ECRI tracks around a dozen leading indices for different parts of the economy. Some of these have a longer lead time relative to economic activity than others, so the firm publishes both a longer leading index, which signals changes in activity about a year in advance, and a shorter leading index, which looks six months ahead. This is one advantage over the Conference Board's leading economic indicator, which conceals useful information by combining longer- and shorter-term indicators in a single measure.

A second advantage is that ECRI publishes a weekly index and is thus always up to date. Moreover, this weekly index largely comprises indicators that are not revised later, such as share prices, bond yields and commodity prices. In contrast, the OECD's and the Conference Board's current historical time series differ from what was published at the time, because the data from which they are constructed have been revised. As a result, the indicators' predictive power is not as good as the revised series imply.

ECRI is perhaps the only organisation to give advance warning of each of the past three recessions; just as impressive, it has never issued a false alarm. In their book Beating the Business Cycle, Lakshman Achuthan and Anirvan Banerji (both of ECRI) claim that the firm's weekly leading index has consistently predicted downturns before the Conference Board's measure. Indeed, in April 2001, after ECRI had called a recession, the Conference Board said: “No recession is on the horizon.” ECRI's indices for other economies have also fared well. For instance, it correctly forecast the Japanese recessions that started in 1997 and 2001.

ECRI has no simple rule for determining when the economy is headed for recession. It compares recent movements in its indices with previous business cycles. In order to signal a genuine turn in the cycle the indices must change direction in a way that is pronounced in size, pervasive across individual components and persistent. For example, the weekly leading index fell sharply in 1987 but no recession was called. This was because the decline was not pervasive, but due almost entirely to that year's stockmarket crash.

So how should one interpret the plunge in the past year in all three series of leading indicators? When America's growth weakened in mid-2004, the economy was widely thought to be in a soft patch that would soon pass. The persistent fall in the leading indicators suggests that growth could remain disappointing in the first half of this year. The more timely and (so far) more reliable ECRI index, however, rose in both November and December, tentatively suggesting that the economic slowdown could end by around the middle of this year. A new recession is certainly nowhere in sight. But keep checking those leading economic indicators over the coming months.