Over three generations of business cycle research, we have helped to advance the understanding of business cycle dynamics, some of which we have shared publicly.
The excerpts and papers collected here reflect a sampling of important concepts pioneered by our research group.
March 2012 | by ECRI
The convergence of two cyclical patterns virtually dictates an era of more frequent recessions in developed economies. As a result, and because of the Bullwhip Effect, growth in developing economies is going to be jerked around more than people think, making for a good deal of cyclical economic contagion. In other words, we are now in the yo-yo years.
March 2011 | by ECRI
Alan Greenspan accepts ECRI's long-held criticism that the Fed is chronically behind the curve because of its reliance on core inflation and the output gap. But he is wrong that no indicator can predict when inflation is about to take hold.
March 2010 | by ECRI
The convergence of lower trend growth and higher cyclical volatility will lead to more frequent recessions, keeping the jobless rate cycling around high levels and spelling the death of buy-and-hold strategies for stocks.
October 2009 | by ECRI
Selling (buying) stocks before recessions (recoveries) based on ECRI's real-time calls would have doubled the returns from a buy-and-hold strategy, beating the S&P by more than eight percentage points a year over the past decade.
June 2011 | by ECRI
The yield spread's popularity is due to its "success" in predicting U.S. recessions. Based on ECRI's international recession dates, we find it to be an unreliable predictor of international recessions - especially with rates at zero.
November 2001 | by ECRI
By the turn of the century, many were proclaiming the death of the business cycle. But risk has returned. Because technology and globalization can both reduce and increase risk, both economies and markets will stay volatile.
March 1999 | by ECRI
Because leading indexes are intended only to forecast the timing of cycle turning points, they should not be evaluated on the basis of standard parametric statistics like R-squares. We suggest an alternative, nonparametric approach.