In Memoriam: Michael J. Dueker [Review (Federal Reserve Bank of St. Louis) (MO)]
(Review (Federal Reserve Bank of St. Louis) (MO) Via Acquire Media NewsEdge) Our friend and former colleague, Michael J. Dueker, passed away on Wednesday, January 29, 2014. Mike was a well-respected economist and policy advisor, a sought-after colleague and coauthor, and a close friend of many current and former employees of the St. Louis Fed.
Mike joined the Research Division of the Federal Reserve Bank of St. Louis as an economist in 1991, following his graduation from the University of Washington, Seattle, where he earned a Ph.D. in economics. At the Bank, Mike was promoted three times-to senior economist, research officer, and assistant vice president-in recognition of his strong performance as a research economist and policy advisor. Mike left the Bank in 2008 to return to the Seattle area to work for Russell Investments. He then also became a regular contributor to the Blue Chip Economic Indicators panel of professional forecasters. Mike remained a good friend of the St. Louis Fed after leaving for Seattle, returning for several short visits to participate in Bank conferences and to work with coauthors.
Mike was hired by Anatol B. ("Ted") Balbach, who was the St. Louis Feds director of research at the time. Mike was a well-trained econometrician who had demonstrated strong technical skills, particularly in econometrics and statistics, in his Ph.D. dissertation. However, Ted was concerned whether Mike would make a good economist. "I know Mike is a good metrician,"' Ted said, "but I don't know about the ecori part-is he a good economist?" Mike clearly was the most talented of the new Ph.D.'s that the St. Louis Fed interviewed that year, however, so Ted was willing to take a chance on Mike. Ted was not disappointed.
Mike's research focused on developing new econometric methods and applying those methods to important economic policy questions. He was not an economic theorist, but he had a deep understanding of economics and knowledge of financial markets. Mike shared generously with his colleagues. In seminars, Mike asked penetrating questions. Frequently, he offered constructive comments and criticism that helped his colleagues build better economic models. Mike really was the economist that Ted hoped he would be, and then some.
Mike's understanding of economics, strong technical skills, and friendly manner made him a sought-after colleague and coauthor. He published over 50 articles on a broad range of topics. As of January 2014, Mike ranked among the top 5 percent of economists worldwide in terms of number of publications and measures of publication impact. Mike had coauthors all over the world. Early in his career, he spent a year at the Swiss National Bank, where he gained insights into the implementation of monetary policy in different countries and developed close working relationships with the Bank's economists that led to several published articles. He also wrote with several of his St. Louis Fed colleagues on a wide range of topics, including the impact of price-level shocks on financial stability, Federal Reserve actions to smooth interest rates, and excess returns in foreign exchange market. Mike continued to do research and write with coauthors from the St. Louis Fed and elsewhere after he moved to Seattle in 2008.
Many of Mike's papers were highly technical, but he always found important applications. In one of his most highly cited works, Mike found that the difference between the yields on longand short-term government securities is a good predictor of recessions.1 Mike built on this research to develop methods of classifying and forecasting the business cycle and other variables with discrete outcomes. Discrete outcome models parse the business cycle into phases-say, expansion and recession. One of Mikes interests was in forecasting transitions from expansion to recession using currently available data. He recognized that output growth, employment, and other economic and financial market variables forecast business cycle transitions, while the behavior of those variables in turn reflected the current phase of the cycle. Mike developed a model he called the Qual-VAR that combines discrete variables such as the business cycle phase with the commonly used vector autoregression (VAR). The Qual-VAR led Mike to construct a business cycle index from which he assessed the probability that the economy would enter recession at some given date in the future.2
Mike also studied monetary policy and financial markets, first by modeling volatility in financial markets and then by identifying trends in foreign exchange rates. He also studied how Federal Reserve discount rate changes affected market rates, the setting of federal funds rate targets, and how explicit inflation targets might improve economic performance.
Mike had a particular interest in modeling the behavior of variables, such as central bank policy rates and commercial bank prime lending rates, that change by discrete amounts. Mike proposed a methodology that both greatly simplified estimation of time-dependent classification models (such as the "dynamic probit" model) and incorporated features common to financial time-series models (autoregressive conditional heteroskedasticity). He applied the model to study the behavior of commercial bank prime lending rates and showed that periods of high volatility in the prime rate coincided with market uncertainty about monetary policy. Further, Mike showed that models of discrete variables that allow for changes in variance outperform constant variance specifications.3 This creative joining of disconnected models in the literature is something at which Mike excelled.
Mike had a terrific career. He gave sound advice to policymakers and wrote important research papers. He was an outstanding, well-liked colleague who offered freely of his time and talents, and he will be missed. *
1 "Strengthening the Case for the Yield Curve as a Predictor of U.S. Recessions." Federal Reserve Bank of St. Louis Review, March/April 1997,79(2), pp. 41-51; http://research.stlouisfed.org/publications/review/article/3183.
2 "Dynamic Forecasts of Qualitative Varia bles: A Qual VAR Model of U.S. Recessions." Journal of Business and Economic Statistics, Ja nuary 2005,23(1 ), pp. 96-104.
3 "Conditional Heteroscedasticity in Qualitative Response Models of Time Series: A Gibbs Sampling Approach to the Bank Prime Rate!'Journal ofBusinessand Economic Statistics, October 1999, 77(4), pp. 466-72.
(c) 2014 Federal Reserve Bank of St. Louis
[ Back To TMCnet.com's Homepage ]