Elmar Mertens
 
Search this site

Research

My research interests are monetary economics, asset pricing and time series econometrics.

Here you find my

None of the material posted here does necessarily reflect the views of the Board of Governors or the Federal Open Markets Committee.

Research Agenda

My research interests are in macroeconomic, finance and applied econometrics with particular focus on the links between monetary policy, business cycles and asset prices. Asset valuations reflect investors perception about fundamentals and the state of the business cycle, which is again influenced by monetary policy. Moreover, monetary policy directly controls the real payoffs to an important asset class: nominal bonds.

There is a lot of uncertainty in the conduct of monetary policy. In one form, this relates to the policymaker's efforts to understand both the general workings as well as the current state of the economy. (Economic research is both an observer and an actor here.) On the other hand, central banking is a form of delegated management with considerable potential for information being hidden from the public; be it internal views about the economy, or preferences and policy targets. Crucial for the effectiveness of monetary policy are public perceptions about the targets and credibility of central bank policies. Understanding the optimal management of public beliefs about the central bank's hidden knowledge / targets is the focus of my recent work. This is important in several respects: First it sets a benchmark against which to evaluate the outcomes under a fully transparent policy. As my job market paper showed in a fully forward looking model, ``don't ask don't tell'' about output targets is the right approach when policy is discretionary. Secondly, the paper develops a simple but general method for computing optimal policies when the policymaker has better information about structural shocks than the public. The method is designed to handle the kind of large-scale, dynamic models used in policy analysis.

Other work for my thesis includes an empirical VAR analysis of the nature and transmission of structural shocks driving output, inflation and interest rates. In U.S. data, standard shocks like ``technology'' and "monetary policy" behave broadly in accordance to textbook models of the business cycle (RBC/ New Keynesian). But a better understanding of the overall co-movements requires to understand shocks associated with inflation persistence. Another motivation for studying public learning about policy shifts as one possible explanation.

In a methodological paper, my thesis takes a deeper look at the methods commonly used to estimate responses to structural shocks. A growing literature has started to criticize the validity and accuracy of VARs when used for identifying shocks based on their long-run effects. Recently, a new method has been proposed by Lawrence J. Christiano, Martin Eichenbaum and Robert Vigfusson in the 2006 Macro Annual of the National Bureau of Economic Research, with the promise of delivering better results. My study asks whether these estimators are useful. Alas, the answer is no. The key problem for empirical analysis with long-run restrictions is not so much the specification of the VAR, but the well-known difficulties in disentangling permanent trends from long-lasting cycles in the data.

Prior and during my studies, I also acquired a background in banking and consulting which shaped my work in financial economics, particularly asset pricing. Early empirical work on return predictability includes my diploma thesis (``Estimating Time-Varying Risk Premia of European Asset Classes'') at the University of St. Gallen and a job at a quantitative investment firm where I designed and implemented the backtesting process for their model. At the thesis stage, this line of research was continued in a joint paper with Philippe Bacchetta and Eric van Wincoop linking return predictability in stocks, bonds and forex markets to survey errors (pdf).

In terms of future work, I intend to pursue further the implications from hidden information in policy problems. Having demonstrated the versatility of my approach in a stylized setting, I plan to work on applications related to inflation scares and the term structure of interest rates evidence from the ``Great Inflation'' of the 1970's. Recently, I have been working on a companion paper about the optimal commitment policy when the central bank is better informed about economic fundamentals.