There is widespread agreement that monetary policy should be evaluated by using forward-looking Taylor rules estimated with real-time data. For the case of the U.S., this analysis can be performed using Greenbook data, but only through 2002. In countries outside the U.S., central banks do not regularly release their forecasts to the public. I propose a methodology for conducting monetary policy evaluation in real-time when forward-looking real-time data is unavailable. I then implement this methodology and estimate the resultant Taylor rules for the U.S., Canada, the U.K., and Germany. The methodology consists of calibrating models to closely replicate Greenbook forecasts, and then applying them to international real-time datasets. The results show that the U.S. output gap series is well described by quadratic detrending, while Greenbook inflation forecasts can be closely replicated using Bayesian model averaging over Autoregressive Distributed Lag models in inflation and the GDP growth rate. German and U.S. Taylor rules are characterized by inflation coefficients increasing with the forecast horizon and a positive output gap response. The U.K. and Canada interest rate reaction functions achieve maximum inflation response at middle-term horizons of about 1/2 year and the output gap coefficient enters the reaction functions insignificantly. Estimating the U.K. and Canadian Taylor rules as forward-looking is crucial, as backward-looking specifications produce nonsensical estimates. This is not the case for the U.S. and Germany.

Although the persistence of inflation is a central concern of macroeconomics, there is no consensus regarding whether or not inflation is stationary or has a unit root. We show that, in the context of a “textbook” macroeconomic model, inflation is stationary if and only if the Taylor rule obeys the Taylor principle, so that the real interest rate is increased when inflation rises above the target inflation rate. We estimate Markov switching models for both inflation and real-time forward looking Taylor rules. Inflation appears to have a unit root for most of the 1967 – 1981 period, and is stationary before 1967 and after 1981. We find that the Fed’s response to inflation is also regime dependent, with both the pre and post-Volcker samples containing monetary regimes where the Fed both did and did not follow the Taylor principle. This contrasts to recent research that suggests the Fed’s response to inflation has been time invariant, and that changes in monetary policy only occurred with respect to the output gap.

Most studies which focus on explaining central bank behavior using interest rate reaction functions fit a single specification of a monetary policy rule using the longest available span of historical data. This data, however, is revised and therefore does not reflect the information available to monetary authorities at the time they are formulating policy. We find that differences in estimated Taylor rules based on revised and real-time data are more important for Germany than for the U.S., Taylor rules using real-time data suggest significant differences between U.S. and German monetary policies, and Taylor rules for the U.S. using inflation forecasts are nearly identical to those using lagged inflation rates. We then investigate the implications of the use of real-time data for evaluating out-of-sample exchange rate predictability. Using a model for the dollar/mark nominal exchange rate with forecasts based on Taylor rule fundamentals, we find strong evidence of predictability of exchange rate changes at the one-quarter horizon using real-time, but not revised, data.

We look at how well several alternative Taylor-rule specifications describe Federal Reserve policy decisions in real time, using the newly developed Giacomini and Rossi (2007) test for non-nested model selection in the presence of (possible) parameter instability. Further, we isolate those Taylor-rule features that are most important for achieving a strong relative real-time performance. A hybrid of the Koenig (2004) and Clarida-Gali-Gertler (2000) models seems to perform consistently better than alternatives. Key features of the hybrid model are partial adjustment of the federal funds rate toward an equilibrium rate that depends on the unemployment rate and forward-looking measures of inflation and real growth.

What makes a good Taylor rule? We addressed this issue by using a recently developed econometric technique to determine how the original rule and subsequent variations perform using different measures of inflation, output and unemployment. We found that the rule remains relevant today, despite the changes wrought by globalization, financial market innovations and technological advances.

Abstract currently not available.

Yes; maybe.

Can AIDS be beneficial for the society? May well be! In this paper I argue that in developed countries AIDS might have positive influence on various aspects of people's life. Specifically, using US state level data I show that recent decrease in abortions rate can be explained by a decrease in unwanted pregnancies, which is the result of the threat of getting AIDS. In this prospective, AIDS improves women's health, amends children "quality", ameliorates human capital, and though these channels, promotes growth.

This paper studies a general equilibrium growth model with stochastic shocks to technology, where consumers maximize separable in labor and consumption utility with non-equal corresponding risk averseness parameters. Similar to Hansen (1986), I assume labor to be indivisible, so all variation in hours worked is due to fluctuations of employment. The model is solved using Parameterized Expectation Algorithm (PEA) enhanced with moving bounds. Results show that this economy displays relatively small volatility in hours worked and correlation of hours worked with GDP, significantly improving classical Hansen results.

The paper examines the causes of bank failures in Ukraine during 1998–2003 using micro-leveldata. Two basic models are used and their performance compared. The first model is the Giant Logit model, which models the probability of bank failure. The second model focuses on estimating time-to-failure, and uses an Accelerated Failure Time framework with time-varying covariates. In addition to standard financial ratios suggested by the C.A.M.E.L. scheme, I concentrate on the role of managerial quality, reflected by DEA efficiency, which is preliminarily tested and adjusted for environmental differences. Finally, a bank ranking is built and the forecast for potential failures in July 2003 is made. The results show that inefficient banks tend to fail; so do banks, which over-invest in securities, and hold significant value of demand deposits; bank’s size has the opposite effect. Further, an analysis of hazard function shows that for an average bank the probability of bankruptcy increases until the bank reaches the age of approximately 2.5 years, then steadily declines.