Published papers

Which labor market specification is better able to describe inflation dynamics, a widely used sticky wage model or a recently investigated labor market search model? Using a Bayesian likelihood approach, we estimate these two models with Japan's data. This article shows that the labor market search model is superior to the sticky wage model in terms of both marginal likelihood and out-of-sample forecast performance, particularly regarding inflation. The labor market search model is better able to replicate the cross-correlation among inflation, real wages, and output in the data. Moreover, in this model, real marginal cost is determined by both hiring cost and unit labor cost that varies with employment fluctuations, which gives rise to a high contemporaneous correlation between inflation and real marginal cost as represented in the New Keynesian Phillips curve. © 2012 Western Economic Association International.

This paper analyzes the revision to Japan's labor productivity, measured using Japan's System of National Accounts (SNA) data. We draw three main findings from our analysis. First, SNA data has been substantially revised in and after the second comprehensive revisions, as well as at the earlier stage of revisions. We find that the past absolute revisions to the annual growth rate of labor productivity often went beyond 1% point. Second, the annual growth rate of labor productivity has been revised upward by 0.4% points on average. We show that part of its upward revisions reflects an underestimation of employment through an increase in 'non-response people,' people who do not respond in the Population Census. Third, revisions to source data such as the Population Census and the Employment Status Survey are helpful to predict revisions to labor productivity growth. Our regression results suggest that labor productivity is likely to be revised upward during expansions or with low real-time estimates of value added. We conclude that the three findings indicate that labor productivity during the 2000s will experience substantial revisions in the future. This conclusion takes into account the fact that the SNA after 2000 has experienced at most one comprehensive revision. The upcoming revisions to labor productivity can be positive rather than negative. © 2011 Elsevier Inc.

This paper (i) reviews the different channels of transmission of prudential policy highlighted in the literature and (ii) provides a quantitative assessment of the impact of Basel III reforms using several policy-oriented DSGE models. It shows that the long-term effects on GDP of higher capital requirements are positive when the associated benefits are accounted for in addition to their costs. However, the results crucially depend on assumptions about crisis probability and severity. For liquidity regulations, only models capturing benefits of increased liquidity (e.g., preventing bank runs) show a net benefit.

This paper evaluates professional forecasters' behavior using a panel data of individual forecasts. We find that (i) professional forecasts are behavioral, and (ii) there exists a stock-bond dissonance: the forecasting behavior seems to be stubborn in the stock market, but jumpy in the bond market. Even in the same country, forecasting behavior is quite different by market. © 2013 Elsevier B.V.

Standard theoretical models predict that higher inflation expectations generate greater current consumer spending at the zero lower bound of interest rates. However, recent empirical studies using U.S. micro data find negative results for this relationship. We use micro data for Japan, which has experienced low interest rates for a prolonged period, to estimate ordered probit models with a variety of controls. We find robust evidence supporting the prediction of standard models: survey respondents with higher expected inflation tend to indicate that their household has increased real spending compared with 1year ago but will decrease it in the future. This relationship appears to be stronger for asset holders and older people. (JEL E20, E21, E30, E31, E50, E52)

In this paper, we calculate the potential output and the output gap using a Bayesian-estimated DSGE model of Japan’s economy. For bridging the gap with conventional measures, we define our measure of potential output as a component of the efficient output generated only by persistent growth rate shocks. Our potential growth displays a high degree of smoothness and moves closely with conventional measures. Moreover, the output gap from our measure of potential output shows better forecasting performance for inflation—in particular, at short horizons—than other measures of output gap.

In contrast to affine term structure models, Black's (1995) model of interest rates as options has properties suitable to examine the yield curve when the short-term interest rate is near zero. We estimate a Black's model with Japan's data to extract market expectations about duration of zero interest. We find that expectations about duration have substantially varied, which contradicts with the assumption utilized in the literature. We also find a tight link between expectations about duration and survey measures of inflation expectations, which appears to be attributable to the Bank of Japan's commitment conditional on inflation.

We use a regime-switching model to examine how exchange rate volatility is related to the failure of uncovered interest parity. Main findings are as follows. First, exchange rate returns are strongly influenced by regime switches in the relationship between the returns and interest rate differentials. Second, low-yielding currencies appreciate less frequently, but once it occurs, their movements are faster than when they depreciate. Third, depreciation of low-yielding currencies and low volatility are mutually dependent on each other. Finally, these three findings are more evident for shorter horizons. The second and third results are consistent with a market participants' view: short-term carry trades in a low-volatility environment and their rapid unwinding substantially influence exchange rates. We consider the effects of funding liquidity to explain these results. © 2011 Elsevier Ltd.

Many studies estimate risk premiums on federal funds futures to extract monetary policy expectations by assuming that average forecast errors of the expectations are zero or that survey forecasts are good proxies for the expectations. These assumptions, however, may fail due to an unanticipated declining trend in the federal funds rate and to survey respondents' strategic behavior. Consequently, the premiums estimated under these assumptions may be biased. We propose a new method to estimate the premiums and find that the premiums have been often negative since 2000, which is generally consistent with the negative betas observed in the 2000s. © 2009 The Ohio State University.