When keeping volatility of the shocks constant Sample Clauses

When keeping volatility of the shocks constant. In this section, we examine the time variation in the frequency of price adjustment while the disturbance shocks are constant across time, Mfreq. Although this version of the model does not fit nearly as well as our best-fitting model, there are several reasons to examine its implications. First, this model’s fit does is not worse than the constant-parameters model. Hence, this setting can still be used for economic analysis. Second, the model reflects the dominant view that firms adjust more frequently their prices in periods of high-inflation, and its economic implications allow us to better understand the role of such private-sector changes in the economy. Table 1 reports the mode for each parameter with a 90 percent probability interval for both the structural parameters and shock processes of the model Mfreq. The estimates for most structural parameters are quite similar to those in M freq+vol . There are, however, some noticeable differences. First, the estimate for the Xxxxxx-rule coefficient for inflation (ψπ) is equal to 1.3207, which is slightly lower than the estimate of the model Mfreq+vol . The second notable difference concerns the persistence of the markup and productivity shocks. The estimate for the persistence of the markup shock (ρp) decreases from 0.3798 in Mfreq+vol to 0.1209 in Mfreq . The estimate for the persistence of the productivity shock (ρz) increases to 0.3974 when the size of shocks is not taken into account. Regarding the estimate for shocks variances, the highest variance is the preference shock (σb) with a mean of 2.8059. The estimate for other standard deviations of shocks lie between 0.1583 and 0.6506. The 0.90 percent error bands for all the shock variances indicate that uncertainty about them is extremely small, except for the preference shock standard deviation with the probability interval [2.1092; 3.6600]. It is striking to observe that price setting behavior by firms has significantly changed over time. Our estimates provide two distinct regimes of the Xxxxx pricing parameter [γp(st)] with a value around 0.8720 under Regime 1—the “low-frequency” regime, and 0.7162 under Regime 2—the “high-frequency” regime. Their tight 90 percent error bands do not overlap, suggesting that the estimates are robust. The estimate for the Xxxxx parameter under the ”low frequency” regime, implying a price duration around 9 quarters, is consistent with the previous macroeconomic literature [Smets and Wouters (2007) and Xxxxxxxxxx and Pr...