Real Time Forecasts of Inflation: The Role of Financial Variables

Abstract

The working paper series promotes the dissemination of economic research produced in the Department of the Treasury (DT) of the Italian Ministry of Economy and Finance (MEF) or presented by external economists on the occasion of seminars organised by MEF on topics of institutional interest to the DT, with the aim of stimulating comments and suggestions. The views expressed in the working papers are those of the authors and do not necessarily reflect those of the MEF and the DT.

Now-casting inflation using high frequency data

“…This allows us to fully exploit the co-movement in our dataset, without losing any information. Moreover, in contrast to MIDAS models, like the one proposed in Monteforte and Moretti (2010), the proposed framework does not require to impose high frequency data as predetermined variables. It therefore allows to disentangle model-based "news" from each release and then to assess their impact on forecast revisions.…”

Nowcasting U.S. Headline and Core Inflation

“…Modugno (2013) and Monteforte and Moretti (2013) both nowcast year-over-year inflation rates rather than monthly inflation rates. Using the same underlying monthly models and the same cases described above, Table 5 assesses the ability of the models to nowcast yearover-year inflation.…”

“…In one exception, Modugno (2013) applies a dynamic factor model-with a larger number of monthly, weekly, and daily data series compared with the limited set of variables we work with-to nowcast year-over-year U.S. CPI inflation. Monteforte and Moretti (2013) use a combination of a dynamic factor model to construct a measure of core inflation and mixed frequency data in the context of a mixed data sampling (MIDAS) regression model based on Ghysels et al (2004Ghysels et al ( , 2005 to nowcast year-overyear euro area inflation. We present results for nowcasting U.S. monthly, year-over-year, and quarterly inflation, especially because the latter is the usual jumping-off point for economists doing quarterly forecasting exercises.…”

“…Preliminary analysis favored this parsimonious approach over MIDAS regression models featuring a larger number of high-frequency indicators along the lines of Monteforte and Moretti (2013). 18 To take advantage of potential gains coming from model averaging, some of which are documented in Andreou et al (2013), we run two separate MIDAS regressions using our two high-frequency data series-the first with daily oil prices and a second with weekly gasoline prices, both of which enter the model in natural log first differences-and then construct an average nowcast or forecast for the monthly inflation series at horizon h.…”

Nowcasting US inflation using a MIDAS augmented Phillips curve

“…In particular, Marcellino and Schumacher (2010) and more recently Ferrara and Marsilli (2014) showed that combining factorial analysis with MIDAS approach (into the so-called factor-augmented MIDAS model) successes in addressing large databases with mixing frequencies. Monteforte and Moretti (2013) showed that including daily variables using the MIDAS framework helps to improve inflation forecasts with respect to models that only consider monthly variables in the euro area.…”