Evan F. Koenig

The Use and Abuse of 'Real-Time' Data in Economic Forecasting

We distinguish between three different ways of using real-time data to estimate forecasting equations and argue that the most frequently used approach should generally be avoided. The point is illustrated with a model that uses monthly observations of industrial production, employment, and retail sales to predict real GDP growth. When the model is estimated using our preferred method, its out-of-sample forecasting performance is clearly superior to that obtained using conventional estimation, and compares favorably with that of the Blue-Chip consensus.

VAR Estimation and Forecasting When Data Are Subject to Revision

We show that Howrey’s method for producing economic forecasts when data are subject to revision is easily generalized to handle the case where data are produced by a sophisticated statistical agency. The proposed approach assumes that government estimates are efficient with a finite lag. It takes no stand on whether earlier revisions are the result of “news” or of reductions in “noise.” We present asymptotic performance results in the scalar case and illustrate the technique using several simple models of economic activity. In each case, it outperforms both conventional VAR analysis and the original Howrey method. It produces GDP forecasts that are competitive with those of professional forecasters. Special cases and extensions of the analysis are discussed in a series of appendices that are available online.

Real Money Balances and the Timing of Consumption: An Empirical Investigation

This paper examines the correlation between changes in consumer spending on nondurables and services, and levels or changes in a variety of other variables that might be expected to enter directly as arguments of the household utility function or to serve as measures of household liquidity. Empirical results strongly suggest that an increase in real money balances raises the marginal utility of consumption. Once the influence of real balances is accounted for, there is little evidence that other variables have a direct impact on the timing of consumption.

Targeting nominal income: A closer look

Abstract I derive conditions under which the monetary authority should target nominal spending. The relevant spending target is a weighted average of income and consumption. Despite ‘sticky’ nominal wages, under optimal policy the economy behaves as if all markets clear.

Long-term interest rates and the recent weakness in M2

Abstract The gap between long-term market interest rates and M2 deposit rates appears to be an important short-run determinant of M2 growth. The post-1990 weakness in M2 growth can be largely attributed to a widening of this yield gap and to a gradual acceleration in the trend rate of financial innovation. The empirical importance of the yield gap is robust across a variety of alternative money-demand specifications and is discernable even in sample periods which ended well before a breakdown in standard opportunity-cost measures became obvious.

Is the markup a useful real-time predictor of inflation?

Abstract Unlike their revised counterparts, real-time markup data fail to improve the performance of a simple Phillips-curve inflation forecasting equation and have little predictive power beyond that of survey measures of inflation expectations.

Interest rates and the recent weakness in M2: An extension to the P* model of inflation

Abstract A P ∗ model of changes in the deflator for nondurables and services consumption expenditures is estimated simultaneously with the error-correction model of M2 growth described in Koenig (1996). The inflation forecasts generated by the joint model have marginal predictive power for changes in the GDP deflator and exhibit none of the systematic bias which has plagued the original P ∗ formulation in recent years. Output-market slack is of little use in predicting changes in the consumption deflator, but there is evidence of a speed effect.

The roles of inflation expectations, core inflation, and slack in real-time inflation forecasting

Using state-space modeling, we extract information from surveys of long-term inflation expectations and multiple quarterly inflation series to undertake a real-time decomposition of quarterly headline PCE and GDP-deflator inflation rates into a common long-term trend, common cyclical component, and high-frequency noise components. We then explore alternative approaches to real-time forecasting of headline PCE inflation. We find that performance is enhanced if forecasting equations are estimated using inflation data that have been stripped of high-frequency noise. Performance can be further improved by including an unemployment-based measure of slack in the equations. The improvement is statistically significant relative to benchmark autoregressive models and also relative to professional forecasters at all but the shortest horizons. In contrast, introducing slack into models estimated using headline PCE inflation data or conventional core inflation data causes forecast performance to deteriorate. Finally, we demonstrate that forecasting models estimated using the Kishor-Koenig (2012) methodology-which mandates that each forecasting VAR be augmented with a flexible state-space model of data revisions-consistently outperform the corresponding conventionally estimated forecasting models.