Revisiting useful approaches to data-rich macroeconomic forecasting
This paper analyzes the properties of a number of data-rich methods that are widely used in macroeconomic forecasting, in particular principal components (PC) and Bayesian regressions, as well as a lesser-known alternative, partial least squares (PLS) regression. In the latter method, linear, orthogonal combinations of a large number of predictor variables are constructed such that the covariance between a target variable and these common components is maximized. Existing studies have focused on modelling the target variable as a function of a finite set of unobserved common factors that ...
Parsimonious estimation with many instruments
We suggest a way to perform parsimonious instrumental variables estimation in the presence of many, and potentially weak, instruments. In contrast to standard methods, our approach yields consistent estimates when the set of instrumental variables complies with a factor structure. In this sense, our method is equivalent to instrumental variables estimation that is based on principal components. However, even if the factor structure is weak or nonexistent, our method, unlike the principal components approach, still yields consistent estimates. Indeed, simulations indicate that our approach ...
An Examination of U.S. Dollar Declines
Although the dollar strengthened somewhat recently, its level relative to the currencies of the United States’ main trading partners is nonetheless 11 percent lower than it was at the start of 2009. This represents one of the more pronounced periods of dollar weakness over the past two decades and consequently has garnered considerable attention from market participants and policymakers alike. In this post, we examine the role of market uncertainty and currency risk premia in the pace and size of episodes of dollar weakness since 1991. We find that the most recent bout of U.S. dollar ...
Alternative Indicators for Chinese Economic Activity Using Sparse PLS Regression
Official Chinese GDP growth rates have been remarkably smooth over the past decade, in contrast with alternative Chinese economic data. To better identify Chinese business cycles, we construct a sparse partial least squares (PLS) factor from a wide array of Chinese higher-frequency data, targeted toward variables that are highly correlated with important aspects of the Chinese economy. Our resulting alternative growth indicator clearly identifies Chinese business cycle fluctuations and it performs well both in out-of-sample testing for China as well as when applied to other economies. Using ...
Forecasting Inflation with Fundamentals . . . It's Hard!
Controlling inflation is at the core of monetary policymaking, and central bankers would like to have access to reliable inflation forecasts to assess their progress in achieving this goal. Producing accurate inflation forecasts, however, turns out not to be a trivial exercise. This posts reviews the key challenges in inflation forecasting and discusses some recent developments that attempt to deal with these challenges.
Uncertainty about Trade Policy Uncertainty
We revisit in this note the macroeconomic impact of the recent rise in trade policy uncertainty. As in the literature, we do find that high trade policy uncertainty can adversely impact domestic and foreign economic activity. In addition, we identify an alternative business sentiment channel that is separate and distinct from the impact of trade policy uncertainty, which provides a complementary explanation of the recent developments in the U.S. and global economic activities. This sentiment channel also implies that subsiding trade policy uncertainty does not necessarily result in a recovery ...
Oil Prices, Global Demand Expectations, and Near-Term Global Inflation
Oil prices have increased by nearly 60 percent since the summer of 2020, coinciding with an upward trend in global inflation. If higher oil prices are the result of constrained supply, then this could pose some stagflation risks to the growth outlook—a concern reflected in a June Financial Times article, “Why OPEC Matters.” In this post, we utilize the demand and supply decomposition from the New York Fed’s Oil Price Dynamics Report to argue that most of the oil price increase over the past year or so has reflected improving global demand expectations. We then illustrate what these ...
Real-time inflation forecasting in a changing world
This paper revisits the accuracy of inflation forecasting using activity and expectations variables. We apply Bayesian-model averaging across different regression specifications selected from a set of potential predictors that includes lagged values of inflation, a host of real activity data, term structure data, nominal data, and surveys. In this model average, we can entertain different channels of structural instability by incorporating stochastic breaks in the regression parameters of each individual specification within this average, allowing for breaks in the error variance of the ...
A New Approach for Identifying Demand and Supply Shocks in the Oil Market
An oil-price spike is often used as the textbook example of a supply shock. However, rapidly rising oil prices can also reflect a demand shock. Recognizing the difference is important for central bankers. A supply-driven increase in the price of oil can result in higher unemployment and inflation, leaving central bankers with the difficult decision to loosen policy, tighten policy, or not respond at all. A demand-driven increase reflecting global growth may support the case for tighter policy. In this post, we describe an approach for decomposing oil price changes into supply and demand ...
Model selection criteria for factor-augmented regressions
In a factor-augmented regression, the forecast of a variable depends on a few factors estimated from a large number of predictors. But how does one determine the appropriate number of factors relevant for such a regression? Existing work has focused on criteria that can consistently estimate the appropriate number of factors in a large-dimensional panel of explanatory variables. However, not all of these factors are necessarily relevant for modeling a specific dependent variable within a factor-augmented regression. This paper develops a number of theoretical conditions that selection ...