Today, we are pleased to present a guest contribution written by Filippo Natoli And the Fabrizio Venditi Directorate General of Economics, Statistics and Research of the Bank of Italy. The opinions expressed in this note are those of the author and do not necessarily reflect the views of Bank of Italy.
The negative reading of the advance estimate of US GDP growth in the second quarter surprised private forecasters, contributing to recession talks. Using a standard forecast model, we show that when accounting for the high level of CPI inflation and tight labor markets, the next recession seemed highly likely even a month ago, in both the US and UK.
The second quarter of 2022 saw a sharp deterioration in the momentum of the global business cycle As global trade continues to be stifled by ongoing supply bottlenecks, commodity prices have been boosted by the war in Ukraine and high inflation continues to erode consumers’ purchasing power. Meanwhile, financial conditions have tightened considerably, as central banks have responded forcefully to stubborn inflationary pressures. Against this background, the US economy recorded a negative GDP reading for the second time in a row, after economic activity declined in the first quarter. The prospects for other advanced economies look equally worrisome, particularly for energy importers, whose terms of trade have rapidly deteriorated in the wake of the invasion of Ukraine. Forecasters have taken into account these headwinds and lowered their forecasts for global growth in 2022 and 2023 (see, for example, the International Monetary Fund) and It seems that the question when instead of if Recession will happenThe search for the word “stagnation” has been on Google since last March (Figure 1).
Figure 1: Google search intensity for the word “stagnation”
(The number of results ranked between 0 and 100 during the studied period). Source: Google Trends
In the last work (Natoli and Venditi, 2022), We contribute to this debate by jointly assessing the importance of financial and macroeconomic factors in forecasting recessions in the United States and the United Kingdom since the late 1990s. In our analysis, we rely on the standardized probability prediction framework devised by Estrella and Hardouvelis (1991),
where the dependent variable is dummy equals one (zero) if the economy (not) is in recession in time r + hAnd the h is the horizon of expectations, Q is regression group, F(.) is the standard normal cumulative distribution function and is a normally distributed error.
The benchmark in the literature, which we take as a baseline, considers the slope of the government bond yield curve – the difference between the 10-year bond yield versus 3-month bond yield – as a unique indicator. However, the ability of this simple model to predict stagnation has recently been questioned in the literature (Karnezova and Lee, 2014; Ercolani and Natoli, 2020; Kelly, 2022, among others). We therefore begin by adding indices of the basic specification of market stress—financial conditions and stock market volatility, according to the view that the slope of the yield curve alone may not be able to accommodate the deteriorating financing conditions leading to a crisis. The Financial Condition Index (FCI) was constructed as an unweighted average of 10-year returns, monthly stock returns, and corporate bond yield spreads, in the spirit of Arrigoni et al. (2022), while stock market volatility is captured by VIX. In the third specification we also add two variables that summarize the macroeconomic environment – CPI inflation and the unemployment rate. Figure 2 shows coefficient estimates for the United States over the different forecast horizons (from 1 to 12). They confirm that financial indicators and macroeconomic conditions provide additional predictive power. In particular, they show that periods of (1) fixed yield curves (2) tight financial conditions (3) high uncertainty in financial markets (4) high inflation and (5) low unemployment (which is very similar to the economic environment) are likely to follow. current) from the slack. UK estimates are very similar.
Figure 2: Mean Marginal Effects, US Model
When specifications based on the slope of the yield curve are enriched with key financial and macroeconomic indicators, forecast performance improves dramatically, and in the current environment, the probability of an incoming recession jumps to values very close to 1. Figure 3 shows the time series for the expected recession probability (6 months ago) for the model based only on the slope of the yield curve (blue lines), intermediate specifications including VIX and FCI (green line), and for the complete model that also presents inflation and unemployment (red line). By standard metrics of fit, this last model is the best performing one. All models are estimated with data from January 1998 through May 2022 – that is, already available around mid-June. It turns out that financial conditions, historically, played a more important role than inflation and unemployment in predicting recessions in the early 2000s; It was also very tight before the pandemic shock that finally toppled the global economy in 2020. In the Great Recession of 2008, financial and real factors played a role. At the current juncture, on the other hand, the strongest signs of recession are coming from record high inflation and tight labor markets, in line with what is found in Domash and Summers (2022). An interesting note is that the sample period includes years of well-established inflation expectations and reliable monetary policy: our results suggest that, even in such an environment, a strong monetary response—aggressive enough to generate recession—would be required to tame inflation. Altogether, our results suggest that, at the current stage, a small downturn—re-engineering inflation without causing a recession—is unlikely. The actual mix of hot labor markets, high inflation, and tough financial conditions usually follows a recession.
Figure 3: Recession probabilities by six months, time series
Note: The slack ranges for an economy are based on the OECD stagnation indicators
Source: Natoli, F. and Venditti F. (2022). The Role of Financial and Macroeconomic Conditions in Recession Prediction (29 July 2022). Available on SSRN: https://ssrn.com/abstract= 4176581
This post was written by Filippo Natoli And the Fabrizio Venditi.