In an earlier post from last December, I noted the correlation between growth in RV sales and recession. Extrapolating October YTD data for 2018 as a whole, I obtained 11% probability of recession. Extrapolating June YTD data for 2019 (thanks to reader AS for compiling the time series data), I obtain 70% probability of recession for 2020.
Here’s the time series:
Figure 1: Growth rate of Recreational Vehicle sales, in log first differences (blue). NBER defined recession dates shaded gray. 2019 observation extrapolated based on first six months’ data. Source: rvia.org, NBER, and author’s calculations.
I estimate a probit regression over 1983-2019 period:
Pr(recessiont) = -1.09 – 5.36(salesgrowtht-1) + ut
McFadden R2 = 0.26, NObs = 37. Bold denotes significant at 5% significance level. Figure 2 shows the relevant predicted probabilities. (Pretty close to AS’s results, if I use exact formula instead of log approximation for growth rate)
Figure 2: Probability of recession predicted by probit regression (blue). NBER defined recession dates shaded gray. 2020 prediction based on 2019 observation extrapolated based on first six months’ data. Source: rvia.org, NBER, and author’s calculations.
The probability of recession in 2020, assuming the rest of 2019 is the same as the first half in terms of RV unit sales, is 70%.