Most measures of the economy being released presently are reporting astonishing April gains. Of course, many will correctly assert that such extravagant percentage gains are the result of comparing the economy during its darkest moment of the COVID pandemic against more recent numbers which, even if not impressive during normal times, are certainly better compared to a year ago.

It is for this reason that Gardner would caution manufacturing decision-makers from placing too much weight on recent and near-term rate of change (ROC) readings when making forward-looking decisions. Rather, we would recommend examining the data as reported at present against results immediately before and after the worst of the economic contraction caused by the regulatory shutdown of much of the country over COVID.  

The latest U.S. Auto and Other Motor Vehicle Sales data from the Census Bureau is a paradigm of the reason we are cautioning manufacturers to look more carefully at the data they rely on for decision-making. April’s results are up 112% from the same month a year ago. Of course, this is NOT to say that the market is growing at 100% on a go-forward basis. So what is the answer?  Examining the latest results against those released just prior to the pandemic and those after the initial rebound show an industry which has grown around 30%. This however begs the question: why should you even believe this?

1.) Is there a long-run trend in the data? How does that trend differ from recent months or the latest reading?

2.) Is there seasonality in the data that is being captured in the ROC calculation and thus giving an improper interpretation of the market’s true movement?

A simple, linear trend of the historical data over many years (or more importantly business cycles) provides a reasonable basis from which to make a more informed assessment about recent data releases of the U.S. Auto market.  This technic is quick, simple and more informative than using the extreme results created by the unusually depressed figures posted during the worst months of the COVID pandemic.

In the case of this series, we don’t see a strong seasonal trend from the data and a long-run trend can be easily superimposed on the reported data. Applying a simple, linear trend to the above data provides a reasonable idea of the industry’s performance less much of the shock caused by COVID.  (Whether or not to include data after February of 2020 in the calculation of the trend is outside the scope of this example, but it is a worthy consideration). Comparing the latest actual of $130B against both the trended number for the same period ($103B) and the trended year-ago period ($100B) both point to industry growth of roughly 30% after removing much of the volatility in the data caused by COVID.