Yesterday’s update of the Chicago Fed National Activity Index (CFNAI) strengthened the case for what’s become obvious in recent weeks: economic conditions overall improved modestly in September. Last week’s review of the numbers published to date certainly looked encouraging, as tracked by The Capital Spectator Economic Trend Index (CS-ETI). Not surprisingly, the September read on the economy via CFNAI tells a similar story.

The three-month moving average of CFNAI, a weighted average of 85 indicators, increased to -0.37 last month from -0.53 in August. A reading of -0.70 or below for the CFNAI’s three-month moving average is considered a signal that a recession has begun. By that standard, we have yet another data point that tells us that the economy wasn’t contracting last month and so the NBER is unlikely to label September as the start of a new downturn. To be sure, September’s economic profile is relatively weak, but not weak enough to argue that the economy was shrinking.

The rush by some analysts in the recent past to declare that a recession has started is curious when you consider that a sober reading of a wide range of economic and financial indicators offered little support for such a dark view. When recession risk is truly elevated, the numbers tend to speak loud and clear. Keep in mind that a high-confidence signal will be slightly dated, which is to say that the arrival of a new recession will only be obvious after the fact.
A careful review of history suggests that the best-case scenario is declaring that a recession has started two to three months after the cycle has peaked. That still leaves enough time to prepare for the worst—most folks don’t recognize that a recession has started until much later.
The alternative is to speculate what might happen in the weeks and months ahead. Of course, that requires making decisions without any data. Not surprisingly, the record with this approach isn’t much better than tossing a coin. Why’s that a problem? Because preparing for a recession that doesn’t strike can be as troublesome as ignoring the warning signs when a new downturn is a virtual certainty.
A much better solution: carefully monitor a broad set of data for the earliest signs that a new recession has started in the recent past. By that standard, according to CS-ETI and similar techniques, September wasn’t the beginning of a new downturn, based on the data in hand. When the numbers tell us differently, you’ll read about it here.