Going through the myriad of indicators and models I review each week, I was surprised to find that my primary economic model (which is developed and maintained by Ned Davis Research) had slipped from neutral to moderately negative this week. In fact, the current weekly reading of the model had fallen to the lowest point since mid-2013. I reported the change in my weekly review that is published on Monday. This prompted a call from one of our financial advisors, who wanted to know if we shouldn’t now be worried about a recession.

On the surface, such a question makes complete sense. When an economic model moves into a moderately negative zone, it is logical to project that the economy might be in trouble.

However, it is important to note that this model is designed to use economic indicators to “call” the stock market – not the state of the economy.

Historically, the model’s hypothetical buy and sell signals (which I have been watching for more than a decade) would have been pretty good. Assuming you went long the S&P 500 when the model was on a buy signal and then moved to cash when the model was on sell signals, the results are pretty good. Going back to 1965 and applying the signals would have produced a hypothetical annualized gain of +12.3% per year, which is nearly double the buy-and-hold annualized return of +6.4%. And 79% of the hypothetical trades would have been profitable.

The disconcerting part about the model’s most recent reading is that the S&P 500 has lost ground at an annualized rate of -8.5% per year when the model has been in the current moderately negative mode. While not good by any means, this also isn’t as bad as the -23.0% annualized return seen when the model is rated as outright negative.

In looking at the historical signals of the model, I note that the sell signals tend to be either early or false. On the positive side, good signals (circled on the chart below in red) were given well in advance of the big stock market declines seen in 1966, 1968, 1973-74, 1977, 1987, and 2008.