Everyone's been chatting about a recession recently.
Some argue that most signs point to a slowdown.
Others argue that there's no way, as other parts of the economy are still doing well.
Wells Fargo has taken somewhat of a win-win stance on the issue.
In a recent note to clients, a Wells Fargo team led by chief economist John E. Silvia wrote that there is a 23.5% chance of recession in the next six months.
This is the highest reading in the post-Great Recession era, but still isn't that ominous. The team writes that a number above the so-called warning threshold of 50% would suggest a recession could be coming within six months.
Notably, this forecast uses their official Probit model, which uses the LEI, the S&P 500, and the Chicago-PMI employment index as predictors. Wells' latest reading is based on January 2016 data.
However, the team also poses a somewhat open-ended question on how one can actually quantify the risk of a near term recession.
They write that in addition to their official probit model (which was built in 2007), they have seven other models that all predict different probabilities. The eight models' forecasts range from a 3.6% chance of recession (using the yield spread model) all the way to 76.2% (using IP, S&P 500 Index, and the CRB Index).
So basically, depending on the model used, you're looking at either a definite "no" or a solid "probably" and everything in-between on whether the US economy is heading into recession.
Which doesn't really tell you that much.
Wells Fargo Average of all the models.
For starters, Silvia concedes that there have been a few false positives, including Q1 1996 and Q4 1998 (meaning the models gave a reading above 50% but no recession followed in the six months after).
Moreover, the team notes their Q1 reading is based only on one month, January 2016 and monthly data can be quite volatile.
In light of all this, Silvia suggests that maybe the best thing to do is to take the average of all their models.
"One possible way to summarize the results from all of these models is to calculate the average of these probabilities and then examine historical performance of the average probability," he wrote in the conclusion of the note.
"The current average probability is 37.3% and this method predicted all recessions successfully since 1980 without producing any false positives," he writes, which you can see in the chart above. "Different models utilize different predictors to capture the state of different sectors of the economy and therefore an average probability may reflect the average risk posed by these sectors."
In any case, overall, Silvia writes that the key takeaway is:
"At present, we are not calling for a recession within the next six months. However, given that the recession probabilities based on our official model and average of all models are somewhat elevated, it is not wise to dismiss recession risk. We will closely monitor the upcoming data in the coming weeks and months to assess where the U.S. economy is headed-stay alert!"
In other words, a recession is probably not coming... but maybe it is.