The consensus narrative around the latest vintage of the Fed’s senior loan officer survey is a “could’ve been worse” story.
Although banks continued to tighten standards over the last three months, the situation isn’t as acute as many feared considering “recent developments” (as the March FOMC minutes euphemistically described the fallout from SVB’s collapse) in the banking sector.
I’m not sure what, precisely, market participants expected. The survey wasn’t encouraging by any stretch, and 2023 isn’t 2008. Or not yet anyway. To my mind, the survey was consistent with the reality of the situation: Standards are as tight as they’ve ever been outside of major crises and demand for credit is quite weak.
But that wasn’t “good” enough to match the hype, and as such, some fear the situation could get considerably worse, or at least on the credit “supply” side (i.e., standards could tighten further).
What would that mean for corporate profits? Well, I’m not sure, exactly. The figure below suggests that on a kind of naive interpretation (which is all overlay charts are good for in many cases), the bottom could fall out.
Goldman’s David Kostin did the math (or some math) and came away with what sounded like a relatively benign assessment in terms of what a further tightening of lending standards in the Fed survey (say, to match 2008 and 2020 recessionary levels) would imply for the bank’s baseline EPS forecast. In short, he didn’t predict a catastrophe even in the event the next survey looks worse than the one markets received this week.
I’m not sure it’s especially useful to extrapolate a specific mechanical relationship, even as a relationship plainly does exist. Rather, this might be a rare case when “naively” eying an overlay chart (like the one shown above) is more useful than trying to determine what the mathematical relationship is.
The same goes for the figure below, which shows the same series from the Fed survey with high-grade corporate credit spreads.
The grey shaded area ostensibly depicts a pretty yawning disconnect. The “fundamentals” might indeed be sound for IG borrowers, but theoretically there should be some connection between the premium high quality companies pay to borrow in capital markets and credit conditions facing large and middle-market firms looking for loans.
I should be clear, though: I don’t love either of the two charts shown above. I felt compelled to editorialize around them because, in light of the bank stress, and considering the never-ending cacophony of recession warnings, the Fed survey is being scrutinized relentlessly.
High-grade credit spreads and reported earnings for S&P 500 companies are a function of too many factors to plausibly enumerate, which makes these sorts of comparisons quite difficult. As Kostin wrote, “commentary from Q1 earnings season suggests that while managements expected tightening lending conditions, most large-cap companies have not yet seen evidence of it.”
About the best we can do is offer the (somewhat trite) observation that if lending standards continue to tighten and demand for loans is nonexistent, it doesn’t bode well for business investment, or for the economy more generally. You don’t need any overlay charts to come to that conclusion, though, or if you do, a simple visual of the C&I lending standards series with shaded areas denoting recessions will work just fine.
I think econometrics should be paired with humility.
For example, recession probability models seldom get higher than 50-60% right before recessions start, which implies that those models are at best 50% predictive.
I took econometrics as part of my MBA. The prof was awful. When I arrived at class the first day I was regaled with 100 green-bar paper charts taped to every square inch of wall space in the room. No humility there. Of course, 50% right is what we would get from guessing. The next semester I spent the summer in a private seminar with the full prof who was responsible for running the bus college intro stat course. Nine of us met at the boss prof’s house, two hours daily, three days a week, spending the summer immersed in the useful task of really understanding the subject we were all going to be teaching for the coming year. Eight of us would be assigned four sections each of the course for all working sessions. One other guy was also involved. He already had a PhD in Metallurgical Engineering and managed quality in the headquarters of Timkin Roller Bearings. He was an operations PhD student and was to do the once weekly stat lectures. If I had been through this summer exposure earlier I would have had full knowledge of the futility of those hundred stupid econ charts. My late wife was a statistician for the Ohio Department of Labor and helped me when I needed it. There ought to be rules for economists playing with this stuff. It can be dangerous. Interestingly, the best stat at Ohio State was not taught in the math department or in econ. Rather, it was taught in graduate level Psych courses. Courses over there prepared me for my dissertation.