This morning, the WSJ published an article, Few Banks Are Hedging Interest-Rate Risk , about a recent SoCal academic paper on bank hedging activities.

I’m posting a reference to this article to give context to my comments: not as an endorsement.

Sometimes academics become so enamored with the data that’s available in regulatory filings that they find “interesting” things that aren’t there or that they misapply in their conclusions.

It’s true, most banks don’t hedge mortgage-backed securities (recall these are mostly the problem in the banking system today). Why? For several reasons:

1. Because it is hard to hit a moving target. When these securities were booked, they may have had an expected WAL of 3-4 years. Once rates started rising, these securities extended and enjoyed a special cruelty known as negative convexity, which means that their WALs extend as rates are rising–a double whammy in terms of depreciation. The skill level requisite for hedging a debt security with negative convexity is high and involves a significant amount of basis risk. Get it wrong, and you could be losing money on your hedge and your underlying instrument.
2. Banks bought these securities for yield. If they would have done the analysis for a proper hedge, they would have found that the net yield would have been low enough that they would have seen that it wasn’t worth buying them to hedge them.
3. Related to #2, hedging a negative convexity debt security efficiently usually involves dynamic hedging techniques, which is very difficult to get right and tend to result in hedging becoming prohibitively expensive just when you need it most.
4. Accounting regs for hedging are idealistic and disincentive hedging.

The SoCal paper makes it seem like the issue was that banks bought appropriate securities but failed to hedge them. This isn’t the issue at all.

The issue is that banks failed to recognize the unique nature of the interest rate transition that was inevitable, and chose inappropriately negative convex debt securities at the absolute worst time.

I’ll leave this discussion with an answer from Daryl Bible (former CFO of Truist) to an analyst’s question about their thinking when they bought $5 billion in MBS in 3Q2020:

“so our current duration of our (MBS) portfolio because of prepayment speeds picking up were just a tad over 3years…right now, 3.1. But they do have negative convexity, so it can move in and out from that perspective (in other words, the WAL will change). What I would say is that we are in the midst of moving some more of our liquidity that we have. In fact, we have a little over $30 billion at the Fed.

Currently, we are moving that over, some of it, this quarter, maybe more of it into early next year.

We are layering in some hedges. Now I would tell you the hedges that we’re putting on are pay-fixed hedges. We’re buying mortgage-backed securities, which, as you know, have cash flows that pay out over the life of those assets. The way FASB has approved hedge accounting on this, we’re only allowed to use bullet swaps. They do have a task force that they are working on, trying to look for other ways to allow for this. It’s called last layer of hedging, and we’re hopeful that we’ll be able to put on a little stronger hedges. But the hedges we’re putting on will mute some of the OCI volatility. If they get come through and allow us to use maybe amortizing swaps instead of just bullet swaps, that would significantly improve the performance of those hedges. So we’re hopeful about that. But we are trying to hedge it the best that we can. Right now, the cost of these pay-fixed swaps are really low at 12 basis points. So it doesn’t really impact it. So we are in the midst of [indiscernible].

Yes, I always look at it as an opportunity cost right now. We could have lower rates for the next 3 years. That’s what’s in the forecast; 5 years, if you just don’t know. And I think it’s good to be deployed. The way I would think of it, though, is that if rates were to go up or we started to lose some of the “surge” deposits, our cash flows from this investment portfolio we’re building could be easily $10 billion a quarter. So we could just not reinvest. If we have strong loan growth, we could use that cash flow to deploy into loan growth. So it gives us a lot more flexibility, a lot more optionality. And it also helps protect our margin and help run rate.

You pay us to run our company and do what we think is best. We think this is a good balanced approach to managing the company.

TFC’s 3Q2020 Earnings Conference call transcript

Truist went through all this to try and pick 100 bps of yield on its cash, and Daryl is as good a CFO as there is in banking. At year end, Truist was sitting on unrealized losses on its securities book that amount to nearly 60% of tangible common equity.