2023-03-28 10:47:25 ET
Summary
- The current situation has a lot in common with the old S&L crisis.
- Securities at banks are being funded at a loss.
- The implied capital losses on bond purchases are massive.
- It sure looks like a lot of the problem is systemic and not just idiosyncratic to one California bank.
- Three charts provide GPS on the issues.
SVB failure: Is it generalizable?
There are several unusual things about the Silicon Valley Bank situation that may be related to specific problems that SVB Financial Group (SIVB) has. That might cause us to think that it would be a standalone problem, however, there are other aspects of what happened to SVB that make us wonder if it isn't one of many problems that might exist and plague other banks as well. In short, the timely designation of SVB as a systemically important bank allowing all of its deposits to be insured may have stopped contagion for the moment but it may not have solved the problem indefinitely.
Treasury Secretary Janet Yellen has made statements and then backtracked on herself, saying at first that there were no discussions to ensure all bank deposits and then saying that it was something that could be considered. The question is what needs to be done to shore up confidence. And that gets to the question of how widespread and how deep are the problems at banks. We offer some insight on this below…
Idiosyncratic Problems at SVB
There are a lot of elements of SVB that are in the press that are critical of the bank. Some of them are related to various political criticisms of the kind of business that the bank conducted but that are not in any way related to the bank's ultimate problems. The bank had received several letters from the San Francisco Federal Reserve district bank telling it to improve certain situations. One of the letters warned the bank that it had been too slow in correcting its problems and it was going to be prohibited from engaging in any acquisitions until it had addressed the situations brought up by the Fed bank examiners.
Clearly, Federal Reserve officials knew that there were things wrong at SVB. The Fed had been concerned for a while that SVB did not seem to understand its own risk model. And this situation seemed to go on for a while. Part of that problem undoubtedly was that SVB's risk manager had quit and had not been replaced. The bank claims that it had parceled out the tasks of the risk manager, but let me assure you that's not possible.
Idio-synchronicity at SVB
There are two jobs at a bank that cannot be broken up and parceled out to local units. One is that of the risk manager, and the other is of the credit manager. These two individuals need to be independent and report directly to a high-level bank management. These are employees who ultimately become among the least-liked employees in a financial institution because it's their job to say "no." Traders and salespeople and loan officers only make money when the risk manager and the credit manager say "yes." Yes, you can do that trade. Yes, you can do business with that client. Yes, is the magic word.
But "no" is a very important word because it protects the bank. This is why you can't break down a risk manager's job to a local level, because the local levels are always going to want to do business; they're always going to want to do things that a risk manager might not allow. The risk manager must have independence and must report to very high-level management.
…and simple excess
The other thing that SVG had done was to have exposed too many assets to interest rate risk. The bank was operating in treasury securities which are always referred to as the safest assets. And this is because it's assumed that the U.S. government never goes broke and so there is zero credit risk. However, treasury securities can be imbued with a great deal of interest rate risk, and this is what the bank was doing; it had accumulated huge losses and even had dismantled some hedges that had been in place to protect it against some of the trades that had on its books. This exposed it to the full ravages of the market, un-tempered by hedging activity. These are among the idiosyncratic circumstances of this bank.
Some general risk issues for all banks
Large banks are more closely regulated than small banks or regional banks. They have to run stress tests; they have to hold more capital; they are more closely scrutinized. However, even for large banks, stress tests apparently have not been constructed to explore the sensitivity to interest rate risk. In an interview on Bloomberg this weekend, Dan Tarullo, former Fed governor who was engaged in banking oversight, and former Treasury Secretary, Larry Summers, were discussing these stress tests on banks and opining that they had become progressively less rigorous. Tarullo confirmed that the tests were not run on interest rate vulnerability. In these circumstances, that's somewhat shocking.
"Framed" by the new framework
It is well documented that the Federal Reserve ahead of the COVID crisis changed its operating framework in a way that emphasized employment more than it had in the past. When Covid dropped, there was massive dislocation and massive fiscal and monetary stimulus was launched to try to support the economy and, in particular, people who were specifically being disallowed from working in certain parts of the economy during the time of peak COVID risk. It is much easier to argue that the fiscal policy monies spent were more justifiable than the monetary stimulus. In fact, the fiscal monies were excessive, and the worst part about them is that they were not targeted. But since one objective of the government was to put a blanket over economic activity, it's far from clear the role easy monetary policy played at all. But that's what the Fed did. And this highly stimulative monetary policy was an important contributor to the problems that have since developed at banks .
We know that the combination of fiscal and monetary policy helped to encourage a rise in inflation. We also know that the Federal Reserve with the mandate to hit an inflation target of about 2% waited 12 months with inflation over the 2% target before it hiked interest rates at all. The Fed was also engaged in balance sheet expansion and it dragged its feet mercilessly, continuing to expand its balance sheet while it reduced its stimulus very gradually all the way through March of 2022. This is long after inflation had risen and even well after the Fed had agreed it was a clear problem.
Of course, inflation is a global phenomenon, and I don't want to portray the inflation that rose as only due to U.S. fiscal and monetary policy. But when there is a generalized global impulse, domestic policy can either make it better or make it worse, and policy in the U.S. was conducted in a way to make it worse.
More money means more bank deposits
We need to think about the mechanics of this a little bit. It's not simply that the Fed created the biggest boom in money supply growth since at least 1960 and followed it with the biggest bust and the first decline in M2 since 1960 early this year. Think about what a boom in the money supply means…it means a surge in bank deposits! Banks in this period became flush with funds. Meanwhile, the economy was not very active. Much of the money transferred to people wound up in banks. Municipalities too wound up with excess funds they could put on deposit - so banks had "money to burn." Deposits bore a near-zero rate of interest. So, bankers took to purchasing treasury securities that yielded more than they were paying on bank deposits. As inflation began to rise, a problem began to develop.
Fed's inflation fighting lollygags…
The problem is that the Federal Reserve did not take an aggressive action toward the inflation. In fact, the Fed had first denied inflation and then it told us it would go away on its own. Eventually, the Fed admitted that there was inflation, and in its periodic forecast known as "the dots", the Fed began to show us its plans for reeling in inflation. These plans showed very small increases in interest rates that had very little impact on the economy and on the unemployment rate and yet brought inflation down relatively quickly. The Fed was forced to modify these forecasts again and again, as inflation continued to move up into surpassing its expectation. However, the Fed's guidance was always that it would do very little and that it had a good chance of achieving a soft landing and controlling inflation.
Bad communication = Bad policy = Bad Fed
I regard Fed communication during this period as one of the biggest problems that the markets had to face. The Federal Reserve did not see the inflation threat, and while in his recent testimony and at the Fed's recent FOMC meeting and press conference Chairman Powell asserts that the Fed warned the markets, I think the record is clear that the Fed did not warn the markets adequately. Even at the Fed Chairman's last testimony 2 weeks ago before the Senate House banking/financial services committees, he was talking about how it was still possible to create a soft landing when Elizabeth Warren was discussing with him her position that the Fed desired to create a recession.
There's a lot of dissonance in this in terms of what the Fed says, what the Fed claims that it said, and what was actually communicated to markets. These three things should all be the same…they are not. Market interest rates were steady and low for a long time in 2021 even as inflation rose. Markets were not preparing for the kinds of interest rates the Federal Reserve finally adopted. Those hikes came as a shock to the market. And even once the Fed adopted these stronger rate hikes, its language never really became that aggressive. The Fed continued to talk about not harming the economy any more than it "had to," which is very different from the rhetoric of the past when it would talk about making sure it extinguished inflation.
Getting back to the banks…
In this environment, banks were buying longer-term treasury securities, funding them with these short-term low interest rate deposits and not worrying too much about the future because the Fed was downplaying the risks from inflation and downplaying what it was going to have to do in the future.
The chart below shows us what was tantalizing the banks during this time.
Chart 1
Investment incentive gaps are gone (Haver Analytics and FAO Economics)
This is a plot of the difference between the federal funds rate, which I take to be a proxy for the deposit rates of banks, and the yields of various treasury securities from two years out to the 30-year bond. In early 2022, you see great incentives for banks to be invested in these securities because the interest rate gap was enormous. But as time went on, the Fed began to raise rates and long-term rates did not move up in step with short-term rates; eventually, by early 2023, the attractiveness of this trade was completely gone as the federal funds rate had moved up above the yield on the treasury securities.
Chart #1 is a pretty clear story of the diminished incentive for banks to take the short-term deposits and place them in longer-term treasuries. However, it doesn't begin to address the real problem or the real risk of doing that.
The risk of maturity mismatch (the Mis-match game!)
Let's look at that. Take the case of a bank buying a 2-year note. The bank will buy the two-year note and it will earn the yield on the two-year note at the time that it purchases that note. The yield on that bond will last as long as the bank holds it and doesn't sell it; that yield will never change; it is fixed, it is not a floating rate note. However, over time it will be funding that two-year note at various changing short-term deposit rates - whatever they are. That's a risk for sure…
In this example, I'm using the federal funds rate as a proxy for what deposit rates became. It's not exactly right but it makes the point clearly enough. And what happens is the two-year note in March of 2022 started out with a substantial premium to the funding rates, but as the Fed hiked short-term interest rates, that two-year note yield remained the same, and eventually short-term interest rates rose enough to surpass the original yield on the two-year note and eventually they inflicted losses - cash flow losses - on the bank that purchased such as security… if that security remained on the books and was unhedged.
Chart two shows the example of the various securities purchased in March of 2022, but those yields are fixed, and then at each month in the future, we take that fixed yield and we subtract the prevailing federal funds rate (funding cost) from it to see what margin the bank was earning. Around August and September, the positive earnings on this trade turned negative; in January 2023, banks are looking at cash flow losses on the order of 2% to 2 1/2%.
Chart 2
Progressive net funding cost cash flow sours (Haver Analytics and FAO Economics)
That, however, is only part of the story. And remember that the story we're telling here is the story that pertains to all banks, so any bank during this period that was engaging in this sort of trade was going to be facing this dynamic. Now, of course, some banks are more sophisticated than others, there are various hedging tools that banks could use to try to protect their investments as they saw market conditions changing. SVB did not hedge; other banks probably did.
Next, the impact on Bond and Note prices
The next thing we're going to consider is what happens to bond prices when rates rise. What we see in the chart above is that the cash flow situation had deteriorated very sharply, but in addition to that, as interest rates rise bond prices fall. Bond contracts are written in such a way that the interest rate that is paid every six months and the dollar value of that payment never changes. This payment refers to the coupon rate. The coupon "rate" isn't fixed as a percentage; the coupon is set in dollar terms and it's expressed in percentage terms when the bond is first offered. But the coupon is actually paid in dollar terms. So, if a bond is issued with a coupon that initially yields 2% and if interest rates in the economy rise say to 3%, that bond would not be attractive to anyone because bond yields are now 3%; the dollar value of the coupon payment is too low. For this 2% security to become attractive, it would have to be sold at a discount to face value. Since the coupon cannot be altered, the trick is to lower the bond price, thereby pushing the fixed coupon payment up in percentage terms. And, in fact, a bond at a big enough discount can raise the coupon yield to the market yield, and this is what happens in the bond market regularly. This is why bond prices and bond yields always move inversely. During the period that interest rates were rising and banks were losing out on their cash flow margin; they also were also getting clobbered because the prices of their bonds were falling. That is, if banks were to mark their securities to market, there would be blood…
Chart 3
Implied mark-to-market downgrade (Haver Analytics and FAO Economics)
Chart #3 explores this issue by selecting one of the securities in the previous example, the five-year note. We compare its current yield to its previous four-year average yield. If the current yield is substantially higher than the four-year average yield, that tells us that interest rates in the economy have risen and that those previous five-year notes that have been purchased by an institution would have to be at a discount in order for them to trade at market value. So this green line on the chart is basically an indication of how much distress past bond purchases are put in because of current changes in interest rates. The line marks the difference between the current yield and the four-year average yield. The current gap is massive!
Substantial losses or mark-to-market issues
This is the chart that is the most frightening. What we see is that the five-year note yield has moved up to about two percentage points higher than its four-year average. Looking back historically, there was only one previous period when a change like this occurred. Then, it was worse, and that was in the 1980s, and we know what that was. Since the 1980s, there have been very few instances of this measure being any higher than 1% except for a very fleeting moment or two. What we have here is evidence that any bank that purchased treasury securities during this period is now facing tremendous pressure on their bond-holding portfolio unless they hedged it. At one time, SVB had hedges on its securities positions, but it let them lapse. This would only be done if the bank had decided that interest rates had risen enough and that they were about to go down and they no longer needed protection from rising rates. In that case, they could save money and add to profitability by not paying for hedges. However, the Federal Reserve has not been on the verge of cutting rates, the Federal Reserve continues to raise rates. And the trap snapped shut on SVB.
The three charts
So these three charts show us the three basic stages of the temptation of the bond investor. Chart one shows you the temptation of long rates premium over short rates and how there's a great incentive to take those short-term deposits and to fund a longer-term security. Chart two shows you that if you do that you could wind up at risk and something that looked like it was going to be a big money maker could turn out to be a money loser even in comparing the simple coupon rate to the funding rate you wind up paying on deposits. Chart three reminds you that it's not just the cash flow and the coupon rates at risk, it's the price of the bond itself. Chart three flags what is looking like systemic risk to the banking sector for banks that bought securities - and you can replicate this chart for the two-year, for the three-year, for the 10-year and it looks pretty much the same for all of them.
Yields on treasuries securities have moved up very sharply, and they're much higher than they were a few months ago, much higher than a year ago, so the less banks have been hedging their portfolios, the more they're dealing with losses. But funding costs have escalated even faster…
Banks do have various accounting options. A bank that purchased a security under a hold to maturity protocol would not have to reflect any mark-to-market loss it would occur on that bond. However, if a bond was procured for another purpose, the bank might find that it has to report this loss and might even find itself in a position where it has to realize the loss and potentially raise more capital.
One additional problem to consider is that the largest banks have the most regulation and the most capital. They also have the most sophisticated software and best-trained staff. Large banks are the least likely banks to have been caught up in the problems at hand. It is more likely that regional or smaller banks might have misread the tea leaves on interest rates and were caught with losses that were not properly hedged or possibly not hedged at all.
Systemic risk?
This suggests there's a great potential out there for systemic risk to have grown out of proportion with what we have seen over the last 50 years. Only the period around 1980 shows us worse risk. And that occurred over 40 years ago. How many people in trading rooms have that experience and remember that?
In the 1980s, the thrifts were caught in a similar kind of vice. Savings and loan institutions were in the business of making 30-year fixed rate mortgage loans. With all of these relatively low interest-rate paying assets on their books, suddenly inflation was terrible and the Federal Reserve was raising interest rates into the 15% to 20% range and thrift institutions quickly ran out of capital and became insolvent or required special treatment from regulators. That period has a lot in common with this one. Now, there also are a lot of mortgage products that have been issued on a fixed rate basis, and a lot of them were at extremely low interest rates; in addition to that, there are simple treasury securities that are out there that banks as well as credit unions would have on their books and have become subject to the same fate. In some sense, this is the savings and loan crisis all over again.
Does the bank crisis preclude good Fed policy?
The question that has arisen is whether the bank crisis is going to keep the Fed from properly addressing the inflation risk. Chart three suggests that The Fed will have to be careful in raising rates and possibly even in how long it keeps rates high because there are already significant depressed asset values as well as negative cash flow funding from past securities investments (Chart 2). The Federal Reserve's plan when it started was to raise interest rates to some higher level and to let them sit there and to bring the inflation rate down eventually. What is not clear is whether the banking system is going to be strong enough to be able to take an extended period of rates this high or a little bit higher, especially if the economy dips into recession and banks have to deal with other kinds of losses in addition to these negative positions on their securities accounts. The worst-case scenario would come if inflation ticked up again on oil or on any adverse war events, for example.
The road ahead
Of course, looking ahead like that gets tricky, because if the economy goes into recession, typically inflation comes down a little bit faster. Typically, interest rates come down a little bit faster, too, and so banks would receive a little bit of added relief from those factors even though they would be operating in a more difficult economic environment. On balance, it still looks like a lot of rough sledding ahead.
For further details see:
Is It The S&L Crisis All Over Again? 3-Graphs Put The Situation In Perspective