The Real Banking Crisis
MARCH 31, 2023
The single most emblematic statistic for the quarter, computed by Carl Quintanilla, is that Apple, Nvidia, Microsoft, Meta, Tesla, Amazon, Alphabet, Salesforce, and AMD contributed 160% of the total S&P 500 gain, which means that it would have been negative without them.
Three stocks, Apple, Nvidia (partly fueled by Artificial Intelligence excitement), and Microsoft alone contributed 91% of the gain.
These were Nasdaq (and S&P 500) stocks, but even within Nasdaq only three quarters of the stocks advanced while a quarter fell below their 100 day moving averages. By the end of the quarter, all the money that had been taken out of technology companies since August 2022 had been put back in, but this time into far fewer names. Conversely all the money that had flowed into energy had been withdrawn.
While mega tech was soaring, venture capital funded companies without positive earnings were crumbling, just as we had predicted in our WSJ article a year ago….
[Recent] ups and downs may also have been influenced by the presidential cycle. Historically, government stimulus has made the half year following U.S. midterm elections a particularly good time to be in the market. This continues to a lesser degree right up to the presidential election. No Fed has ever tightened in the six months preceding a presidential election. On the other hand, radical stimulus has now become an all years, all the time phenomenon, which has made this kind of analysis less useful.
The Banking Crisis
The banking crisis is by now old news, but news accounts have missed its essence. The story in short is that Fed attempts to fix the last (Fed-induced) crisis once again backfired. We will try to explain this briefly.
Easy money policies created the dot.com bubble in the late 1990s. When that bubble burst, economist and New York Times columnist Paul Krugman suggested the Fed create a real estate bubble (he used the word bubble) to reboot the economy, which is exactly what the Fed did.
Fed chair Ben Bernanke then compounded the error by attempting some bank reforms designed to make banks “mark to market” all their assets. This had the predictably irrational result that when one bank under stress sold a loan for half nominal value, auditors felt they had to mark down all bank loans. The problem was that most bank loans had no market and thus no identifiable price. All this plunged the financial world into crisis in 2008. When Bernanke stopped being stubborn and finally revoked mark to market on illiquid assets, the stock market rout ended within a few days.
What did the Fed do next? It kept easing and easing long after the crisis had passed. This was ostensibly to support the economy, but was really intended to backstop the treasury market. How does the Fed backstop the Treasury market? It cannot create money out of thin air to buy bonds directly from the Treasury. That is illegal and would also make it too obvious that the government is buying bonds from itself.
Given a legal prohibition of the government selling bonds to itself, the government instead sells bonds to banks which then sell them to the Fed. In operational terms, easing thus means using newly created money to buy from banks a flood of new government bonds used to finance growing federal budget deficits. In return for transferring bonds, banks receive huge fees but must in turn hold trillions of dollars (eventually amounting to over a third of GDP) in initially zero interest cash reserves.
What were the banks to do with all this cash? They could not possibly lend it all out. The Fed did not even want that result, which would have been too stimulative. The Fed then began paying modest interest on reserves for the first time. Banks accepted this but were not satisfied with the interest and began using some of it to replace the treasuries they had sold. In order to boost the tiny increment over reserve interest, they began to buy longer maturities.
This led to the Silicon Valley Bank trap. When the venture market broke, uninsured venture firm depositors at Silicon Valley Bank began a run on the bank, demanding their money back. The bank could only meet these demands by selling the treasuries it held, but these were now far below purchase price because the Fed had forced up interest rates to control inflation, despite promising in 2021 not to do this. In a short time, the bank ran out of capital and had to be closed.
About half of bank deposits are uninsured. Banks in general had greatly increased the risk of a run by paying depositors meager interest rates, far below those available from money market funds, often below 1%. They did this in order to meet “earnings guidance,” but by doing so risked insolvency. Some banks were in far better shape than others and it was the same for “shadow banks” such as brokerage firms. On March 23, Jefferies reported the following:
Unrealized bond losses as % of tangible common equity:
US Bank: 26%
Truist: 30%
Schwab: 185%
Although the banks were enticed into their treasury bond portfolios by the Fed, they could have kept them shorter. This again was a problem of trying too hard to boost earnings. In addition, there are other regulatory barriers to managing a bank bond portfolio. Bonds sit in two buckets, those “held for sale” and those “held to maturity.” The former is marked to market, the latter not. If even one bond is sold from the “held to maturity” bucket, all the bonds there must be marked to market. Banks dared not do this, which also contributed to the recent crisis.
The U.S. government said it would not intervene and then immediately intervened. The government guaranteed uninsured depositors and the Fed opened the discount window (this has erroneously been described as another round of quantitative easing but is not). In a few weeks, three quarters of the sum pumped into banks in 2008 was lent out.
Much of this was done under a new program called The Bank Term Funding Program. The Fed often operates without legal authority under the Federal Reserve Act, for example when it began buying mortgages, and this was another example. The Economist called the program “shocking” for a different reason: it created huge moral hazard for the future. Through this program, which is supposed to be only for a year, but will no doubt be extended, the Fed has now removed all mark to market requirements, even for the highly marketable treasuries, by agreeing to lend based on book, not market value. So, an unrealistic effort to mark unmarketable assets to a fictitious market has now resulted in the end of all marking to market, and in effect the end of all financial discipline.
Interest Rates, Earnings, Recession Risk, Fragility
The Fed currently pays 4.65% on remaining bank reserves, far higher than the recent past. As previously noted, this money is created out of thin air and measurably increases the money supply despite Fed “tightening” and “quantitative tightening.” Despite the cross currents, the money supply has fallen by about 6%, which is unprecedented.
The market is betting that the Fed has stopped raising rates and will start reducing them in the second half. Even if it does, markets do not always respond positively to a cut. They did after the Covid Crash, but quite often markets fall after a cut, because the cut is correctly interpreted as a confirmation of economic weakness.
Interest rates alone are a blunt instrument with which to try to control inflation. If rising interest rates produce recession, that reduces inflation more, although stagflation is always possible. Looking more deeply, rising unemployment and increasing credit spreads bring down inflation more than rising interest rates alone. We have not yet seen much of either.
The Fed’s own preferred measure of (economy not earnings) recession risk according to chairman Jerome Powell is the “near term forward spread.” This is the difference between a three-month and an eighteen-month treasury bill. It is currently quite negative.
Earnings declined at year end. If they decline again with this quarter’s reporting, we are officially in an earnings recession. One must, however, keep in mind that the market expecting flat or negative earnings is not necessarily negative for stock prices. The market looks ahead, so may be focusing on subsequent earnings recovery.”
Another potential tremor to watch is the likely loosening of Japanese control over interest rates. There is now significant inflation in Japan and as rates are allowed to rise there, some Japanese may wish to bring their capital home. They currently have $1.5 trillion in the US, much of it invested in U.S. government securities. If U.S. treasuries do not find buyers, that will put upward pressure on U.S. rates independent of Fed policy. One could even see the anomaly of rising rates during a recession.
Something else to watch is global real estate, which has been in a major bubble. The Green Street Commercial Property Price Index reports that U.S. office space has fallen in value by 25% over the last twelve months and even apartment properties have fallen 21%. The latter figure is partly because of large amounts of supply coming on stream. BlackRock has closed a large real estate fund to redemption and some shareholders are learning to their surprise that they really have no redemption rights.
Many buyout loans look shaky. We have an article on buyouts in the Financial Times that addresses this as of April 19, 2023.