Bank Earnings and QE/QT
New York | In the most recent edition of The IRA Bank Book, we note that the rate of increase in funding costs for US banks in 2018 was a bit over 70% year-over year. The rate of change in this key component of bank earnings is not particularly correlated to the broad swings in market interest rates and, by implication, investor confidence. But the normalization of bank funding costs is relentless. As we describe in some detail in the IRA Bank Book for Q1 2019, rising bank interest expense is a structural trend that heralds the end of extraordinary monetary policy, at least for now.
Whereas the quarterly cost of funding for all US commercial banks was just shy of $40 billion at the end of 2018, by the end of 2019 funding costs for the $17 trillion in US bank assets should be closer to $70 billion, especially for the larger banks. Asset earnings are rising at one quarter of the rate of funding expenses, BTW, the result of a continued sellers market in collateralized loan obligations (CLOs) and straight debt. At year end 2018, the average cost of funds was well over 1.3% for large banks such as Citigroup (1.27% Q4’18) but the average for the top 100 banks is still below 1%.
Despite the full stop retreat by the Federal Open Market Committee, the impact on the bank earnings is likely to be negative, albeit gradually. The fact of a flat yield curve and continued downward pressure on yields for loans and securities is not particularly helpful when it comes to keeping asset returns even with increase funding costs.
The chart below shows our projections for bank interest income and expense from Q1 2019 through Q4 2019. Note that we see net interest income stop growing in Q1’19 and begin to shrink in dollar terms as the year continues. We currently project that US bank funding costs will exceed $70 billion per quarter by year-end 2019.
Source: FDIC/WGA LLC
"[T]he Fed already seems to have embraced the idea that inflation should be allowed to exceed 2% without immediately triggering a tightening," notes Nouriel Roubini over the weekend in Project Syndicate. Roubini correctly notes that with the addition of Governor Richard Clarida on the FOMC has the effect of “tipping the balance of the FOMC in a more dovish direction.”
Additional evidence of the change in the Fed's policy orientation toward greater intervention is found in the March 20, 2019 document released by the FOMC entitled “Balance Sheet Normalization Principles and Plans.” The FOMC has slowed the rate of decrease in the central bank’s balance sheet. Specifically:
The Committee intends to slow the reduction of its holdings of Treasury securities by reducing the cap on monthly redemptions from the current level of $30 billion to $15 billion beginning in May 2019.
The Committee intends to conclude the reduction of its aggregate securities holdings in the System Open Market Account (SOMA) at the end of September 2019.
The Committee intends to continue to allow its holdings of agency debt and agency mortgage-backed securities (MBS) to decline, consistent with the aim of holding primarily Treasury securities in the longer run.
In addition, the increasingly dovish FOMC is going to continue reinvesting in both long-dated Treasury securities and agency RMBS. Again, the FOMC:
Beginning in October 2019, principal payments received from agency debt and agency MBS will be reinvested in Treasury securities subject to a maximum amount of $20 billion per month; any principal payments in excess of that maximum will continue to be reinvested in agency MBS.
Principal payments from agency debt and agency MBS below the $20 billion maximum will initially be invested in Treasury securities across a range of maturities to roughly match the maturity composition of Treasury securities outstanding; the Committee will revisit this reinvestment plan in connection with its deliberations regarding the longer-run composition of the SOMA portfolio.
The upshot of all this is that the FOMC is going to continue the ill-advised policy of Operation Twist, buying long dated securities as part of the asset mix. It is not clear, for example, how the FOMC will reinvest MBS in Treasury securities. By purchasing long duration paper for the system open market account (SOMA), the FOMC directly contributes to the further flattening of the yield curve and maintains an effective cap on bank asset returns. This is why we see bank industry NIM growth going negative after Q1 2019.
Yet of note, in the very same document, the FOMC states: “It continues to be the Committee's view that limited sales of agency MBS might be warranted in the longer run (emphasis added) to reduce or eliminate residual holdings. The timing and pace of any sales would be communicated to the public well in advance.” That's nice.
Meanwhile, even a relatively modest $15 billion per month in net runoff from the SOMA will still squeeze overall liquidity in the system. As George Selgin at Cato Institute noted back in 2016 (citing Walter and Courtois) once the Fed started paying interest on excess reserves (IOER) banks “would be more willing to hold on to excess reserves instead of attempting to purge them from their balance sheets via loans made in the fed funds market." But the level of excess reserves in aggregate is still a function of the size of the Fed's securities portfolio.
The decision to slow (and eventually reverse) the rate of SOMA portfolio shrinkage indicates that the more dovish FOMC has decided to make permanent the nationalization of the short-term money market in the United States. The FOMC accomplishes this end by continuing the reverse repo approach to managing interest rates and dealing with a larger number of potential non-bank counterparties, such as money market mutual funds. Why bother with those messy private markets, eh? As with IOER, the Fed is subsuming the functioning of private unsecured markets for cash onto its own federal balance sheet.
When the FOMC mechanically reduces excess reserves by allowing the SOMA securities portfolio to shrink, the big banks will return to private markets and bid more aggressively for REPO and other low-risk capital weight substitutes, driving yields down. As we noted last week, the FOMC’s plan to maintain a “ceiling” on short-term rates via IOER may be doomed to the same fate as the Fed’s plan of a year ago to reduce the size of the balance sheet. In a market where the 10-year Treasury note already is trading below 2.5% yield, a ceiling may become a floor very quickly.
Markets are likely to be a bit sloppy this week as investors digest the latest statements from the FOMC. Of course, the FOMC under Chairman Jay Powell is committed to full and complete public disclosure. But the markets may not care for the message – especially as the decidedly mixed outlook for financials become more apparent. Thus the markets will tighten and yields will fall, as started in earnest in December 2018 and has continued since then. But bank funding costs will continue to rise.