Two years ago, the financial markets were flooded with billions of newly minted dollars care of the Federal Open Market Committee in response to COVID. Banks put aside $50 billion in loan loss reserves for potential defaults on mortgages and everything else due to the pandemic, but thanks to the FOMC, the credit crisis was delayed for a couple of years.
Because of the Fed’s outsized response to COVID, interest rates on loans and also mortgage backed securities plummeted and home prices soared. For a while, mortgage bankers were selling government loans into Ginnie Mae 2% coupon MBS. There are even a few Ginnie 1.5s kicking around. These securities will be around for many years, even decades to come, as prepayments slow to a crawl and extends the maturity of the MBS.
The U.S. central bank owns about two-thirds of this ultra-low coupon COVID vintage production of MBS, of note, roughly one-third of total System Open Market Account (SOMA) holdings. How and when the Fed sells or does not sell part of its $2.7 trillion mortgage bond trove is a topic of conversation among the chattering classes on Wall Street, but for mortgage bankers it is a crucial question. The table below shows the total SOMA holdings:
The surfeit of income in the mortgage sector cause by the growth in the SOMA was a gift from the Fed. But what the Lord giveth, the Lord may also taketh away. Now the Fed is worried about inflation and is slowing purchases of MBS to only reinvestment of dwindling principal repayments. Soon the Fed’s SOMA will be running off, with no reinvestment of principal prepayments from Treasury securities or MBS.
Because the FOMC is raising the target for Fed funds, the on-the-run contract in the to-be-available market has gone from 2% and 2.5% MBS coupons to 4.5% and 5%. TBA pricing has gone from 3- to 4-point premiums to a discount. The 10-year Treasury is nearing a 3% yield, a rather dramatic reversal from the under 1% performance during 2020, and rates are moving higher. Meanwhile, mortgage lending volumes are falling.
With
The big threat to the mortgage banking community comes from any future Fed sales of MBS. Sales of low-coupon MBS from the SOMA will push mortgage rates higher and faster than already seen. These low coupon, high beta securities are more volatile than the broad market, difficult to hedge, and will soon be underwater in terms of net carry.
Bank of America projects $595 billion of net issuance and $2 trillion of gross MBS issuance in 2022, according to an April 8 research note. So while the secondary mortgage market is shrinking compared with last year, the Fed may eventually try to sell amounts of older MBS that is equal to a large fraction of current production. Fed sales of MBS will also hurt banks. Many banks that chose to retain low-coupon loans and MBS will soon be deeply underwater as durations explode.
The prospect of large Fed sales of MBS is worrisome for the mortgage industry and housing regulators, but selling low coupon mortgage debt is also a problem for the central bank. Every time the Fed receives a prepayment or insurance payment at par for a Ginnie Mae 2 that originally cost 104, for example, the FOMC is recording a four-point capital loss. If the Fed today sells a Ginnie Mae 2% MBS at 92, for example, it will record a 12-point capital loss.
Another problem for the industry arising from the end of QE is what to do with loans that were modified during the period of low interest rates during the
Servicers can modify a COVID forbearance loan more than once, but rising interest rates are forcing some consumers into a full-doc process. Imagine you have a borrower that was on COVID forbearance for 18 months, then calls and says that they are still experiencing hardship and ask for another COVID modification. You modify the loan and put the arrearage on the back end, but keep the old coupon rate on the new loan in order to "help the consumer."
When you try to sell that newly modified 3% loan into the current market, the price is in the low 90s and the servicer is facing a $10,000 cash loss on the sale of the loan. Will the Federal Housing Administration or the GSEs reimburse the servicer for the loss? Specifically, will the FHA allow more than one partial claim for COVID loans? The answer seems to be no, but nobody in the industry is quite sure. This is an especially amusing question given that many observers are calling for a reduction in FHA mortgage insurance premiums on government loans.
Once President Biden declares an end to the COVID emergency, the special servicing waterfalls for pandemic-affected loans also go away. Government loans that qualified for COVID forbearance will go down the traditional waterfall for FHA foreclosure, creating opportunities for the Consumer Financial Protection Bureau to penalize servicers for following the FHA rules regarding loss mitigation.
Although QE provided some respite for the U.S. economy during 2020 and 2021, now the bill is coming due as the FOMC pivots aggressively to fighting inflation. Rising mortgage rates will lower lending volumes, but increase the value of conventional and government mortgage servicing rights.
Whereas investors considered the option to refinance loans as part of the calculus in purchasing MSRs at premium prices over the past year, today that option is worthless. Now considerations about the cost of loss mitigation on both conventional and Ginnie Mae MSRs are coming back into focus. Loss mitigation, lest we forget, used to be about protecting note holders and taxpayers, not consumers.
Industry insiders tell NMN that roughly a third of the government loans that were modified during COVID will re-default, meaning that servicers which bought Ginnie Mae MSRs at a premium valuation may soon experience rising expenses and possibly negative returns. Conventional MSRs will likely feel this impact as well, thanks to QE, since many FHA borrowers were able to migrate to conventional loans without mortgage insurance.
The moral of the story is that Fed’s policy of QE is a decidedly blunt and double-edged instrument. The Fed drove down mortgage interest rates during 2020 and 2021, pushing asset prices up and making default rates negative. Now this process is reversing and the challenges of the next few years could be as difficult as were the benefits of QE over the past 24 months. At the end of the day, there is no free lunch even if the Fed pretends to pick up the tab.
