Updated: 4 days ago
June 20, 2022 | Q: Is the Federal Open Market Committee preparing to sell mortgage-backed securities (MBS) from the system open market account (SOMA) via that most notorious (and useful) of all products, the CMO or collateralized mortgage obligation? Is this a good idea? Yes it is, but hold that thought…
During the most recent FOMC press conference on June 15, 2022, Federal Reserve Board Chairman Jerome Powell described the mounting carnage in the house industry with considerable detachment. Maybe the folks at the Fed’s Board of Governors in Washington think they are playing with a science experiment in high school and not an industry that impacts every American. Said Powell:
“Recent indicators suggest that real GDP growth has picked up this quarter, with consumption spending remaining strong. In contrast, growth in business fixed investment appears to be slowing, and activity in the housing sector looks to be softening, in part reflecting higher mortgage rates. The tightening in financial conditions that we have seen in recent months should continue to temper growth and help bring demand into better balance with supply.”
Headline: The housing sector is in recession and accelerating down in terms of volume and headcount. Of course, it was the FOMC that encouraged an increase in supply of mortgage loans by dropping interest rates several hundred basis points in March of 2020. The industry increased headcount and operating expenses by 50% in response.
Now in 2022, the residential mortgage industry is downsizing as falling revenue and operating expenses converged in Q1 2022. We described the unprecedented change in the secondary loan market last month in The Institutional Risk Analyst (“FOMC vs TGA; PennyMac Financial & United Wholesale Mortgage”). Message to Chairman Powell: How is this helpful?
We also commented last week on some related matters in National Mortgage News (“Mortgage industry liquidity risk returns”). What should investors and risk professionals focused on the mortgage industry expect in coming months? Something like 2019, the worst year for secondary market profitability in a decade, only a bit worse in terms of the speed of change. Large issuers will survive, small issuers will become canon fodder for the Fed’s decidedly idiosyncratic policy machine.
First, the latest up and down cycle in mortgages is more extreme than the 2002-2006 bubble, as shown by the historical data from the MBA. In 2020 and 2021, we saw larger lending volumes but fewer, bigger loans. Average loan size has grown more than a third since 2008 and this increase has accelerated over the past five years, but drew little notice from the FOMC or the Fed’s staff in Washington. Home prices are up more than 200% since 2008. Below are the MBA’s historical annual production number going back to 1990.
Source: Mortgage Bankers Association
Because of the magnitude of the Fed’s error in terms of running QE purchases too long and at too high a level, the mortgage market ran very hot in 2020 and 2021, leading to bigger peak lending volumes than in 2005. Again, in real, inflation adjusted terms, the 2002-2006 bubble is bigger and, significantly, was comprised 50% of private label loans. Today’s residential loan market is 99% government-insured and conventional loans.
Another way to think about the problem created by the FOMC is home prices. The average purchase mortgage is now $450,000 today vs less than $300,000 a decade ago. The chart below from FRED shows the Case-Shiller Index going back to before the crisis. In nominal terms, the most recent skew upward is bigger, but adjusted for inflation it is actually about the same magnitude increase as the 2002-2006 housing bubble.
The next question to ask is about profitability. There are many ways to show this, but the bottom line is that production profits in 2022 will be about a tenth of the peak rates seen in 2022-21. The weak profitability is a function of the fact that refinance loans will drop sharply. The two charts below tell the story. First is the MBA’s production profit and the second is the MBA’s forward production estimates. Below is MBA’s “Chart of the Week”.
As you can see, production revenue is about equal to expenses. Many firms are losing money. Joe Garrett of Garrett McAuley in San Francisco said in a note: “2022 is about costs more than revenue. Put another way there’s a limit to what you can do about revenue. There’s a lot you can do about costs." The MBA’s Marina Walsh and Jenny Masoud put the surge in production costs in perspective:
“While lower production revenue contributed to scant profit margins, the primary driver was cost. Total loan production expenses – commissions, compensation, occupancy, equipment and other production expenses and corporate allocations – increased to 345 basis points in the first quarter, from 315 basis points in the fourth quarter. On a per-loan basis, production expenses ballooned to a new study-high $10,637 per loan in the first quarter, up more than $1,000 per loan from the fourth quarter and more than $2,500 per loan from one year ago.”
Again, the inflation of lending costs since 2008 speak to a secular inflation that ought to be of interest to the Fed and the research community. The MBA’s actuals and projections for 2022-2023 are shown below. The MBA has production cratering this year and bouncing in 2023 as capacity runs off and pricing power begins to be restored. Likely the FOMC will also be lowering interest rates in 2024 before the presidential election.
Source: Mortgage Bankers Association
At present, there are a number of marginal lenders making loans in competition with healthy companies. Just as thrifts doubled down in the 1980s, independent mortgage banks (IMBs) facing extinction will fight to the bitter end.
The marginal lenders must exit the business before the healthy lenders can restore profitability. That is the harsh reality of the mortgage business, but amplified thanks to the FOMC and quantitative easing (QE). The dying among lenders will make loans until they are completely dead, preventing the industry from re-pricing efficiently until next year.
In general, the annual loan coupon paid by a consumer is 1% higher than the debenture rate on the MBS that will eventually finance the loan. With the industry moving to 5% or 5.5% coupons for future MBS issuance, implying loans with at least 6-7% annual percentage rates, lenders that have been writing loans in the 5s will take losses when the loan is sold. The TBA swaps page from Bloomberg as of the close on Friday is shown below. Notice that the 4.5% contracts for Fannie Mae MBS for delivery in July are already trading three-quarters of a point below par.
Stronger lenders can lose money on a new mortgage loan at close and at the point of sale, but recoup by collecting servicing fees and the occasional refinance opportunity. IMBs without servicing portfolios are not long for this brave new world fashioned by the FOMC. The plummeting production levels due to higher interest rates and the spread compression caused by industry overcapacity will require a one-third reduction in lending capacity in 2022.
Our view of home prices, which is informed by conversations with some of the smartest operators in the industry, is that sharply falling lending volumes are unlikely to affect home prices initially. Sticker shock will slow demand in some cases, but pent-up demand for homes remains powerful and yet to be even partially satiated.
Home prices will plateau for a few years, then pop up again when the Fed drops interest rates ~ 2024, followed by a macro reset for the housing market a la 1988 and 2008 in 2026. Home price cycles tend to run about 20 years. That reset in 2026 could force down home prices to 2019 levels – but again inflation may change this calculous.
Finally, both stock prices and bond yields for the mortgage industry suggest that the 1-4 sector is headed for a serious restructuring in 2H 2022. Mortgage issuers with term debt are trading at the widest spreads in years. We expect to see some forced sales and bankruptcy filings in the next several months. The big issue facing the IMBs and the mortgage industry as a whole in 2023 is loan delinquency, which we expect to rise significantly as the economy slows and dips into a recession.
As usual, Chairman Powell saved his best comment on the housing sector for last:
“[T]he supply of finished homes, inventory of finished homes that are for sale is incredibly low, historically low. So it's still a very tight market. So prices may keep going up for a while even in a world where rates are up, so it's a complicated situation. We watch it very carefully. I would say if you're a home buyer, somebody or a young person looking to buy a home, you need a bit of a reset. We need to get back to a place where supply and demand are back together and where inflation is down low again and mortgages rates are low again so this will be a process whereby we ideally we do our work in a way that the housing market settles in a new place and housing availability and credit availability are at appropriate levels. So, thank you very much.”
Listening to Powell, it’s as though we all live in Superman’s Bottle City of Kandor, which sits on a shelf at the Fed’s HQ on Constitution Avenue. Of note, Chairman Powell had nothing to say about providing liquidity to the market for government loans and Ginnie Mae MBS as the economy slides into recession. As we’ve noted several times in past comments, credit costs for 1-4s are likely to spike significantly higher as the economy slows.
Neither was there any discussion in the press conference of sales of MBS from the Fed’s system open market account or SOMA and how this might further adversely impact the already reeling mortgage market. Nobody in the Big Media thought to ask Chairman Powell about the most important factor in US monetary policy, namely the SOMA portfolio. That omission, however, may be quite significant.
The Federal Housing Finance Agency (FHFA) just announced what may seem a purely technical change to the pricing for collateralized mortgage obligations (CMOs) issued by Fannie Mae that contain Freddie Mac MBS and vice versa with Freddie Mac as the issuer:
“Effective market-wide July 1, 2022, Fannie Mae will begin charging a new fixed upfront fee to create certain Supers® and real estate mortgage conduit (REMIC) securities that have Freddie Mac Uniform Mortgage-Backed Securities (UMBS) collateral underlying those Supers and REMIC securities. A fee of 50 basis points will apply solely to the portion of the Supers and REMICs that are Freddie Mac UMBS collateral. This fee is a one-time fee applying to Freddie Mac UMBS collateral that has never incurred such a commingling fee from Fannie Mae in the past.”
In other words, if the FOMC wants to sell some of its low coupon MBS now in the SOMA via a comingled Super CMO containing Fannie Mae and/or Freddie Mac collateral, then the pricing is in place to cover the cost of the insurance and other expenses. Since both Fannie Mae and Freddie Mac are in runoff in terms of their own balance sheets, they are unlikely to be issuers of CMOs on their own account. And these could potentially be very large, investment grade deals.
By allocating part of the duration risk among investors with progressively longer investment timeframes, the stand-alone CMO that we described in the book Financial Stability actually "kills" part of that duration risk. Issuing CMOs via the GSEs could help the FOMC out of its trading mess, but represents a monumental irony. By entering the CMO market in order to hide growing losses on low-coupon MBS, bonds acquired during the latest round of QE at 103-104 but now trade in the low-80s, the Fed will concede the reckless nature of its policies.