Stocks are flattish as we await the Fed decision. Bonds and MBS are down small.
The Fed decision is due at 2:00 pm EST today. Kevin Warsh will hold a press conference afterward. The market is looking for a hawkish pause, where the Fed maintains policy but signals further hikes ahead.
New Home Sales fell 0.6% MOM and 5.3% YOY to a seasonally adjusted annual rate of 680,000 units. At the end of September, there were 476,000 new homes in inventory, which represents a 8.4 month supply. The median new home price fell 3.9% YOY to $400,400.
The inventory of new homes for sale is still elevated:
Housing isn’t going to be much of a contributor to GDP growth until some of this inventory gets worked off.
Microsoft is offering buyouts to about a quarter of its global workforce – 60,000 employees in all. The company is expected to report earnings today. Capital expenditures for AI are taking a toll on the company’s finances, so it is cutting its workforce aggressively. I wonder if Microsoft is the first salvo in tech downsizing. If so, we could see the labor market break out of its low hire / low fire equilibrium and weaken.
Home prices rose 2.1% YOY according to the Case-Shiller Home Price Index. The hip-to-be-square trade continues with Chicago seeing the biggest gains and Seattle seeing the biggest losses. Real, inflation-adjusted home prices continue to decline, however affordability remains stretched as incomes haven’t caught up with the 2022 spike in prices and mortgage rates continue to work their way higher.
Mortgage applications rose 3.4% last week as purchases rose 4% and refis rose 3%. “Mortgage rates declined for the second week in a row, with the 30-year fixed rate decreasing to 7.37 percent, the lowest in three weeks. The decline in rates gave borrowers some relief, with applications rebounding slightly, particularly conventional applications,” said Joel Kan, CMB, MBA’s Vice President and Deputy Chief Economist. “Mortgage rates are still more than a percentage point higher than the same week last year and most homeowners still have rates much lower than prevailing market rates. The ARM share of applications increased to 11.3 percent, the highest level in more than a year, as more borrowers applied for ARM loans to lower their monthly payments.”
Purchase applications are down 18% YOY while refis are down 60%.
________________________
During the MBA Conference, there was a lot of chatter about non-QM loans and speculation that some aggregators lost money during the month by being underhedged. I wanted to take a look and see if that was true. If you aren’t a secondary guy you will probably find the following of limited use.
A typical mortgage banker hedges a pool of Fannie / Freddie or Ginnie loans with TBAs. Non-QM aren’t deliverable into TBAs so the lender needs to think about duration risk, which refers to a loan’s expected life and its interest rate sensitivity.
During September, rates moved up violently, and many lenders with lower-coupon non-QM loans found they were unable to find an acceptable bid for them. Basically anything with a coupon below 7% was highly unattractive to the buy-side due to extension risk and pricing was hit accordingly. Does that mean that the hedges didn’t work? Not necessarily.
Below is a snippet from a well-known NQM correspondent’s DSCR rate sheet on 8/31 and 9/30 for a few coupons. You can see how much the lower note rates got slammed (the last column):
Below is a table of the Eris SOFR Futures prices based on 8/31 and 9/30 prices
The lower coupon 6.5s probably have a duration of 10 years while the 8s have a duration of 5 years or lower. If you have a Bloomberg terminal you can calculate them directly, or your broker-dealer may be able to give you an estimate. Knowing the duration of the underlying loans is a critical component to hedging non-QM loans, and getting this wrong will affect hedging performance.
Below I looked at how Eris SOFR futures performed as a hedge for these loans. The 6.5 loan had a 10-year hedge, the 7 had a 7-year, the 7.5 had a 5-year, and the 8 had a combo of 5s and 2s.
The lower coupons struggled and shows why lenders may need some out-of-the-money options to hedge the tail risk of a big move. But for most coupons, the hedges were within 25 basis points. An important note about duration and extension risk: Duration is not a set-it-and-forget-it number. It changes with rates. This is what happened to the lower coupons. They might have a duration of 10 years with the 10 year at 4.8% and a much longer duration when the 10 year hits 5.3%.
Like any mortgage backed security, non-QM loans have negative convexity and therefore hedged portfolios will be vulnerable to interest rate shocks. Non-QM loans are hedgeable, but they have to be handled differently than conforming loans.
If you have any questions about hedging non-QM with Eris SOFR futures, please contact John Douglas john.douglas@erisfutures.com or Geoffrey Sharp geoffrey.sharp@erisfutures.com at Eris Futures.
