The weekend has been consumed by speculation over what might happen to Silicon Valley Bank’s depositors. On Sunday night we finally found out. Jason Calacanis saved the financial system.
However, one little-discussed issue that still warrants attention is the impact on mortgage-backed bond markets, and the aftershocks that could cause elsewhere.
SVB is still sitting on a $50bn book of MBS. It might need to dump those bonds to help give depositors some of their money back if it isn’t sold — or possibly even if it is. That has already caused ructions in the US mortgage market on Friday, when other investors understandably rushed to avoid getting squashed by SVB’s potential MBS dump.
You won’t see this cleanly in the headline price of MBS or overall mortgage rates because Treasuries also rallied hard due to the obvious flight-to-safety dynamics. But spreads exploded on Friday as SVB went down the drain.
Here’s JPMorgan before the government’s intervention on Sunday night:
Though we think the collapse of SIVB has relatively limited implications for the broader financial system, the announcement on Wednesday that it had liquidated its AFS portfolio and the subsequent receivership announcement left the mortgage market wondering if (when?) the ~$50bn HTM MBS portfolio might hit the market. Mortgages sharply widened on Friday in anticipation of a liquidation, with lower coupons 10bp T-OAS wider and production coupons 4-5bp wider as of mid-afternoon. That leaves lower coupon 30yrs at the cheapest levels (high 30s T-OAS) that we’ve seen in a while and the current coupon basis similarly wide at 50 T-OAS. The gradient of spread changes on the stack reflects the market’s take on the coupon distribution of SVB’s HTM holdings. Given that their portfolio grew sharply during the QE4 period, and that their holdings appear to be tilted toward shorter-term, lower book yield MBS, it makes sense that the pain in mortgages was felt more acutely down the stack and in 15yrs. While it is unclear at this point what will happen with SIVB’s holdings, it seems likely that the bonds, or at least their duration, are going to come back to the market in one form or another. Valuations do look attractive to us across most of the stack at this point, but there are clearly unknowns and stresses in the market that are hard to quantify.
Most people don’t need to care about this — and the government’s efforts could make it a moot point — but there is one area that was already feeling the heat and could end up being incinerated if we have more days like Friday: mortgage REITs.
MREITs are like traditional real estate investment trusts, but they buy individual mortgages and MBS instead of property. They hedge out their duration risk, meaning they are purely exposed to mortgage spreads. MREITs are much smaller than the broader REIT industry, but they’re chunky players in mortgages as they use a lot of leverage to invest in what is usually pretty steady-eddie, low-return stuff.
For example, Annaly Capital Management, the biggest one, had total assets of $81.9bn at the end of 2022. AGNC Investment Corp has $51.7bn, The Starwood Property Trust manages $28.3bn, Rithm Capital has $32bn, and the Blackstone Mortgage Trust another $26.8bn.
Now look at what the share price of some of the public ones did late last week.
The problem is that with a lot of leverage, it might not take many days like Friday before their capital is toast? And then we could face a tsunami of MBS portfolios washing over the market. This is probably at least part of the explanation for why the US government acted so forcefully.
Now, Friday was exceptional, and initiatives like the Fed’s Bank Term Funding Program — allowing banks to offer Treasury and MBS collateral to the Fed at their par value — means that things could be entirely nipped in the bud. As DB’s George Saravelos said this AM:
. . . taking a step back, the Fed’s new Bank Term Funding Program can be interpreted as re-establishing a temporary QE program by offering to absorb UST and MBS from banks at above-market prices. This is in effect a self-regulating break to QT whereby if funding pressure becomes too acute as bond prices sell-off, liquidity gets reinjected into the system offsetting the reserve drain from QT.
But just because SVB’s depositors will be made whole doesn’t mean that this saga can’t still end up causing aftershocks.
