Why are spreads widening despite a lower probability of Fed sales?
There has been heated market debate about potential outright mortgage sales by the Fed this year. We believe Fed Chair Powell’s comments at September’s FOMC meeting put that speculation to rest for the time being, specifically, his remark that [sales are] “not something we are considering right now and not something I expect to be considering in the near term.” Part of the Fed’s rationale is that the doubling of mortgage rates this year has contributed to a major slowdown in the housing market. As such, there may not be a need for additional tightening at a time when the risks of such an unprecedented move appear to outweigh the likely benefits. We believe the Fed will continue to use interest-rate hikes as its primary policy tool to tighten financial conditions.
Some have speculated that the Fed may seek to impose a floor on the size of its balance-sheet reduction. Recall that it now has a cap: If its portfolio’s prepayments rise above US$35 billion per month, it will reinvest the difference, capping its monthly balance-sheet reduction. A floor would ensure the opposite, with Fed holdings shrinking by at least US$35 billion per month. This would have the Fed selling agency MBS equal to the shortfall, in prepayments below that threshold. We do not think this is likely.
We anticipate natural paydowns in the current rate environment to be in the range of US$20-25 billion per month. To reach the floor, the Fed would need to sell close to US$10-15 billion of mortgages per month, or around US$150 billion per year. This is not a trivial amount for the market to absorb, but it’s not large enough to meaningfully change the trajectory of the Fed’s US$2.7-trillion mortgage portfolio. Based on this asymmetry, it does not seem worth introducing a new tool with limited benefit to a market already worried about liquidity.