§ Given the 50bp move higher in rates over the last few weeks, we analyze MBS convexity risks. Even though the structure of the mortgage market has changed significantly from the pre-crisis period due to Fed ownership and GSE conservatorship, we think that there are still some holders of MBS convexity risk that would need to hedge the duration extension of their portfolios due to the sharp move higher in rates.
§ We estimate paying needs of $46bn 10yr equivalents due to the 50bp rise in rates since the election. The peak of MBS convexity is about 50bp higher from here, and we see paying needs of $10bn in 10yr equivalents for every 10bp rise in rates until the primary mortgage rate gets to 4.5%. Any further increase in rates above 4.5% should result in less convexity selling.
§ Convexity paying should argue for higher rates led by the belly of the curve, wider 5-10yr swap spreads, higher levels of implied vol and richening of the payer skew. We suspect some of this flow has gone through, even though the move in swap spreads looks counterintuitive. Even though convexity paying should have move spreads wider, the flow was more than offset by fears of higher deficits (which should tighten spreads).