Trevor Cranston: Hey, thanks. Good morning. Can you talk a little bit more about the movement in the duration gap positioning during the quarter and the decision to shift to a slightly negative position? As a second part of that, if you can maybe comment on if you think going forward, agency spreads are likely to remain correlated with the directionality of rates. Thanks.
Peter Federico: Oh yes, that’s a really good question. Let’s start with that, because that’s really a significant factor, and that’s going to ultimately dictate to a large extent where we operate from a duration gap perspective, is what do we think mortgages are going to do with respect to interest rates and that correlation. Chris can talk a little bit about how that correlation has been and what we expect it to be going forward, and what that means from a hedging perspective.
Chris Kuehl: Yes, so as Peter mentioned, our duration gap as of today is slightly positive, right around zero given the move higher in rates since year end. But given the shift in sentiment with rate cuts now on the horizon, mortgages are trading shorter than they did last year, much of the last year and a half relative to model durations, which makes sense as some of the higher rate tail scenarios have been sort of clipped or had become less probable, and so given that mortgages can trade to shorter durations and with our portfolio having somewhat more call risk than extension risk, we’re likely to try to maintain a near zero to slightly positive duration gap in the current environment. I’d say fast forward another six to 12 months, and if we find ourselves at rate levels with significantly–you know, lower rate levels with significantly lower refinancing activity and the Fed still running off its portfolio, then I think a case could be made for running a significantly longer duration gap as mortgage supply, both organic and Fed run-off, would be much more sensitive, or spreads would be much more sensitive to rate levels with spreads likely widening in a rally, tightening into a sell-off.
But for moderate moves in rates from here over the near term, I think spreads are less correlated with rates than they have been for the last year or so, and so we’re likely to try to maintain close to a zero duration gap to a slightly positive duration gap over the near term in the current environment.
Trevor Cranston: Got it, okay. That’s very helpful, thank you.
Operator: Our next question comes from Rick Shane from JP Morgan. Please go ahead with your questions.
Rick Shane: Thanks, guys, for taking my questions this morning. Look – you talked a little bit about the call risk within the portfolio. When we look at Slide 23, there are probably twice as many coupons on there as there were three years ago. The issue is that stack gets bigger and we’re in this incredibly strange environment, there’s much, much more concentration of call risk at the higher coupon. You guys were pretty deliberate about moving up the stack and being willing to realize losses when coupons were moving. Should we expect that you will start to realize some gains, and again I realize it’s sort of indifferent economically, but will you start to move back down the stack in order to mitigate some of that call risk?
Peter Federico: Yes, sure. I’ll let Chris talk about that.
Chris Kuehl: Yes, our weighted average coupon is still quite low relative to the production coupons – I think it’s just over 480. The bottom line is the convexity risk on our portfolio has gotten worse, given the 100 basis point rally over the last quarter, but it’s still very low and very manageable. Generally speaking, sourcing convexity through pool selection is still very attractive. We did add a small receiver option during the quarter as vol traded lower, but implied vol is still relatively–you know, it’s still very high relative to historical norms, and so again we’re likely to manage the negative convexity on the portfolio simply through delta hedging in this environment and through asset selection on the pool side, which almost always is cheaper than buying explicit protection on rates through the options market.
If implied volatility declines materially, you could see us do a little more on the rate option side, but over the–you know, for this environment, convexity risk, very manageable largely through delta hedging and asset selection.
Peter Federico: And just to add to that, Rick, when you look at that Page 23, you can see we provide our positions in those coupons, and you can see that the biggest positions are really in the 5% and 5.5% coupon, so you think about those positions relative to the current mortgage rate, which is still between 6.5% and 7%, to Chris’ point, in a sense we’re more exposed to the middle coupons as opposed to the highest coupon, so we still have a lot of room to rally in the mortgage rate before those refinance.
Chris Kuehl: Just real quick, what I’d add is when you look at the percentage of specified pools in some of those higher coupons, look at the end notes on the composition, so 47% on average specified pools, but the next 20, actually close to 30% of the portfolio is also in pools with favorable convexity characteristics. They’re just not the lowest loan balance cuts. They’re specified pool characteristics like low FICO, certain geographies, investor or second homes, modified pools. We do have a–you know, again, the call risk on the portfolio in the current environment is very manageable, given the composition of our holdings.
Rick Shane: Got it. Yes, it’s interesting – when you look at the portfolio in aggregate, I agree with you 100%, and definitely note the concentration in the 5s and the 5.5s, very slight amortized cost premium there. I was just more curious on the tail, the 6 and the 6.5s, there is more premium and perhaps a little bit more risk in this particular environment.
Chris Kuehl: Yes, I mean, the last thing I’d say on the prepayment environment – I mean, when you step back, obviously the vast majority of the universe has almost no incentive to refinance, 98% of the universe. But the origination over the last four or five months is going to be very interesting just to observe speeds, given that on average that cohort will have a 50-ish basis point incentive to refinance, and lenders certainly have capacity, they’re probably going to be willing to work thin to capture what little volume there is, and so that would argue for pretty fast speeds on the highest coupons that were originated, call it though the August through November time frame, but there are other arguments as to why the S-curves could be quite a bit flatter as well relative to what we saw in 2020 and ’21.
