Greetings!
Part one of the QRA is now out and Treasury has provided us with their updated Q1 and new Q2 financing numbers as well as their end of qtr projected TGA balance. This article will review those numbers, explain how they relate to what I projected in my post last week and address the question of whether we will see continued coupon increases in the QRA and recommended financing tables released Wednesday morning.
First though, some quick housekeeping. I have renamed this blog/newsletter/substack to Reverse Engineering Finance. When I started writing 18 months ago, I quickly needed a place to publish long form explanations (initially QT) that just wouldn’t work in a twitter thread, and put no thought into the name of it, thus John’s Musings. But to be blunt it was an awful name that did little to describe what this newsletter seeks to do or be. I am correcting that now. Ill keep the FKA in the title for a while for continuity’s sake and drop it at some point in the future. Hopefully if you recommend this newsletter to your friends or colleagues in the future you wont have to justify to them why they should read random musings by some guy named John.
Also, I reckon most folks reading this also know about and have watched Jack Farley interview me on the latest episode of Forward Guidance. If not, check it out here
. We go into depth about this QRA, QTs end game, the Total Treasury model etc. Jack is a phenomenal interviewer, I super enjoyed and appreciated the opportunity to chat with him. (and the that shot of me and Janet is the current leader in the clubhouse for next years Christmas card)
Now on to the financing estimates announcement.
So what did Treasury actually tell us yesterday?
1. The expected financing need in Q1 – 760b
2. The assumed end of Q1 TGA level – 750b
3. The expected financing need in Q2 – 202b
4. The assumed end of Q2 TGA level – 750b
Sure, there is other information in the announcement but that’s what really matters to understand what the issuance composition is likely to look like. Four pieces of the puzzle now filled in that I think helps us better predict the issuance composition that will come out in the recommended financing tables tomorrow for Q1 and Q2.
First though, lets review how my previous predictions stacked up against actuals.
Q1
My projected borrowing estimate = 780.87b vs. QRA of 760b.
I overprojected Treasury’s borrowing need by 21b over with both Treasury and I assuming and end of Q1 TGA level of 750b.
My projection split between 340.89b in net new coupons and 439.98b new bills. Treasury wont confirm the Q1 coupon issuance until tomorrows recommended financing tables (and actual refunding announcement) but if Treasury does not change bill issuance between now and end of March they will end at +422 new bills end of Q1/ Thus I am very confident that the coupons for Q1 will be as projected and my 21b overprojection is accounted for with 18b fewer bills. Uhh but John, didn’t Treasury just this morning announce one further 5b increase to the 4w bill. Ahh, astute reader you are. Yes they did and at first glance this puts a fly in the ointment. But only so much, because while that increase adds 20b net new bills to the quarter, Treasury can get it back in the last week of March by reducing bill issuance across the tenors in the normal 5b (shorter tenors) / 2b (longer tenors) pattern. Since I had identical decreases slated for the first two weeks in April, its easy to just slide them up a week. This is now what I think will happen. In a literal sense, the extra 20b from the increased 4w bill slightly nudges up the possibility of a downside coupon surprise tomorrow, but that outcome remains solidly in the highly unlikely camp imo, given the other information (1 qtr of similar increases etc.). So I reiterate my coupon forecast for Q1 made in last weeks post.
Why would Treasury want that extra 20b now? My guess is to create a little extra cushion to remain safely above the TGA minimum policy “high point” in late February. Tax refunds start to flow out of the TGA in February, this extra 20b provides additional cushion if they flow extra heavy/earlier.
Q2
My projected borrowing estimate = 269.8b vs QRA of 202b
So it would seem I overprojected Treasury’s borrowing need by 68b but that is actually not the case. Why? Because I incorrectly assumed an end of Q2 TGA level of 875b. Treasury did not, continuing to assume a 750b level. Removing the 125b of my expected financing required for a TGA build that isn’t going to happen, my adjusted borrowing estimate was ~145b or -57 vs. the QRA. Or put differently I underprojected Treasury’s financing need by 57b in Q2 (after accounting for the same lvl TGA end of quarter), which seems to be much closer to actuals than the “street consensus” estimate was. I think that results reflects the value of the ground up approach to modeling each of the individual flows that the Total Treasury Model takes.
Ok but what does that say about the thing everyone really cares about tomorrow? That being the coupon issuance amounts for Q2 of course.
Well, nothing directly, but it does change some of the parameters for my model. Starting with, it pegs the end of second quarter TGA level at 750 (instead of my previously assumed 875) So we need to find an issuance pattern across bills and coupons in Q2 that ends there, still carries the TGA over the high point in the policy minimum level for the TGA in late May, accounts for the difference implied between Treasury’s financing need and the current output of the withdrawal/deposit flows of my model (I assume that will play out as a little less tax receipts in April and June then my model projected) and generally fits with gradual and predictable issuance.
Well I found one. Two across the tenors cuts to bills the last week of March and first week of April, followed by one more cut the week of May 21/23 seems to do the trick, but it does that trick with no increases to coupons. So am I now projecting no coupon increases in Q2? Not exactly. I think my position is now very much on the middle of the fence with a teensy slight lean to no coupon increases because the issuance pattern seems to work. I haven’t found one yet that works nearly as well that includes another round of coupon increases in Q2 though there may well be one. I need to develop a quicker way to test the different issuance configurations which is something I will work on before the next QRA cycle but for now Im running out of time and I want to get this piece out where folks have a chance to read it before it stops being relevant. So the overall change in my position is really from expecting another round of coupon increases in Q2 (last week) to being very much unsure today.
To quote the specific language from the TBAC report to Secretary Yellen from the Nov QRA “At this point, the Committee expects that the need for similar further increases at the Q2 FY2024 meeting is likely, but increases beyond Q2 FY2024 may not be required”
The attorney in me can plausibly argue both that this implies increases in Q2 Calendar Year 24 (which is Q3 Fiscal Year 24) and that it does not. We will find out tomorrow which it was.
Thank you for reading!
John
Great work John and congrats on the interview. Can’t wait to see how Q2 issuance impacts RRP.
Thanks as always John. Question for you. Have you modeled out how the bill/coupon share will evolve if there are no further increases to coupon auction sizes after the next increase (ie CY Q1) vs one further increase in CY Q2?