The size of the Tarp turns out to be one of the small things congress hasn’t agreed on yet. There was talk of a Republican move to halve its size late yesterday.
Indeed, it might well be that the whole package is about to fall through.
If – and it’s now a big if – the bailout happens, here’s a few back of the envelope calcs from Michael Cloherty at Bank of America (emphasis ours):
The Troubled Asset Relief Program means the Treasury faces a massive cash outlay in the coming year as assets are purchased from banks. Once the assets are in Treasury hands, the assets will generate a cash flow that reduces Treasury financing needs in the coming years. The exact cash flow to the Treasury will depend on the average price the Treasury pays, CPRs, delinquency/default rates, recoveries, and when the Treasury starts to sell assets back to the market. This means the range of error in any analysis is extremely wide. Nonetheless, it is clear that the TARP leaves the Treasury with a colossal financing need in FY09, followed by a return to a manageable issuance need in FY10.
There is though, a problem. The Treasury bill market has already seen a “massive supply shock” this year. As this graph illustrates:
The bottom line of the analysis is that the Treasury does not have an easy task ahead in financing the bailout. The increase in supply will likely push bills beyond their “normal” investor base and will put the market under “considerable strain” according to BoA.
It’s evident that the law of unintended consequences is very much on the cards.
We cannot think of any solid metric for gauging where bills outgrow their normal investor base and the marginal investor becomes a CP buyer (which would make bill yields move toward LIBOR). However, our guess is that the market would show significant signs of strain if it grew another $400bn, putting bills held privately (Fed holdings excluded) at three times its level in June 2007.
This would leave roughly $525bn to be handled with increases in coupon issuance. The Treasury has already signaled a move to monthly 10yr auctions and quarterly bonds, but these will only raise roughly $54bn per year. Introducing a $20bn quarterly 3yr note would bring the residual financing need to slightly over $390bn.
In addition, we could see the Treasury create an issuance window which would offer an unlimited amount of supply at a given yield (either an absolute level, or at a spread to the Fed’s Constant Maturity rate similar to a savings bond). Callables and TIPS might be good vehicles for a window.
From a long term horizon, this is all pretty significant in its effect on Treasury yield curves. On the one hand, notes BoA, the US could finance through massively increased offerings in the existing auction cycle. The effect of that would be long-lasting disruption to the curve. On the other hand, the Treasury could introduce a series of off-cycle “orphan issues”. Those though, would have a “significant impact” on the futures market:
The contract specifications for 10yr futures require that the deliverable must be a note (old bonds are not deliverable) and that the remaining maturity as of the first of the delivery month be at least 6 ½ years but not more than 10 years. If we assume that the Treasury issued a one-time 7yr with a November 15th maturity, that issue would be the CTD for the December 10yr contract if it was 18.7bps cheap on a forward basis to the 4.5s of Nov 2015 (using the September 24th closes). Since this issue would have a shorter duration than the 5.125s of May 2016, we would expect Dec 10yr futures to trade somewhat cheaper and with a somewhat shorter duration if the Treasury starts talking about funding the TARP with an off-cycle 7yr.