‘This Is More Unpredictable Than Usual’: Goldman Is Worried About The Debt Ceiling

Ok, lots of folks are worried that “this time is different” for the debt limit debate.

And indeed this time is different. Because Donald Trump is president. So you know, there’s no precedent for anything anymore.

A couple of days ago, we posted a simple chart from Deutsche Bank that shows what the VIX did in two previous fiscal showdowns. Here’s that chart again for anyone who might have missed it:

“Equity markets reacted during [the] two [previous] episodes as VIX increased from 13 to 20 in 2013 and from 16 to 25 in 2011 (and rose further as the eurozone crisis worsened in August 11),” Deutsche wrote, in the color that accompanied that visual. “So far, nothing seems to be priced in this market.”

Right. And indeed “nothing is priced into this market” would accurately describe how investors are pricing any and all potential event risks. Simply put, they’re not – pricing anything in that is.

Well, Goldman’s Jan Hatzius and his 6 dwarfs are out on Wednesday explaining why when it comes to the debt limit, this time is “more unpredictable than usual.” Here is their simple assessment:

The Treasury projects that the debt limit will need to be raised by September 29, and we expect Congress to take at least that long to raise it. At this stage, we believe Congress will combine the debt limit increase with a spending bill that must pass by September 30 to avoid a government shutdown.

Below, find some details from the note…

Via Goldman

When does the debt limit need to be raised?

The Treasury has projected that Congress will need to raise the debt limit by September 29, which corresponds closely with our projections over the last couple of months. While the fluctuations in daily tax receipts and outlays make precise projections impossible, our current estimate, shown in Exhibit 3, suggests that the Treasury will be able to make payments scheduled on September 29 and can probably make payments scheduled on October 2, but runs the risk of exhausting its funds by October 3.

What is the plan to raise the debt limit?

Trump Administration officials have endorsed a “clean” debt limit increase, meaning that Congress should not attempt to negotiate a package of spending cuts or other gestures of fiscal responsibility alongside the increase. That said, debt limit increases are generally passed as part of larger legislative packages and this looks likely again this year. Congressional appropriations expire at the end of the federal fiscal year on September 30, and Congress will need to pass a spending bill to avoid a partial federal shutdown starting October 1. In addition, several other programs need to be reauthorized by September 30, including programs under the Federal Aviation Administration (FAA) and the Children’s Health Insurance Program (CHIP). Given the proximity of these deadlines, it seems very likely that the issues will be addressed as part of the same legislation.

The increase looks likely to pass very close to the deadline. In the past, many conservative Republican lawmakers, particularly in the House, have opposed debt limit increases and spending legislation and it is unlikely that Republicans could muster a majority in favor of either this year. Moreover, since the debt limit and the spending bill will need 60 votes to pass in the Senate, at least 8 Democratic votes would be needed in that chamber even if all 52 Republicans supported them; the compromises needed to win 8 Democratic votes might reduce Republican support further. We expect Republican leaders to exhaust other legislative options before passing legislation that a large number of Republican lawmakers oppose. Complicating matters further, the House of Representatives is scheduled to be on recess the week of September 18-22.

The debt limit and extension of spending authority might be addressed only temporarily. While the Trump Administration and congressional Republicans would undoubtedly prefer to pass a full-year spending bill and a long-term (e.g., 2-year) debt limit increase, a short-term extension might be all that is possible. On the spending side, Congress usually fails to enact a full-year appropriation ahead of the start of the new fiscal year and an extension until December or so is typical. While the debt limit is usually extended for longer periods, a short-term extension is a clear possibility this year as well.

Overall, the outlook for the debt limit feels a bit more uncertain this year than it did the last time around in 2015, in light of the new political dynamic under unified Republican control of government, a new President whose approach to the issue is not yet clear, and a number of substantial policy debates, like tax reform and ACA repeal, in the background.

What happens if Congress does not raise it in time?

Scheduled federal payments would be delayed and financial markets would be disrupted. The primary consequence of a failure to raise the debt limit before October 2 would be a likely failure to make some payments on that day or soon thereafter. This could include, for example, the main monthly payment to Social Security beneficiaries, which occurs on the 3rd of every month. All told, the decline in federal payments that would be necessary to avoid breaching the debt limit would be roughly equal to the budget deficit, resulting in a temporary fiscal contraction of 3-4% of GDP for the period the debt limit is binding. A sharp decline like this would likely be disruptive to financial markets and could have consequences for the real economy, which is why Congress has managed to avoid such an outcome in the past.

That said, there is probably less risk that scheduled Treasury coupon payments might be missed, for two reasons. The first is the particular timing of this year’s debt limit deadline. Coupon payments on 2-year, 5-year, and 7-year notes are scheduled to be made October 2 (the usual payment date of September 30 falls on a Saturday). Absent a debt limit increase, the Treasury might have enough funds to make this and other payments expected that day, though the variability in the Treasury’s daily cash flows make it hard to be certain. The next interest payment is not due until October 16, on 3-year notes and 5-year and 30-year TIPS, and it seems very likely that Congress will have increased the debt limit by then. Exhibit 4 shows the projected daily cash flows from late September, by category, based on the same period in 2016, adjusted for calendar effects and growth in various types of revenues and outlays.

The second is the potential for prioritization of payments. The transcript from an FOMC conference call held August 1, 2011–just ahead of the debt limit increase that year– suggests that prioritization of payments could be feasible. Fed officials described procedures that would allow principal and interest on Treasury securities to be made on time while delaying other payments. This would involve rolling over maturing securities into new issues and delaying non-interest payments, even when funds were available to make them, in order to conserve cash for interest payments. While the transcript resolves some of the uncertainty regarding the technical feasibility of prioritization of payments, it is far from clear that prioritization of payments–for example, delaying benefit payments to make interest payments–would be politically sustainable.

Are there any signs of a market reaction thus far?

The Treasury bill curve has already started to reflect some aversion to holding bills that mature shortly after the deadline. As shown in Exhibit 5, market participants appear to have initially expected the deadline later in October, but have adjusted their expectations in light of the Treasury’s guidance to Congress on July 28 that the debt limit would need to be raised by September 29. That said, the kink in the bill curve is still fairly minor so far compared to the experience in 2013, for example, when bills maturing just after the stated debt limit deadline yielded as much as 50bps more than comparable securities that matured slightly earlier.

But then again, this should be no problem, right?

I mean after all…

******