Sometimes the bond market gets it wrong too.
Earlier today, when Trump "flipped" on the GOP and aligned with congressional Democrats at a White House meeting to fund the government and raise the debt ceiling through Dec. 15. despite objections from virtually all Republicans, the threat of a late September/early October debt ceiling crisis disappeared. The bond market realized this first by sending the yield on the October 5 BIll tumbling from 1.20% to 1.00%, as repayment on this maturity was no longer in jeopardy.
However, at the same time as the October T-Bill yield was tumbling, bond traders sent the December 21 T-Bill yield surging, since after all all Trump had achieved was kick the can by three months...
... which in turn inverted the Oct 12-Dec 21 curve.
There"s one problem with this kneejerk reaction: it was only half right, because while bond traders were right to buy the October bills, they made a mistake in dumping the December Cusip.
Why? The answer lies in the cash flow calendar.
Recall that the reason why early October Bills were being sold is not because the "debt ceiling" would be breached - that happened in March of 2017, with a last minute agreement in May kicking the can through September, which in turn has now been pushed through December) - but because early October is when the Treasury"s "X-Date" would finally be hit: that"s the date when the Treasury would run out of cash and emergency measures.
The problem is that the upcoming three month extension to December 15 is not also an extension of the X-Date: there will be a buffer of at least a few months between December 15 and when the next X-Date hits.
As Jefferies" Ward McCarthy explains, Congress’s standard operating procedure for addressing the debt ceiling over the past few years has been to suspend it for a defined period of time. Assuming that this agreement centers on a suspension of the debt ceiling through December 15th, the following will happen once President Trump signs the bill:
- The “debt issuance suspension period”, the process by which Treasury accesses the “extraordinary measures”, will end. Treasury can then replenish the G-Fund, CSRDF, and EFSF with the issuance of nonmarketable securities that they have been redeeming since mid-March. Effectively, this re-loads the extraordinary measures for when the debt ceiling is hit again in mid-December. It
- Treasury will ramp up issuance of bills in order to replenish supply in the market and boost their cash balances. Under normal circumstances, Treasury maintains a minimum cash balance of $150 bln, but cash balances had fallen to as low as $32.1 bln as of September 1st. This is an uncomfortably low level and Treasury will waste no time in re-establishing a cash buffer.
- Lastly, there will be no debt limit. However, Treasury will be bound to issue only that which is necessary to meet current obligations. The translation of that is that Treasury will have to have the same or less cash on hand on December 15th as they did upon the signing of the bill. So, no matter how many bills are issued in September, October and November, Treasury will have to make potentially massive paydowns heading into December 15th.
So if December 15th is not the next "X-date", what happens then, and in the weeks after.
- The debt ceiling will be re-struck at the current level.
- Repeating the above, Treasury will need to bring their cash balances down to the same level (or lower) as was the case at the beginning of the debt ceiling suspension period.
- Treasury will once again declare a “debt issuance suspension period” and have access to the extraordinary measures. It is too early to tell how much will be available at that time, but around $350 bln seems like a reasonable baseline. This estimate will change as we get closer.
In other words, instead of the $32 billion in effective cash and equivalents as of today, the Treasury will have over $300 billion in dry powder, more than enough to buy the Treasury 2-3 of more months, and to take the government well into 2018 before a debt ceiling crisis reappears on the horizon.
Lastly, when will this issue come to a head again? Here"s Jefferies" estimate:
The answer to that question is highly dependent on estimating borrowing needs for Q1, which is always a dangerous exercise since Treasury pays out the lion’s share of tax refunds in January and February. If tax refund outlays are relatively high and Treasury needs to issue a lot of debt in order to fund them, the new “drop-dead” date could come as soon as late-February. However, if Treasury is able to get through Refund Season and make it to the April 15th income tax receipts, then the “drop-dead date” would be a few more months out, potentially the end of Q2, beginning of Q3.
Putting it all together, it is still too early to tell the specifics, "especially because there is a non-zero chance that this “agreement” reached by Trump and Congress falls apart." After all, as Jefferies snydely notes, only a few hours ago Paul Ryan was on television describing a 3-month extension as “unworkable” and characterized the action of tying debt limit legislation to an aid bill as “disgraceful”. Shortly after, the President endorsed it.
In a worst case scenario, the early-October bills would come right back into the cross-hairs although for right now, the crisis appears averted, if only temporarily. One thing, however is certain: there is no reason at all why the December 21 Bills should have been sold off. While it is much too early to determine what is the correct "Fulcrum" Treasury Bill should today"s extension pass - it could be an issue maturing in Q1, Q2 or Q3 2018 - it is effectively "free money" that any purchases of the December Bills that were sold today at the closing price, will be a profitable investment.
So yes, sometimes even the bond market gets it wrong.
