We came very close to a moment when the definition of the word “default” would become very important – and we may come to that moment again, perhaps within a matter of weeks. My definition: Default is what the financial markets say it is. Personally, I hope that the financial markets never need to make that definition specific.
Some background – that is, how the Treasury markets work:
The Treasury sells securities of varying maturities. The shortest-term securities, called bills, have maturities generally of four weeks, 13 weeks, 26 weeks, and 52 weeks. (Treasury sometimes sells even-shorter-term securities, called “cash-management” bills, to deal with minor unforeseen financing needs.) Bills are like “zero-coupon” securities – that is, they are sold at a discount, and the return is delivered when they are redeemed upon maturity at their higher par values. Intermediate-term securities, called notes, have maturities including two years, three years, five years, seven years, and 10 years. Treasury notes bear coupons which are paid twice each year. Treasury also issues 30-year bonds, which like Treasury notes bear twice-yearly coupons. (Treasury also sells Treasury Inflation-Protected Securities (TIPS) of five-, 10- and 30-year maturities, plus various non-marketable securities (including the familiar “savings bonds”) that are not pertinent here.)
The Treasury normally takes some steps to sell securities on almost every business day. Treasury securities are sold at auction; bidders offer a price that they are willing to pay for a specified dollar yield. Treasury announces its tentative financing schedule quarterly, with each tentative schedule extending over the next six months (thus being a partial revision of the previous schedule). Treasury announces the terms of each auction at least one day (for four-week bills) or as much as a week (for longer-term securities) in advance, executes its auctions generally on Mondays through Wednesdays, and then settles the auctions usually on Thursdays (but there are exceptions to this pattern). The major Treasury auction is a quarterly refunding, which establishes ongoing cycles of longer-term securities maturing and being refinanced with new securities. The quarterly refundings include three- and ten-year notes, and 30-year bonds.
So where are we now?
Treasury Secretary Jack Lew has announced that as of October 17, after executing the current financing plan, the Treasury will have exhausted its borrowing authority. That is to say, after using his various extraordinary statutory authorities (see here) essentially to replace internally held government debt (which counts against the debt limit) with informal IOUs (which don’t), the Secretary will have well and truly pushed his head up against the debt ceiling. He will be unable to borrow any net new cash. (In truth, there will be some additional headroom left, but it will be veritable nickels and dimes).
Therefore, without borrowing as a live option, the Secretary will be left with a bare-minimum cash balance, tax revenues coming in, and a greater amount of claims due to be paid arriving daily. The greatest economic power in world history will be living hand to mouth.
What happens next?
The honest answer is that nobody knows. No great financial power, no provider of the world’s reserve currency in markets that trade instantaneously around the globe, ever has been there. The ramifications are so wide-ranging that no one can get his or her arms totally around the situation. I can only give it my best shot.
Perhaps the one certainty is that the federal government’s basic financial infrastructure will be strained as never before.
A fundamental fact: Until now, the Treasury has paid all of its due bills as they arrive every day, and then balanced its checkbook later that night. When you have an ample cash balance backed by a virtually unlimited ability to borrow, you can do that. But starting on October 17, the Treasury no longer will have an ample cash balance, because it will not be able to borrow net new cash. If the Treasury continues its current practice of paying all its bills during the day and then doing the accounting that night, what are the chances that it may discover that it bounced some checks four hours before? Answer: Far, far too great for the world’s greatest economic power. We don’t have a system to give reliable second-by-second cash balances, and we are not going to create one in the next week.
The fallout would extend beyond a potential “oops” end-of-day moment. The Treasury could discover that paying too many of its bills that were due on a Monday and a Tuesday, followed by an unexpectedly weak revenue flow on a Wednesday, could leave it unable to pay a large bill that was due on that Thursday.
The sum and total is, again, that a payments system that was designed for the world’s greatest economic power, which reasonably could be expected never to be cash-short, will not work flawlessly in a hand-to-mouth mode. Presumably, in time, we could build the kind of financial system we would need to count and pinch pennies. But if we go that route at some future date, we should stop claiming to be the world’s greatest economic and financial power.
Which gets us back to the significance of the definition of “default” in the context of the United States Treasury. Some take that word to a fine point, and say that revenue flows are sufficient to pay all interest due on the outstanding debt, and that rolling over the maturing debt is a zero-net-cost operation; and therefore the United States need not “default” in that narrow sense even if left without a debt-limit increase for a very long time. And in fact, the Treasury’s payment system does have two separate channels: all debt finance transactions, which are internal to Treasury; and all other transactions, which come mostly from other agencies. So Treasury literally and operationally could “prioritize” its payments in a limited sense – that is, it could pay debt service first, and then, with what funds were left over, pay what other claims it could. So it is literally true that the Treasury could ensure that the United States did not, in this strict sense, “default on its debt.”
But there are a lot of qualifications.
First, the Treasury would fail to pay a lot of bills, and the stack of unpaid bills would get taller and taller by the day. Then, the definition of “default” would become a real issue. Like the household that would promise to keep its home mortgage current even while it juggled scarce cash and failed to pay its car loan and its credit-card payments, the Treasury would not look very solid to its creditors. This is why default unavoidably would be what the financial markets say it is. If the markets began to react negatively to the Treasury’s cash management perils, it is worth remembering that people who smell smoke in crowded theaters do not typically arrange themselves in orderly two-by-two columns; they run for the exits. Dictionary definitions would quickly become irrelevant.
To put just the roughest order of magnitude on this factor: From the beginning of the month through the middle of this week, before the markets began to smell concessions in the non-negotiations between the House of Representatives and the White House, the annualized yield on 13-week Treasury bills increased by more than 25 basis points (that is, more than one quarter of one percent). That is more than one standard tightening move by the Federal Reserve Open Market Committee. Presumably, if we had stumbled yet closer toward a debt-ceiling collision, that interest-rate increase would have continued. And though it is always perilous to assign causation to market behavior, this episode was tied as clearly as any to the debt-limit standoff.
A second issue is that although the Treasury has a separate channel for its non-debt-service transactions, it has no capability to prioritize among claims within that channel. Treasury has a simple filter to avoid paying claims to entities that are, for example, suspected of fraud; but it has no capability to rank its roughly 80 million payments each month by some measure of priority. And one lesson from the news items emerging from the current shutdown (see here) is that there is a vast number of enormously diverse claims on the Treasury that virtually everyone agrees must be kept current. Prioritizing one such instance leads immediately to prioritizing another, and another, and another. It would be a hopeless task to try to create guidelines for each individual agency to send its claims to the Treasury in some particular order – relative to its own other claims, and relative to all of the claims of every other agency. And all of that complexity does not speak to whether some ostensibly higher-priority claim that is currently due should be paid before some on-the-surface lower-priority claim that was due three weeks ago – noting that both claims were mandated by Acts of Congress, and therefore have equal legal standing.
And speaking of which: Would some legal claimant whose bill was prioritized lower choose to sue, on the ground of being denied equal treatment under the law? Could an entire prioritized Treasury payment system be hauled into court?
But it is time to loop back to the theoretical ability to prioritize the servicing of the public debt. Remember that Treasury revenue flows are highly unpredictable. Suppose that the Treasury decides to pay high-priority non-debt-service claims (perhaps Social Security benefits and salaries for uniformed servicemen and women), and then tax receipts come in lower than expected. The Treasury might then not have the cash needed to pay interest on the debt, even though debt service has its own prioritized separate payment system. Thus, like Arthur Okun’s metaphorical six-foot-tall economist who drowned in a stream that was on average three feet deep, the Treasury easily could fall afoul of data variability and the unpredictability of the future. If the financial markets (and risk-averse investors like insurance companies and pension funds) become concerned about such contingencies, it could easily alter the functional definition of the term “default.”
And one more thought about debt service in the broader sense: The narrowly defined refinancing of the debt (that is, as opposed to borrowing net additional cash) is on the surface a zero-sum swap of securities, and therefore not an increase of debt outstanding; but there are complexities and issues of timing. As was explained at the outset, Treasury bills are sold at a discount. Thus, retiring a $1,000 Treasury bill will require the sale of a bill that will raise somewhat less than $1,000, plus a little bit more. Even coupon-bearing longer-term securities tend to be sold for less than par. The law of large numbers should allow the management of this mismatch, but it adds one more layer of complexity to an already fraught situation.
And as to timing: In theory, refinancing the debt entails swapping one security for another, in one zero-net-debt-change transaction. In practice, to retire a maturing security, the Treasury needs the cash to pay the investor. Thus, technically, the Treasury needs to sell a new bond first to retire a maturing bond later, in two separate transactions. This is a technicality, but we are in a technical situation where the letter of the law matters.
Should the world’s greatest economic and financial power be stuck in this mire? Every citizen is entitled to answer for him- or herself. For sake of discourse, let’s assume that the most common answer will be “no.”
So in the absence of an Act of Congress increasing the debt limit, what can the President and the Treasury Secretary do to avoid such dire consequences? Basically, there are two families of responses that are under consideration.
The first is resort to a part of Section 4 of the 14th Amendment to the Constitution, which reads (in case your pocket Constitution should be stuck in your pocket) “The validity of the public debt of the United States, authorized by law, … shall not be questioned.” This amendment was ratified in the wake of the Civil War, so that the United States could both affirm its constant commitment to honor its debt while denying responsibility for debt undertaken by the newly reunited Confederacy.
Some would argue that this language contradicts the statutory debt limit itself. How can a law restrict the power of the Treasury to honor its debt, when the Constitution itself establishes that power without limit?
In other words, and in a possibly independent argument: If the Congress exercises its constitutional power of the purse to spend a lot and tax a little, how can the Congress then restrict the authority of the Executive to issue the debt that is essential to pay the difference? The Congress thus puts the Executive on the horns of a dilemma: Either the Executive violates the statutory debt limit, or the Executive violates the charge of the Constitution to honor the validity of its debt, authorized by law. This would not be the first time when the Executive had been forced to choose between mutually contradictory legal or constitutional obligations, and the Judiciary has been reluctant to fault the Executive for making one or the other choice. (This interpretation might be challenged as a practice of the law without a license.)
Some would argue that the Executive, so tied in constitutional and statutory knots, should simply issue additional debt – either while invoking the 14th Amendment, or simply citing the dilemma imposed upon it by the Congress and the Constitution.
The President has stated that the Treasury’s attorneys do not agree with this analysis. I do not believe that it is unfair or unkind to say that the Treasury attorneys have a reputation for being cautious. Their argument is that debt issued under such uncertain authority would not be received blithely by the markets. There would be fear that the debt might be ruled invalid in the courts, and that the investors would find themselves holding worthless paper. Again, practicing without a license, the most common analysis would seem to be that no one would have standing to sue to challenge the validity of such securities, on the ground that no one who had not voluntarily purchased those securities could claim to have been harmed. Still, the most conservative analysis might conclude that the Treasury would not want to run any such risk.
A more likely contingency would be that the President would be impeached. One might question whether a House of Representatives that claimed in no way to want to place the United States of America in default would take an action that would threaten at least indirectly to do just that (because the ground for impeachment would be that the President had sold bogus, unauthorized United States Treasury securities). One might also ask whether there would be any chance that the Senate would even engage in a trial. But again, the cautious choice would be to avoid such a contingency in the first place.
But the President and the Secretary might not enjoy such a clean option. If the choice is between failing to pay the nation’s bills and running such legal and constitutional risks, the President and the Secretary might determine that the less-imprudent course would be to issue additional debt and invoke the 14th Amendment – regardless of what the President has said publicly to this moment.
The second broad class of options for the Executive is asset sales. The Treasury could sell (or sell and lease back) just about any piece of federal property. Any private (or other government) investor with some confidence could be willing to deliver cash to take ownership of real estate, military hardware, the gold in Fort Knox, or just about anything else – in return for a piece of the action. That cash would replace an equivalent amount of additional Treasury borrowing – although because the transaction would be (shall we say) irregular, it surely would cost the Treasury more than would equal borrowing through the sale of Treasury securities.
A particular form of asset sale that has achieved considerable attention lately is the minting by the Treasury and sale to the Federal Reserve of a platinum coin. The Treasury is authorized by law to mint platinum coins (see here), according to specifications chosen by the Secretary. Thus, the Secretary could mint coins in denominations of his choosing, and sell them to the Federal Reserve to repay existing Treasury debt. That would give the Treasury headroom under the debt ceiling to sell new debt to the public to raise net new cash.
No Treasury Secretary would choose to undertake such a transaction under normal circumstances; but these are not normal circumstances, and some might argue that the platinum coin would outrank a failure to meet the Treasury’s obligations in full and on time. Again practicing law without a license, there would seem from the statute to be no prohibition against such a transaction, so there might be less legal vulnerability to the Secretary and the President than invocation of the 14th Amendment. Yet there is little doubt that such a financial transaction would seem to the rest of the world to be beneath the United States of America – but again, perhaps not as far beneath the United States of America as playing the deadbeat.
If pushed to the wall, what would the President do? I have no inside information, but the platinum coin feels like too much of a gimmick. Other assets sales seem less likely, because they would entail both greater transactions costs and – perhaps most importantly – time to execute, in a world where the Treasury necessarily will be counting coins and watching the second hand on the clock. The worst outcome – one that I believe no President and no Treasury Secretary of either political party ever would (or should) accept – would be to allow what they would perceive to be the first default (see this for a technicality) by the United States of America. Accordingly my money – whether in cash or Treasury securities – would be on an appropriately solemn speech, invoking the 14th Amendment and citing the Catch-22 that the Congress’s inaction has forced upon him, in which the President accepts the responsibility to sell bonds in excess of the statutory debt limit.
Our nation may have dodged this bog; but that is not yet certain, and we may find ourselves back in this unseemly place in just a few weeks. For myself, I hope that I will not need to research this topic further.