Originally Published: October 2021
The US debt ceiling is in the news again due to Congressional gridlock, along with the possibility of a US sovereign default and the unintuitive idea of the US Treasury Department minting a trillion-dollar platinum coin to bypass the problem.
This article breaks down some of the nuances involved in this strange situation, which comes up every few years.
As of this writing, the US Treasury expects to run out of money by mid-October unless the debt ceiling is raised. However, I’m writing this article in the more evergreen sense so that it can mostly apply to all future debt ceiling situations as well, rather than just this particular event.
The Debt Ceiling Backdrop
Congress is the political body that authorizes spending by the US federal government.
Prior to 1917, Congress authorized individual bond issuance to supplement tax revenue to fulfill specific spending allocations. Eventually, beyond a certain scale for fiscal spending during World War I, this practice became administratively unsustainable.
From 1917 onward, Congress instead would allow the US Treasury Department to issue bonds as it sees fit, albeit constrained by a Congressionally-set debt ceiling. Put simply, Congress (the Legislative Branch) stopped micromanaging Treasury bond issuance to fund spending authorizations (thus leaving it as a function of the Executive Branch) but still retained its authority to ensure a division of powers by limiting the total amount of debt issuance, and still has to authorize federal spending.
According to data going back to 1960 by the US Treasury, Congress has raised or extended the debt ceiling 78 times, including 29 times under Democratic presidents and 49 times under Republican presidents.
A handful of times in recent years, most notably in 2011 but a few other times as well, Congress used the debt ceiling in order to pressure a presidential administration to either extract a bargain or for narrative gain.
Often, when politicians oppose raising the debt ceiling, they do so within the narrative of fiscal constraint, but that is mostly political theater. The debt ceiling on its own is a separate concept from new spending authorization, although it can be combined into the same piece of legislation. Raising the debt ceiling itself doesn’t authorize new spending; it merely allows the government to continue paying its previously-authorized spending obligations. To not raise the debt ceiling means the government either has to not pay previously-authorized spending obligations, or default on its national debt, which is accumulated over time from previously-authorized spending obligations.
In other words, logically-speaking, the idea of fiscal constraint is relevant when deciding on new spending and taxation plans, but not relevant during debt ceiling disputes, which are about prior spending plans. To promote fiscal constraint, politicians can vote against increases in spending, or propose new legislation that reduces existing ongoing spending obligations.
However, out of practicality, the debt ceiling is also a tool that can be used for this sort of purpose, even though it’s technically only about whether the government will pay previously-authorized spending obligations and honor its debt.
For this reason, although it has been in place for over a century, there is a lack of full legal clarity about whether the debt ceiling is even constitutional. During 99% of the time, it is not controversial, and so it has not made its way to the US Supreme Court to be ruled on.
There’s a sentence in the 14th Amendment to the US Constitution that says:
The validity of the public debt of the United States, authorized by law, including debts incurred for payment of pensions and bounties for services in suppressing insurrection or rebellion, shall not be questioned.
When push comes to shove, if the US federal government is on the brink of nominal default due to being constrained by the debt ceiling from issuing new net Treasuries to pay off prior maturing bond issuance and fund existing obligations, the President could potentially cite the 14th Amendment, direct the Treasury Secretary to ignore the debt ceiling and issue net new Treasuries in excess of Congress’ debt ceiling limit in order to service previously-authorized spending obligations and national debt, and then let the US Supreme Court work it out later when this decision is challenged.
Would that presidential decision be upheld by the Supreme Court? I’m not a constitutional lawyer so I don’t know. Either way, it’s always unclear what a specific sitting president will do, and as this article discusses, there are a range of outcomes.
What Would a Sovereign Default Entail?
The US federal government is monetarily sovereign, meaning it issues its own currency, and its debts are denominated in its own currency.
In theory, it should therefore never outright default on its debts, since it can print the currency that its debts are denominated in.
Decades and centuries ago, when dollars were backed by gold, the US could potentially run out of gold.
Indeed, the US eventually did default on those gold standard systems by devaluing the dollar relative to gold in 1934 and outright eliminating any gold-backing of the dollar whatsoever in 1971, without compensating the existing holders of dollars and Treasuries for these sudden changes of contract. The government didn’t have enough gold to keep up with the spending plans they wanted to do without changing the definition of what a dollar is.
In the age of fiat currency since 1971, where a dollar is not directly backed by anything in particular or guaranteed to have any specific value, the government can’t run out of its own printed dollars. That says nothing about the value of those dollars, however. If too much money is printed, it can result in price inflation or in extreme cases, hyperinflation. For a monetarily sovereign country, debts can always be paid back nominally, but not necessarily with the same purchasing power as when they were issued. For example, the purchasing power of US Treasuries fell rapidly in the 1940s inflationary decade and the 1970s inflationary decade.
This chart that I update occasionally shows the historical yield of the 10-year Treasury note (blue line) along with the forward 10-year annualized inflation-adjusted return from the start of that year (orange bars). Those two periods where the orange bars were negative, were periods of significant Treasury security devaluation:
Data Sources: Robert Shiller, Aswath Damodaran
Basically, the fiat currency system still operates to some extent as though the government is constrained by money supply. In the context of a gold standard, in order to retain confidence in the backing of a currency, government spending is bound by the combination of taxation and debt issuance. The government must collect gold or gold certificates from the population through taxation and debt issuance, and then spend it back into the economy, and thus has constraints.
After the gold standard was defaulted on by every country, the system nonetheless continued to operate with that mindset for decades, that government spending must be equally offset by a combination of taxes and debt issuance. Currency is still treated as though it is semi-scarce, in other words, and there are limited ways to create it.
Eventually, MMT advocates and some other economists started to say the quiet part out loud; that in this purely fiat system, taxation and debt issuance are not strictly necessary for government spending anymore, and are just done to keep inflation in check. The government’s ability to spend is not bound by tax collection or debt issuance, other than by its own laws. Those laws merely serve as guardrails to keep the money supply from rapidly inflating (and instead, just moderately inflating). In that sense, the government creates an amount of money by spending it into the economy, and separately destroys an amount of money by taxing it and issuing bonds to the private sector or foreign sector.
This starts getting violated if the central bank creates new base money and swaps it for a big chunk of that sovereign bond issuance, meaning that the government spends money into the economy without extracting it from the economy (but instead extracts it from a void of newly-created dollars). Unless that process is offset by private credit destruction (as it generally has been in Japan), then money supply growth can steeply accelerate.
The country, however, still can outright nominally default on its debt even in a purely fiat currency regime if it handcuffs itself with legal constraints and then sticks to those constraints (i.e. doesn’t cite the 14th Amendment, doesn’t use the trillion-dollar coin loophole, etc).
Specifically, the US government has defaulted once within the existing fiat system. In 1979, the US Treasury experienced a brief outright nominal default, when it failed to pay back T-bill holders on time due to a combination of political gridlock and an IT systems failure. After a delay, those payments were indeed made, and then after some conflict and legislation, the Treasury also paid additional interest to make up for the delay. So, that was a relatively minor default in the grand scheme of things.
The US government came close to similar technical defaults a few times since then, especially in 2011.
The 2011 event infamously caused S&P Global Ratings, one of the three big credit rating agencies, to cut the credit rating of the US government from a perfect AAA score to the second-best score of AA+. To this day, they still rank US sovereign debt as being of slightly higher default risk than a number of other countries (Australia, Denmark, Germany, Lichtenstein, Luxembourg, Netherlands, Norway, Singapore, Sweden, and Switzerland), as well as slightly higher than two corporations (Microsoft and Johnson and Johnson). To that credit rating agency, these moments of gridlock put a US temporary default in the realm of possibility, despite the country being monetarily sovereign.
The impact of such a default depends on the scale and duration of the default, ranging from almost nothing to widespread calamity.
Brief Default Scenario
When the debt ceiling is reached, the US Treasury still has some options. For one, they have their primary cash balance that they hold at the Fed, which consists of money that they already collected from issuing bonds, but haven’t spent yet. They can draw that cash balance down to zero before truly running out of money. They can also perform some inter-government borrowing, basically shuffling the books around.
If politicians go beyond the event horizon where the debt ceiling has been reached for a while and there is no further room by the Treasury to draw down its cash balance or juggle its books around, and if the President doesn’t invoke the 14th Amendment or use any loopholes to direct the Treasury Secretary to bypass the problem at that point, then some payments will not be made. A challenging aspect is that it’s not clear ahead of time exactly on what day that will happen; the Treasury Secretary provides estimates based on incoming taxes, outgoing expenditures, and her department’s ongoing emergency measures.
When they run out of options and hit a threshold of no available cash, the US Treasury would have to prioritize what it wants to keep paying on time vs what it delays/defaults on. They would still have incoming tax revenue, but that tax revenue is insufficient to cover ongoing expenditures without continual net new Treasury issuance, and so with the debt ceiling legally blocking them from issuing net new Treasury debt, a large subset of things would start being defaulted on, with the scale of the problem escalating over time.
This could mean Social Security checks don’t go out on time or in full, and a subset of retirees or disabled people that rely on timely checks without savings, would be in financial trouble. Or it could mean that some Medicare payments get delayed. It would likely entail disorderly shutdowns and missed paychecks for government workers, including potentially soldiers, but unlike prior government shutdowns, would not be contained to just these government workers.
Alternatively, or in addition to those types of missed government expenditures, it could result in missed interest or principal payments for some holders of Treasury securities.
If it went on for hours, it would be historically noteworthy like the 1979 default, but likely not functionally significant. If it went on for days, the impact could be historically significant for the confidence of the US political and financial system, but not necessarily an unrecoverable disruption in and of itself from the first-order effects alone.
If it went on for weeks or more, the situation would begin turning dire. Politicians of various types would likely be pressured by donors and constituents to fix things rapidly. The longer it goes, the likelihood of serious damage increases. With no mitigation plans in place it would be catastrophic, but various mitigation plans could extend the amount of time that the financial system can continue to function.
The US Federal Reserve formed an emergency playbook to try to mitigate this brief Treasury default scenario back in 2013 in terms of the functioning and solvency of the financial system, although their exact course of action would depend on decisions by their leaders at the time. In the playbook, the Fed would ease regulations for banks, allowing the banks to basically pretend that any defaulted Treasuries they own were not defaulted on, as it relates to their capital ratios and other requirements, until the situation is resolved. They would also perform various open-market operations as needed to keep payment systems functioning for a time. If needed, a more controversial part of the playbook is that the Fed could also buy defaulted Treasuries and sell non-defaulted Treasuries on the secondary market, to bring the problem onto its own balance sheet and off of purely-private balance sheets. The reason that second part is controversial is that it more directly invalidates the concept of central bank independence, with the Fed financing defaulted US debt.
From the Treasury’s perspective, when choosing which Treasury securities to default on, the Treasury could also aim most of their defaults on the specific bonds that the Fed is holding whether the Fed likes it or not. This keeps the problem mostly “in the family” to minimize real-world impacts until it gets resolved. This also basically invalidates the concept of central bank independence, since the Fed is technically insolvent on its official balance sheet once about $40 billion of its Treasuries don’t have value. In practice, however, they would just ignore that fact until the default is resolved, and consider it a technical quirk.
Basically, if mitigated in certain ways, a short temporary Treasury default has no meaningful impact, other than threatening confidence of the system, which is hard to measure. The longer it goes, and the less well it is managed, the more damaging it gets, as the mitigation efforts by commercial banks and the central bank get harder and harder to do.
Permanent Default Scenario
If the US were to hypothetically default on all or a large chunk of its Treasury debt and leave it that way for a very extended period of time or indefinitely (a permanent broad default), then just about every commercial bank in the country would be rendered insolvent, since they hold Treasuries for a significant component of their risk-free capital. Bank capital ratios, almost without exception, would be insufficient to prevent their liabilities from exceeding their assets if their Treasuries cease to have value, resulting in a disorderly financial collapse of the entire US banking system.
The US Federal Reserve’s balance sheet consists mostly of Treasuries and mortgage-backed securities on the asset side, and physical currency and bank reserves on its liabilities side. So if Treasuries were broadly defaulted on in the permanent sense, the central bank would also be completely insolvent because they would have a $5+ trillion hole on the asset side of their balance sheet. They are not subject to the same rules as commercial banks in terms of solvency, since they are at the core of the system, but the foundation of the institution would be rendered void in terms of both legal accounting and logical underpinning.
The US dollar itself would cease to have conceptual meaning beyond that point other than some momentum from habits of use, since dollars are liabilities of commercial banks (bank deposits) and liabilities of the US Federal Reserve (physical currency and commercial bank reserves), and all of those institutions would now be insolvent without the asset side that consists significantly of Treasuries. It would be like an upside-down pyramid, crumbling after the underlying capstone is removed.
Here is a useful diagram, but I would draw it upside down, since the capstone actually holds up an inverse pyramid:
Diagram Source: Nik Bhatia, Layered Money
In addition, many foreign central banks hold US Treasuries as a large part of their foreign-exchange reserves, and this section of their reserves would be rendered worthless, which would likely affect the value of their own national currency. The complex and non-transparent offshore dollar system would likely implode as well, rendering many foreign banks insolvent, rippling into the value of their national currencies from that angle too.
The outcome of that type of broad and permanent US Treasury default would be catastrophic, to the point where people in the US and elsewhere would likely benefit from having non-perishable food and clean water on hand, along with physical valuables to trade, if the financial system stops working for a time. ATMs and credit cards could fail, global commerce would be crippled, cash itself might be problematic, etc. It would basically be like a financial nuclear bomb going off.
Eventually, some foreign markets could start to recover first, like perhaps Russia that has mostly de-dollarized by now, or nearly-insolvent emerging markets that have more dollar liabilities than dollar assets anyway and on paper would benefit from a destroyed US dollar. However, with the disorderly halting of the US economy and global economy by that point, it’s unclear what would happen to even those sorts of countries, since they are still interconnected to the world via trade.
Since it’s a rather absurd scenario to begin with, the potential outcomes of it happening are also absurd and highly variable based on decisions that occur in the weeks that follow. The eventual beneficiaries would likely be individuals, institutions, and countries with hard assets and enough self-sufficiency to get through whatever period of economic and societal chaos ensues until some semblance of order is cobbled back together in the global financial system.
To be clear, that sort of widespread permanent nominal US sovereign default isn’t really on the table of options from a mere debt ceiling issue. However, it’s not hard to see why countries that are not on particularly friendly terms with the US are increasingly diversifying their reserves so as not to be insolvent if they are hit with a targeted Treasury default (e.g. the US Treasury could selectively default on the Treasury securities held by certain countries).
The $1T Dollar Platinum Coin Loophole
Apart from invoking the 14th Amendment and seeing how the Supreme Court interprets that later, it has been proposed a number of times that the President could also direct the Treasury Secretary to mint a unique platinum coin, worth any arbitrary amount of face value (often $1 trillion is the default assumption), and deposit it with the Federal Reserve at that arbitrary face value. The Treasury would then be able to continue paying its pre-authorized sending obligations.
Wait, what? And why platinum? Let’s back up a second.
Due to checks and balances that exist within the laws of the financial system, the Treasury has strict limits on how it can create currency. However, it has less restrictions on coinage specifically.
For precious metal coins, which the Treasury still mints, gold and silver coins are defined by law as being limited to face values of $1, $5, $10, $25, or $50. A modern 1-ounce gold coin issued by the Treasury usually has a face value of $50, even though the gold content is worth around $1,700 and the coin is priced on the market by its gold weight along with a slight premium. That face value is just symbolic and doesn’t really mean anything anymore.
But the key point is, the Treasury has no legal authority to mint a gold or silver coin with a nominal dollar face value of over $50.
However, there is no face value restrictions on the Treasury’s minting of platinum coins, which it has been legally authorized to issue since 1996, and does so each year. The law merely says that the face value of such coins is up to the discretion of the Treasury Secretary. The law wasn’t expected to be used for any serious matter. Currently, the platinum coins that are minted have a face value of $100, which is much less than the value of the platinum metal in the coin.
Image Source: US Mint
Unlike gold coins, silver coins, normal coins, or paper currency, the Treasury Secretary has the legal authority to instruct the US Mint to create a platinum coin of any legal tender face value, including $1 trillion or more if she wants.
When the US Treasury creates most types of non-precious metal coins, it sells them to the Fed at face value. Thus, with a platinum coin, a potential loophole exists if the Fed accepts the deposit rather than challenging it. The Treasury can basically create a meme coin to finance itself entirely with seigniorage.
Basically, there are a couple of avenues for how they make the transaction with this hypothetical $1 trillion coin, but if the process is successful one way or another, the scenario ends with the Fed holding the coin at the face value of $1 trillion as an asset, and the Treasury with $1 trillion in dollars in its general account with the Fed (a liability of the Fed, balanced by their new coin asset), which the Treasury spends from. When the Treasury spends the money, it leaves the Treasury’s account with the Fed but goes into the broad money supply and into bank reserves. Bank reserves are another component of the Fed’s set of liabilities. This is outright money-printing (or technically, “money-minting”).
Even with this loophole, the Treasury would still be constrained from new spending since no new spending was authorized by Congress, but the Treasury could continue to pay its previously-authorized spending obligations and debts. The Treasury could conceivably perform the platinum coin loophole an indefinite number of times as needed.
The loophole clearly works if the Fed doesn’t challenge it as legitimate, but there are some proposals that the Fed could refuse the coin if initiated from the Treasury’s side without legislative action. I’m not well-versed enough in the law to sort through how that would play out in practice; ultimately it would depend on a successful phone call between the Treasury Secretary and the Chair of the Federal Reserve.
A relatively close precedent to this was from 1934, when the Gold Reserve Act required the Fed to give its gold to the Treasury. That was a bigger deal though, and it was authorized by Congress. In order to avoid rendering the Fed insolvent with that action (by giving over a large portion of their asset base in exchange for nothing), the Treasury also gave the Fed an equal amount of gold certificates that equaled the value of the gold they were handing over to the Treasury. The gold certificates theoretically represent a claim to the gold, but in this case are non-redeemable, and thus don’t really mean anything. But from a legal accounting perspective, those gold certificates kept the Fed solvent by avoiding any reduction in official value from the asset side of their balance sheet, and have been held by the Fed for almost 90 years now. In fact, inflation eventually grew the Fed’s balance sheet so large, that they don’t even need those gold certificates to be solvent anymore, since their capital would still be positive if those gold certificates were deleted. But for most of the period since 1934, those basically-meaningless certificates were a key part of what made the Fed’s balance sheet technically solvent.
Congress could override the $1 trillion coin Treasury loophole by creating a new law to restrict the face value of platinum coins to, say, $100 or less, but in order to get past a presidential veto, it would require a Congressional supermajority.
If the $1 trillion coin trick is used to avoid a default, and then the debt ceiling is later raised by Congress, and then net new Treasury debt is issued to collect cash and buy back the coin and retire it, then it could bring the financial system back to normal. Things would continue as they always did, and this event would be a historical footnote.
Alternatively, they could just leave the coin as-is, with that new portion of the money supply in place forever. If there is enough consensus between the President and Congress and the Fed to do it, there’s potentially nothing stopping all future US government spending to occur without issuing net new Treasuries, under the current law. Congress could authorize spending, and the Treasury could mint a couple new trillion-dollar platinum coins each year to pay for any portion not made up by taxes, meaning deficit spending is happening without debt issuance, which is money-printing or money-minting.
Personally, it always seemed easier to me to just invoke the 14th Amendment and sort it out later, if a president is seriously considering the platinum coin loophole. It’s arguably less damaging to the perception of how the financial system works to just ignore the debt ceiling and keep issuing Treasuries, than to do accounting gymnastics to create massive seigniorage with a meme coin. But advisors to a president have various expertise and can advise which method has the highest chance of being met with legal success and avoid a nominal default, or if a brief default is simply easier than these workarounds.
The Takeaway
Ultimately, these sorts of scenarios mainly serve to show how arbitrary the financial system is.
Fiat currency systems have historically had a 100% failure rate over a long enough timeline, and this period from 1971-present is the first time in history when all or most of the world is on a fiat standard. Multiple hyperinflations have occurred within this 1971-present period, but not for the major currencies, because they use a system of checks and balances to keep the growth of the money supply at moderate levels.
Within this fiat currency system, the currency itself has no intrinsic scarcity, but it does have some governance restrictions and division of powers to keep its supply growth in check. As I discussed in my article on money printing, and my article on inflation, new broad money is for the most part created in one of two ways.
The first way is that commercial banks make net new loans, which increases the broad money supply. Banks are constrained in their speed of doing this based on ensuring that not too many of those loans fail and render the bank insolvent.
The second way is that the government can run large fiscal deficits (spending far more than they are taxing), and the central bank can create new bank reserves to buy a big chunk of the sovereign bond issuance associated with that deficit. This technically has no limit, other than that it requires the government and central bank to both participate, and if done at a sufficiently large scale is likely to cause broad price inflation as it did throughout the 1940s and as it has done in 2021. For a decade in the 1940s, central bank independence ceased to exist in order to fund the war effort, and this is arguably happening in the 2020s decade in response to high sovereign debt levels.
Adding to that second way, there are loopholes that reduce the checks and balances by essentially allowing the government to create new money without issuing debt, purely via seigniorage, and thus somewhat violating the concept of central bank independence that way as well.
Since there is no physical constraint on the supply of fiat currency, this sort of fragile monetary system is based on trust. Most of the trust rests on the shoulders of Congress, that they will restrict their deficit spending to a reasonable degree, such that inflation is gradual and pervasive rather than spectacular and acutely disruptive. It is historically around wars and the conclusions of long-term debt cycles, that this trust gets broken.
As cracks appear in the financial system, showing users how arbitrary it is, it could erode confidence in the system and the institutions that support it. In the meantime, any individual loophole, like a temporary $1 trillion platinum coin exploit or an invocation of the 14th Amendment, is unlikely to be immediately inflationary in its own right. It’s more about the build-up of cracks and the increasing perception that the existing monetary system is rather arbitrary, and that the checks and balances are gradually losing power.
Overall, my base case remains that the 2020s decade will continue to be one of real asset outperformance, meaning commodities and other scarce assets are the things to have exposure to. This won’t be linear, and there will be periods of time where having some currency allows you to buy dips and rebalance, but overall, the key is to have strong legal claims on real productive assets at appropriate valuations, and/or exposure to other truly scarce things.