Readers may recall that I wrote a Politico Op Ed at a critical moment in the debt ceiling showdown. That piece, was entitled “Biden Can Steamroll Republicans on the Debt Ceiling”, and I aimed squarely at debunking the idea that the Federal Reserve would step on any “unilateral actions” to avoid treasury default. My key piece of evidence was a memo that I had not read, and was not publicly available. But I knew the contents of the memo indirectly through the Federal Open Market Committee Meeting transcripts. Those comments were in some ways especially revealing, since they came from the Fed’s three leaders: Ben Bernanke, Janet Yellen and Jerome Powell. It’s worth quoting the key part of my Op Ed at length:
On that call, Powell and most of his colleagues reluctantly endorsed buying defaulted Treasury securities — an unprecedented move to maintain financial stability — if a legislative debt ceiling solution did not come in time. Here’s the key exchange between Powell and then-Fed Chair Ben Bernanke (options “8 and 9” in the memo are purchases of defaulted Treasury securities and the Fed “swapping” non-defaulted Treasury securities for defaulted Treasury securities):
MR. POWELL: As long as I’m talking, I find 8 and 9 to be loathsome. I hope that gets into the minutes. [Laughter] But I don’t want to say today what I would and wouldn’t do, if we have to actually deal with a catastrophe on this.
CHAIRMAN BERNANKE: So you are willing to accept “loathsome” under some certain circumstances. [Laughter]
MR. POWELL: Yes, under certain circumstances.
Powell’s willingness to purchase defaulted Treasury securities — however “loathsome” he finds it — casts the entire debate over bypassing Congress on the debt ceiling in a new light. No option under discussion is more extreme, from the Federal Reserve’s point of view, than stepping in and buying compromised securities of uncertain underlying value.
When I wrote and published this Op Ed, I was actually working on getting hold of that memo. Using the Federal Open Market Committee Freedom of Information Act official website portal, I had crafted a request for this memo, as well as some older unreleased materials. As an aside, make sure the website you are on is about the Federal Open Market Committee, if you decide to get into the FOMC Freedom of Information Act Request game yourself. There is a website for the Board of Governors of the Federal Reserve. If you submit a request to that website when you’re seeking Federal Open Market Committee related material, your request will bounce and will be redirected to the FOMC FOIA website. I personally make sure to resubmit when I make this mistake in case something gets lost in the shuffle between FOIA portals. If you want to request materials related to the discount window and 13(3) “unusual and exigent circumstances” authority, the Board of Governors website is the one you want.
Anyway, I had not gotten a response at the time of publishing the OpEd — nor frankly did I expect to get one. I thought this material would be considered too sensitive and too recent to release. Happily, I can inform you that I was wrong. On June 30th of this year I received a copy of the memo. It was also quietly linked to in the “2011 memos” section of their website, meaning it has technically been publicly available since I received it. However, it does not show up on google searches (like many other memos do). As far as I can tell, no one has noticed that this memo is now public. Thus, I can confidently say that I am the first to bring the full text of this memo to the public, and analyze its specific contents. The memo is dryly entitled “Potential Policy Responses to the Debt Ceiling”.
Before getting to the policy options it lays out, it's worth commenting on how it elegantly summarizes “the problem”. The first sentence of the memo claims: “With the federal debt ceiling binding and the Treasury running out of the additional borrowing capacity it can achieve under various accounting procedures, a technical default on Treasury securities cannot be ruled out.”. That summary is notable for a few reasons. First, it recognizes that the problem is not “debt” or “deficits”, but rather the lack of legal authority to issue additional treasury securities. Second, as I’ve hammered many times in this newsletter (as well as the Financial Times and Politico among other outlets) the treasury always already uses accounting gimmicks to continue to fill up its checking account, and make payments. What does that mean? That there is no debate about whether the treasury should use accounting gimmicks to circumvent the debt ceiling. It already does that. The real debate is over whether it should use an “accounting gimmick” to permanently and unilaterally avoid default.
With that out of the way, we move on to their policy options. Let’s start with the five basic options listed first. These are interesting because they were uncontroversial to members of the FOMC when they were discussed in 2011, and again in 2013. Therefore the extent to which these options themselves preserve liquidity for defaulted treasury securities, they emphasize the extent to which the Federal Reserve will strive to contain the economic and financial stability fallout from a treasury default.:
We begin by describing how defaults could affect five routine policy actions that are permissible under the Federal Reserve Act and fall within the current authorization of the Desk and the authority of the Reserve Banks. The relevant issue for these actions is the treatment of defaulted Treasury securities – that is, securities that are experiencing a delay in principal or interest payments due to the debt ceiling. The staff’s recommendation is to make no changes to our current procedures, thereby treating defaulted Treasury securities in these transactions on the same terms that apply to other (non-defaulted) Treasury securities. This approach would send a signal to the market that we continue to view principal and interest for these securities as backed by the full faith and credit of the U.S. government. Moreover, in many cases, this approach would help the market cope with the pressures that could emerge in such circumstances.
In other words, there are all sorts of day to day things that the Federal Reserve does that provide liquidity to treasury markets which the Federal Reserve would continue. But what are those options?
1. Outright purchases of Treasury securities
2. Rollovers of maturing Treasury securities.
3. Securities lending activity.
4. Repurchase agreements.
5. Discount window lending.
These tools are very familiar to people who read about monetary policy. The main surprise here is regarding the first “option”. In inferring the contents of this memo from the FOMC transcript, I didn’t realize that the first option included the purchase of defaulted treasury securities. Yet, it does. As the document says: “Accordingly, unless otherwise directed by the Committee, the [New York Federal Reserve Trading] Desk intends to accept defaulted securities in these operations in the same manner as other Treasury securities, with the prices determined through competitive bidding.” This means that Federal Reserve officials have across the board found the purchase of defaulted treasury securities by The Fed to be uncontroversial. This is an important and remarkable fact.
But what of the other options? Option 2 is not really focused on the defaulted securities part of a potential debt ceiling crisis. Instead, it is concerned with how to deal with the Fed’s “reinvestment policy” when the treasury is no longer conductings new auctions. I will skip over that issue to focus on options 3 to 5. These options sound very different, but really they are all about borrowing from the Federal Reserve. They are really concerned with the question of “collateral” for those loans. In option 3, the Federal Reserve is lending securities rather than “cash”. Nevertheless, it wants collateral. The repurchase agreements of option 4 are simply a form of borrowing that has preferential treatment in bankruptcy. Since the collateralized loan is structured like a repurchase agreement, if the borrower fails to “repurchase” the collateral, then the lender gets to keep it. Option 5 is simply the discount window, the most traditional version of Federal Reserve lending.
So what is the Federal Reserve’s concern? In all these cases, they are thinking through how to treat defaulted treasury securities. An obvious approach would be to not accept defaulted treasury securities as collateral. The justification would be… they’ve defaulted. A subtler option is accepting them as collateral, but at their current (depressed) market price, or at some other lower price. The issue with these approaches is that they worsen liquidity for these securities — while causing panic about what might happen to currently undefaulted treasury securities. The Memo’s alternative is to accept them at the same terms they accept undefaulted treasury securities.
While this is a more technical issue than outright purchases of defaulted treasury securities, it is not less important. As we saw earlier this year with the Bank Term Funding Program, changing the collateral policy to treasury securities can have a profound impact. There is no doubt that if the Federal Reserve, for lending purposes, treated defaulted treasury securities like undefaulted securities it would go quite a long way to preserving liquidity for these instruments, and thus the larger treasury market. It is thus notable that Federal Reserve officials treated these options (which are simply three manifestations of the same policy) as uncontroversial
The next three options are about technical issues related to the repo market, as well as Money Market Mutual Funds. I may write about this section of the memo in a premium piece in the future. But for today I’m keeping my focus on the issue of defaulted treasury securities. It is worth saying though, that this part of the memo illustrates the Federal Reserve’s commitment to preventing a treasury default from having a wider knock-on impact on financial markets. In any case, the remaining part of the memo deals with engaging in purchase and sale operations specifically aimed at absorbing defaulted treasury securities. Now that we have the memo in hand, it's clearer that purchasing defaulted treasury securities was not controversial. Instead, it was specifically aiming at, and designing a comprehensive program to solve, the issue of defaulted treasury securities which made Federal Reserve officials uncomfortable. However, the transcript conversations show far more willingness to use these options if circumstances truly required them than staffers expected. For completeness sake, I will quote this part of the memo below:
9. Purchase operations to remove defaulted Treasury securities from the market. To limit the negative impact of defaulted securities on market functioning, the FOMC could decide to purchase a specified amount of these issues from market participants. These purchases would be in addition to those associated with reinvestments (the first policy issue described above). This approach would help market functioning if cash market liquidity were to deteriorate and participants were otherwise unable to sell their Treasury holdings. Moreover, in contrast to the financing operations discussed above, outright purchases remove the securities from firms’ balance sheets, so that the firms no longer have to deal with the operational issues associated with defaulted securities. Unless offset by other actions, such operations would increase the size of the SOMA portfolio and the amount of reserves in the banking system.
10. Outright CUSIP swaps to remove defaulted Treasury securities from the market. One way to avoid the effects of additional purchase operations on the size of the Federal Reserve’s balance sheet and the amount of reserves is to do them as CUSIP swaps. A CUSIP swap would be an operation in which the Desk simultaneously (or at least on the same day) bought a defaulted Treasury security and sold a non-defaulted Treasury security. This operation would be roughly neutral in terms of the size of the SOMA portfolio and the amount of reserves in the banking system. It could also be designed to limit any change in the duration risk of the SOMA. This approach has similarities to securities lending operations against defaulted Treasury collateral (the third policy issue described above), only in this case the operation would remove the defaulted securities from the market permanently. The terms of these operations would be determined through competitive bidding
Now that we know the Federal Reserve’s playbook in case of a treasury default, we can be much more confident in the future when arguing during the next debt ceiling crisis. Powell, or any future Federal Reserve chair, may deny that this playbook has any implications for how they would respond to unilateral workarounds of the debt ceiling. However, now any reader can plainly see that that isn’t true. They are willing to do an incredible amount — if it's necessary to contain the financial instability resulting from a treasury default. They would gladly facilitate an option the Treasury pursues in its role as fiscal agent, since they can always claim that it was the Treasury’s decision and they had no other option as fiscal agent (which happens to be true). To deny the Treasury and then step into the breach themselves with a policy that is clearly their own to follow, or not follow, puts them in a far more untenable political position. I’m glad to have this out there, and I look forward to unearthing many more gems through Freedom of Information Act requests in the coming months and years.