Money market funds, repo, and monetary policy: A mechanism design problem for the Fed

Before there were money market funds — and their finance of banks and big corporations borrowing on commercial paper and other wholesale markets — monetary policy in the US used to be implemented by starving banks of funding and thereby constraining the credit they could provide to the economy (Burns 1979).

How is this possible, you ask, in a world where the money supply is endogenous? This was an environment where current accounts were required by law to pay 0% interest, bank time deposit interest rates were capped at 4-5% and all bank funding in the form of publicly issued debt was legally classified as deposits. When the Federal Funds rises to, say 6%, and other short term rates also rise, (i) the banks don’t particularly want to have to borrow reserves because they are a relatively expensive source of funding, and (ii) holdings of bank deposits become a hot potato, and banks have to adjust to a world where outflows of funds are faster and more variable, so at any given level of lending the risk of having to borrow reserves increases. Historically, banks chose to reduce their lending in this situation, making monetary policy very effective. This is well documented.

This world was completely transformed starting in the late 1960s, when so-called market-based finance first began to develop. Banks were now allowed to use Eurodollar and commercial paper markets for funding. Of course, only a select group of large banks had easy access to this “market-based” funding. These banks facilitated the development of money market funds that could invest in these market-based instruments, thereby freeing the banks from regulatory constraints associated with deposit-based funding. This created a two tiered monetary system in the US, the “money center banks” that were funding on “markets” through money market funds and through their relationships with corporations and their treasury departments, as well as earning income from providing services that made it possible for corporations to fund directly on these markets. In the meanwhile, the rest of the commercial banking system had little or no access to “market-based” funding and was thus still constrained by the regulations governing deposit based funding.

This two-tier system might not have lasted very long, had markets been allowed to function. However, in 1974 when the post-Bretton Woods monetary system had yet to prove itself, the Federal Reserve created “too big to fail” by bailing out a fraudulent bank, Franklin National, in order to stabilize the Eurodollar market — and to preserve the dollar’s position in the post-war monetary system (Sissoko 2019; Spero 1980). In fact, this was just a way of covering up the gross inadequacies of the US bank regulatory system, which continue to this day. (Allowing banks to choose their regulator is not an intelligent way of designing a regulatory system.)

In this era, the Fed’s traditional tools of monetary policy did not have much effect on the money center banks, which after the Franklin National bailout had access to funding at LIBOR, the interest rate on the Eurodollar market, and this funding was understood to carry an implicit US government guarantee. Thus, the reason people like Arthur Burns (1979) doubted that Paul Volcker could tighten monetary policy enough to control inflation was because the Fed had traction over only a portion of the banking system — and not the part of the banking system that provided funding for the biggest US corporations. What Paul Volcker proved was that monetary policy could control inflation even if it was only small- and medium-sized domestic enterprises and the general public that suffered from a significant credit contraction, while the biggest enterprises just faced an incremental increase in the cost of funds.* (Through the worst of Volcker’s interest rate hikes from December 1980 to August 1981, Libor was running 3 to 5% below the Federal Funds Rate.)

This two-tiered banking system continued to operate with the “too big to fail” money center banks receiving ever expanding forms of government support (the bailout of the banks on the backs of LDC countries, the Greenspan Fed’s regulatory lifting of the Glass-Steagall restrictions, the deregulation of derivatives and the hobbling of the CFTC, etc.) so that they grew to make up an ever increasing fraction of the banking system over the course of three decades. While monetary policy was operated as interest rate policy, monetary control was for the most part ceded to the money center banks which were allowed free reign to monetize assets by covering them with “off-balance-sheet” bank guarantees, thereby making them eligible assets for money market funds and so-called market based finance. (Money center banks were able to dominate this activity, because the credit rating agencies explicitly viewed them as having an expectation of government support.)

In 2008, the contradictions at the core of this “market-based” monetary system were exposed. Money market funds are pass-through vehicles. Under no circumstances should they ever be treated as a reliable source of funding for any asset, because it is in their DNA that they are every bit as unstable as money demand itself. The 2007 ABCP crisis took place because this fact was not understood by regulators. On the other hand, the credit rating agencies do apparently understand this instability and as a result, when MMF funding exits the market, the banks have a contractual obligation to support the assets — and as long as the banks in question are “too big to fail” that obligation will fall to the government in extremis. This is the basic structure of “market-based” finance, at least as it applies to the funding of private sector assets on money markets.

Interest rates on money markets are inherently unstable. The most basic task of a central bank is to stabilize the funding of “good” short-term assets, while taking care not to support the value of “bad” short-term assets. The former is important because a lot of very valuable economic activity will be discouraged in an environment where short-term funding rates have a habit of spiking upwards. On the other hand, providing universal access to cheap short-term funding with no credit discrimination at all is a recipe for disaster, because the funds will be misused and bankruptcy and financial instability will result. The short-term funding system has to have some mechanism for distinguishing “good” assets from “bad” assets.

In the pre-2008 monetary policy regime, the problem of distinguishing “good” from “bad” assets was delegated to banks. The Fed ensured that banks could borrow at a stable rate. In theory, a bank that used that facility to lend “badly” was at risk of failure and being closed or sold off. In practice, of course, only small banks were subject to this discipline — and as we saw most “too big to fail” banks engaged in lending practices — including providing contingent guarantees that they were unprepared to meet without regulatory forbearance — that were at best unwise.

Now the Fed is looking for another means of implementing monetary policy. The key problem is, as it has always been, how to stabilize interest rates in a way that is consistent with financial stability goals. Or in other words, to provide stable funding for “good” short-term assets, while avoiding the funding of “bad” short-term assets.

The risk here is that the Fed still seems to be prone to assuming that “markets” will do its job for it. It was caught flat-footed in September, when it learned that “markets” will not stabilize rates by themselves (See BIS 2019 and Coppola 2019 on this).

Now the Fed appears ready to step in to stabilize rates in the repo market. The question is whether the Fed understands that there must be some mechanism for distinguishing “good” from “bad” assets, or whether once again it is expecting “markets” to solve this problem — despite the fact that (i) “too big to fail” is far from having been laid to rest; and (ii) so many businesses have been operating for decades in an environment where it is very cheap to extend and pretend that there has likely been too little feedback and the standard learning process for managing business debt may not be operating effectively. To make the latter point by analogy, just as regular small fires are essential to a healthy forest, so it is important to the economy that regular business failures take place to engender a healthy measure of caution in business decision-makers. As Frances Coppola puts it: “Why are we once again allowing the Fed to provide an implicit backstop for risky non-banks, thus enabling them to misprice risk and gorge on leveraged trades without fear of market penalty? Have we learned nothing from the past?”

In response to Frances, David Andolfatto asks whether a Standing Overnight Repo Facility that serves to cap interest rates in the repo market (as was proposed here and here) is an adequate solution. Unfortunately it seems that an overnight facility is likely to be inadequate to address interest rate volatility on the repo market, since the BIS discussion makes it clear that intraday repo demand was also very high.** Zoltan Poszar’s most recent Global Money Note (#26) explains that this demand for intraday liquidity is likely to due to the fact that Basel III requires the systemically important banks to prefund their intraday liquidity needs. Unsurprisingly this leads a hoarding of reserves in order to preclude the risk of being in violation of Basel III.

Furthermore, simply establishing Fed lending that ensures that money market rates don’t spike seems to be in line with the Fed’s historical tendency to rely on stop-gap measures that have not been thought out at all in terms of their effect on financial stability, but even so become permanent. The Fed needs to think long and hard about what mechanism is going to ensure that the liquidity that it provides is going to the right kind of short-term borrowing. Sixty years ago the answer to the problem was easy: the Fed provided liquidity to the banks and if the banks made bad loans, it was the bank shareholders that were going to eat the loss. Since the rise of market-based lending and bank creditors whom the Fed perceived as needing to be protected at all costs, in the US both the managers and the shareholders of money center banks have been coddled outrageously for decades, and after the experience of 2008 not many people have much faith that they even know how to distinguish good from bad assets any more.

So the real problem is that the Fed has a mechanism design problem that it needs to solve: How is it going to design the market through which monetary policy is implemented to ensure that it is no longer perverted by “too big to fail” and to ensure that any losses on bad assets fall in a way that fully aligns incentives in the market?

* I think it is possible to both approve entirely Martin Wolf’s assessment of Paul Volcker, the man: “Paul Volcker is the greatest man I have known. He is endowed to the highest degree with what the Romans called virtus (virtue): moral courage, integrity, sagacity, prudence and devotion to the service of country.” and yet at the same time to feel that the legacy he left us is very complex indeed. We need great public servants like Paul Volcker, but we need to recognize that they cannot save us when the underlying problem is systemic.

** In fact, the degree to which the Fed is currently providing intraday liquidity (anybody know where this data can be found?) is probably a good clue to whether the stressors in the repo market are mostly overnight or also intraday.