Regulatory Discussion in the 2019 FOMC Transcripts
Featuring a Brainard v. Quarles debate on alleged supervisory preferences for reserves
Today the Federal Reserve released transcripts and other historical materials from the 2019 meetings of the Federal Open Market Committee. The meetings, of course, focused on monetary policy, but also included a few discussions of supervisory and regulatory policy, particularly as it - maybe - began to affect monetary policy.
Is There a Supervisory Preference for Reserves?
Liquidity rules adopted after the 2008 financial crisis generally treat U.S. Treasuries and reserves as equivalently liquid, and as equally valid ways of meeting requirements to hold prescribed amounts of high-quality liquid assets. Nonetheless, some banks have long felt that examiners in fact have a clear preference that banks hold reserves. Some worry that this examiner preference has adversely affected the Federal Reserve’s ability to conduct monetary policy.
The FOMC meeting transcripts from early in 2019 reflect a few comments from Reserve Bank presidents suggesting there might be something to this concern,1 but because Reserve Bank presidents don’t set or vote on supervisory or regulatory policy this on its own is slightly academic.
Where the transcripts get really interesting is later in the year. September 2019 saw significant turmoil in the repo market, so much so that the FOMC held an unscheduled meeting by conference call in early October to discuss.
A few days after that, and in preparation for the regularly scheduled FOMC meeting to be held later in October, Federal Reserve staff prepared a memo on Recent Money Market Developments. The memo is about a lot more than supervisory and regulatory issues, but as relevant to this post the memo shares these preliminary conclusions and other observations:
Declining reserve balances could have dampened the willingness of domestic banks to step in and lend funds that would have helped to alleviate pricing pressures. Firms’ risk management practices, which have been shaped directly and indirectly by bank regulations and supervision, appear to have played a role.
The substitution of reserves for reverse repo collateral would constitute an exchange of one high-quality liquid asset for another, with no change to a bank’s overall asset size or liability composition. Given that, one would expect banks to opportunistically convert excess reserves into reverse repos backed by U.S. Treasuries to earn a higher return when overnight repo rates rise sharply. Banks’ inertia may be indirectly related to the influence of supervision and regulations on banks’ internal risk management practices, which may have made some banks reluctant to step in.
[T]he largest banks also run internal liquidity stress tests, which are examined by supervisors. Supervisors focus on liquidity risk management, including managing buffers to be able to meet outflows as they occur in stress. Some market participants may have interpreted that focus as a supervisory preference for reserves. The heterogeneous approaches to buffer composition in evidence among the large banks, however, is consistent with supervisors taking a firm-specific approach to liquidity risk management practices.
The memo, particularly in that last quote, appears to reflect a bit of a debate between its authors as to the extent that regulation or supervisory preferences played a role. That same debate is reflected in the October 29-30 meeting transcript in the comments of then-Governor Brainard and then-Vice Chair for Supervision Quarles.
Governor Brainard’s Comments
Governor Brainard’s comments on the September repo market events (p. 112) acknowledged “frictions in the pipes,” saying it was “important to understand whether these are associated with changes in business models, risk management practices, or regulatory requirements in order to decide what type of policy response might be appropriate.”
While later agreeing that it would be worthwhile for the Board to “assess whether there are, in fact, some areas of guidance in which we can smooth those frictions,” Governor Brainard suggested that, of the reasons she had mentioned, business models and risk management decisions were perhaps more important.
[F]irms’ risk management appears to be determining the degree of monetization risk that each firm chooses to take on. Risk-management practices appear to reflect a low perceived cost of holding a reserve buffer as self-insurance to meet volatile payments relative to being highly reluctant to use any Federal Reserve credit. […]
Although some large firms occasionally use intraday credit, this usage is limited, compared with pre-crisis, in part because of fears that daylight credit could turn into an overnight discount window loan, which is discouraged by senior management and the institutions’ boards.
Moreover, the internal risk limits and governance processes appear to have some inertia. So the dealers’ incentives to step in when spreads widen because of perceived idiosyncratic short-lived factors are somewhat muffled by risk limits that appear to be changed only at longer horizons in response to more sustained changes in the opportunity set.
As evidence for firm-specific decisions, rather than supervisory expectations, being a key driver, Governor Brainard pointed to differences in both the composition of large banks liquid asset portfolios and in their specific response to the September 2019 repo market events.
The noticeable differences among the large dealer banks in their practices regarding reserve holdings and their responses to the widening of repo market spreads suggest that the preference for reserves in HQLA holdings is not a monotonic function of the requirements such as the LCR, the G-SIB surcharge, or the leverage ratio. Staff discussions with several of the key dealer banks reinforced this finding.
Along similar lines, from a few months earlier, see Figure 3 on PDF page 16 of the June 18-19 presentation materials and the comments from Federal Reserve staff on page 6 of the June 18-19 meeting transcript saying that Figure 3 demonstrates the “heterogeneous approaches” taken by large banks to the composition of their HQLA portfolios.
Vice Chair for Supervision Quarles’s Comments
A short while later in the meeting, Vice Chair for Supervision Quarles offered his own view (p. 120). He started by noting that “many,” though not all, banks had told the Federal Reserve that “a significant element of the market frictions that we saw in September was the instruction that they have received from Fed supervisors to prefer reserves over Treasury securities in the composition of their high-quality liquid asset.”
On the other hand, according to Vice Chair Supervision Quarles, Federal Reserve supervisors “insist … that they do not have such a preference [and] have never told the banks to have such a preference.” As support for this, the supervisors “point to the facts that Governor Brainard has cited—that some banks, in fact, do not hold a particularly large percentage of their HQLA in reserves.”
Calling both sides of the debate “quite sincere and quite vehement,” Vice Chair for Supervision Quarles continued:
I think the way one squares the circle is by looking at the operation of the internal stress liquidity test that we require the banks to run. So we have an overall liquidity regulation that requires the banks to hold a minimum amount of liquid assets. But, quite sensibly, we also require the banks to run an analysis of how those liquid assets—which could be Treasury securities, reserves, or, after the passage of a law in the summer, municipal bonds—will perform under stress.
Now, in the world as it currently stands, there is a modest but measurable difference in the liquidity and usability of reserves versus Treasury securities in extreme stress. Treasury securities settle a day later. There may be problems of severe disruptions to markets and to the economy. And so, required to perform this stress analysis, many banks conclude that they should keep very high levels of reserves. The supervisors say, “We didn’t tell them that. They came to their own conclusion.” The banks say, “We came to that conclusion because of the way you require that we run our stress-testing practices. It’s a direct result of the supervisory framework.”
Vice Chair for Supervision Quarles then went on to discuss how he believed a standing repo facility — if “appropriately designed” — could help to address the issue.
Other Notes
The facts and data referenced in the following comments are all pretty stale by now, but just to highlight three other somewhat regulatory-related things from the transcripts:
Governor Bowman on Private Credit and Agriculture Markets
Again, the specific numbers here are years old by now, but this from the July 2019 meeting transcript (p. 115) was interesting from Governor Bowman, particularly given her rumored consideration in 2025 for a new role.
Private credit funds are also relatively new entrants to the lending market, so their performance is much less understood. During the previous recession, the private credit market was about one-third the size of today’s market, with slightly higher credit losses than leveraged lending. It follows, then, that this credit also tends to be extended to riskier borrowers—that is, firms with higher debt relative to assets or earnings ratios.
From a financial stability perspective, private credit funds may pose a lesser risk than banks because they tend to be less interconnected with other financial institutions. That said, private credit funds do compete with large regional banks, and yet they’re not subject to the same degree of regulatory and compliance scrutiny and requirements. […]
My interest today is more specific and focused on lending to agricultural producers, an area where generations of relationship lending has traditionally been important. Through many cycles—some lengthy and very stressful—banks have relied on a deep understanding of agricultural production, particularly when they have geographic proximity and local knowledge. Currently, total farm debt in the United States is about $400 billion, with about two-thirds backed by agricultural real estate. And 45 percent of agricultural lending is made by the Farm Credit System, which is a GSE, and 40 percent by commercial banks. Nonbanks make up the balance of 15 percent, a portion of which is private credit.
In comparison to banks, private equity funds tend to be more focused on loans to riskier businesses for a higher yield. This additional credit source could be contributing to the elevated land prices we’ve seen in the agricultural markets for the past 10 years. Should agricultural land prices decline, this could lead to insufficient debt collateralization for both banks and nonbanks, potentially creating significant issues for financial institutions.
In my view, an important risk associated with the entrance of these private credit firms is the durability of their presence. They’ve not yet been tested by a major and prolonged agricultural downturn. Experience suggests that these downturns can last multiple years before a recovery, requiring patience on the part of lenders. Should private credit funds prove prone to liquidate and move out when, inevitably, times get tough, the resulting additional contraction in credit supply could significantly deepen an agricultural downturn.
The good news is that private equity funds are closed end and have stable long-term funding from investors, so run risk appears limited. These funds also do not appear to rely much on borrowing to fund their lending. About half of the funds borrow from large banks. It also bears repeating that, so far, this is a small fraction of the wider agricultural market.
President Rosengren on A Large German Bank
In the January 2019 meeting transcript (p.105), President Rosengren commented on “financial stability risks generated by undercapitalized and underperforming European banks that could seriously affect the euro area but would also pose a serious complication for the United States.”
President Rosengren referenced one bank in particular:
Most notable is one of the largest globally important German banks. Its ratios look great. Here, tier 1 common equity is currently 14 percent, and its LCR is 148 percent. Over the past five years, its stock price has fallen 75 percent, and its CDS spread has more than doubled. Press reports indicate that capital and liquidity ratios that are well above minimums are providing little comfort. Financial press reports are highlighting merger proposals, and counterparties have been reported to be reducing exposures. That this bank is under such pressure, once again, even though its capital and liquidity ratios are high, even by our standards, should warn us against complacency following the significant post-crisis progress we’ve made in strengthening our own banks’ capital and liquidity requirements.
He then tied this back to his views on U.S. bank regulation:
We should question whether our supposed gold plating of financial ratios on systemically important financial institutions is really sufficient to ensure financial stability in times of economic stress. Furthermore, gold plating should be thought of not only in terms of calibrating minimum requirements on specific ratios, but also in terms of ensuring that the system of regulatory requirements and supervisory tools as a whole is truly sufficient to protect the financial system and the broader economy against the range of problems that institutions may face in a crisis.
Governor Brainard on Libra
Finally, in the October 2019 meeting transcript (p.153), Governor Brainard shared comments on Facebook’s Libra project, which was formally announced a few months previously. These comments aren’t any different from what Governor Brainard said publicly about the project, but it is still kind of interesting the topic made it to the FOMC.
I will now turn briefly to the nontraditional risks out over the horizon. One risk that bears watching regards the emergence of stablecoins at global scale. Stablecoins aspire to achieve the functions of traditional money without relying on confidence in an issuer such as a central bank to stand behind the money. Indeed, for some potential stablecoins, users may have no rights with respect to the underlying assets overall.
We’ve already seen the growth of massive payments networks on existing digital platforms, such as Alibaba and WeChat, and the issuance of stablecoins on a smaller scale, such as Tether, Gemini, and Paxos. What sets Facebook’s Libra project apart is the combination of an active user network representing more than one-third of the global population with the issuance of private digital currency opaquely tied to a basket of sovereign currencies. […]
For example:
President Bostic in the January 2019 transcript (p.68): “I believe there would be value in exploring whether there are regulatory reforms that could reduce the volatility of reserve demand without adversely affecting bank liquidity and financial stability”
President Kaplan in the January 2019 transcript (p.71): “I would like us to take advantage of the time to—again, as we’ve talked before—tweak banking regulations, if possible, and look at other adjustments that will allow the banks to more easily hold Treasury securities in lieu of reserves, or encourage a more efficient distribution of reserves across banks, or both.”
President Kashkari in the March 2019 transcript (p.43): “And then, finally, one factor driving high demand is likely our own supervisory and regulatory guidance, and I think that that’s something that we should take a look at. Maybe this isn’t the right forum for it, but if we really are telling banks, “You should hold reserves and not T-bills,” I think we need to ask ourselves, why are we doing that? Does it really make sense?”