On (the) QT: What Expanding Central Bank Balance Sheets Mean For The Crypto Industry?
Why the Fed might soon be forced to capitulate
by Ryan Shea
Key Take-aways
- QE is supposed to be only a temporary, or at least, finite policy. Once the economy was back on track, central banks promised to scale back their balance sheets, a process known as quantitative tightening (QT).
- What sounds great in theory has not been matched by reality. Central bank balance sheets in all the major economies are higher now – vastly so in some instances – than they were at the start of the Great Recession some 15 years ago.
- That is not to say that central banks didn’t try. The Fed did in late 2017. However, by 2019 it was forced to reverse its QT programme. Ditto for the BoJ, the ECB and the BoE.
- Despite the recent worries about US regional banks, the Fed has continued to tighten monetary policy via interest rate hikes and additional QT predicated on the belief that it has sufficient policy degrees of freedom to achieve both its financial stability and price stability objectives.
- Based on what happened last time, it is far from clear that this is correct. Not only could the Fed be forced to curtail its QT sooner rather than later but QE may be a financial Hotel California. This has profound implications for everyone and especially the crypto industry.
In the previous research note I made clear that QE is supposed to be only deployed during exceptional circumstances, namely when nominal rates are bounded on the downside. It, therefore, follows that QE should be a temporary, or perhaps better put, finite policy. Once the economy is reflated, the central bank is expected to reverse direction and begin offloading the assets it purchased from the private sector during QE thereby reducing the size of their balance sheet: QE transforms into QT (quantitative tightening).
That is what all central banks promised and it sounds great in theory, the reality though has been somewhat different. For example, Japan introduced QE back in March 2001 and it has been running QE programmes in various guises for much of the intervening 20 years. The direct consequence of these programmes is that the BoJ’s balance sheet has expanded to just shy of 140% of Japanese nominal GDP. Think about that for a moment: the BoJ owns assets whose value is worth considerably more than a year’s worth of output of the entire Japanese economy!
Central Bank Asset/Liabilities (% GDP)
Source: Refinitiv (via Capital Economics)1
Many would argue that Japan’s experience was unique because the bursting of its huge domestic asset bubble burst in 1990 unleashed sustained deflationary forces in the economy, something Richard Koo labelled a balance sheet recession2. However, take a look again at the chart. Around the middle of the last decade, the Fed and the ECB made some progress in reducing the size of their balance sheets, but it was limited. Moreover, even before the Covid pandemic hit, which obliterated any prospects of QT as central banks slashed interest rates back to zero and restarted QE, the Fed, the ECB, the BoJ and the BoE all deemed it necessary at one point or another to increase the size of their balance sheets. The Fed’s experience with QT is actually worth taking a closer look at because it has some significant to more recent developments.
Fed QT 1.0
In 2006 the Financial Services Regulatory Relief Act was passed authorizing the Fed to pay interest on reserve balances3 (IORB) and in 2014 in their Policy Normalization Principles and Plans4 the FOMC announced that would use utilize overnight repos (ON RRP) as a supplementary tool to help keep the Federal Funds rate within the target range5. These tools were supposed to allow the Fed to conduct monetary policy with a much larger-than-normal balance sheet.
In late 2015 when the FOMC decided it was time to finally move away from an ultra-accommodative monetary policy stance, rather than seek to tighten monetary policy via QT they relied primarily on the target Fed Funds rate. The first hike came in December 2015 and it took over a year before the next rate hike occurred. QT came even later, commencing in October 2017.
Part of the reason for this sequencing (rate hikes before QT) is that the Fed considers the target funds rate to be their primary monetary policy tool. It also reflected the fact there was little precedent of central banks reducing their balance sheets in a modern monetary system, something that naturally made them cautious.
The active way to engage in QT is to sell the bonds bought during QE back to the private sector, thereby destroying bank reserves in the process. But, given uncertainties surrounding the effects of QT, the Fed instead chose to deleverage their balance sheet by letting maturing debt on its books roll-off – a passive and predictable approach6. Despite conducting its QT programme in this most benign form, and despite supposedly having the tools necessary to operate with a larger-than-normal balance sheet, they still encountered difficulties.
By late 2019 bank reserves dropped to such a level that short-term money market rates deviated from the Fed’s target range (described as “temporary money market pressures” in the minutes of their unscheduled policy meeting)7, forcing them to reinject liquidity via purchases of short-dated T bills8- see chart.
Fed’s Interest Rate Corridor
Source: Fred database
Cut-and-Paste Policy
Despite this experience, when inflation surged in 2021 they again decided to use the Fed Funds target as their primary tool. The Fed hiked in March 2022 and three months later they began reducing their securities holdings “in a predictable manner primarily by adjusting the amounts reinvested of principal payments received from securities held in the System Open Market Account (SOMA)”, namely QT via benign roll-offs. In effect, they followed the exact same play book as during the previous tightening cycle. The only difference being it was more aggressive. Faced with inflation rising to its highest level in over 40 years, the Fed has implemented its fastest rate hiking cycle over a similar time frame.
As the recent turmoil in the US banking sector demonstrates, the aggressive pace of rate hikes has had unfortunate consequences for some banks. When the Covid stimulus cheques hit bank accounts during the pandemic, the lack of oven-ready lending opportunities meant quite a few commercial banks (including the now defunct SVB) decided to invest in Treasuries. Whether because they were lulled into a false sense of security by a decade of low nominal interest rates, hubris, greed or plain incompetence (Treasuries may not have credit risk but they sure as hell have duration risk) the Fed-induced back-up in Treasury yields resulted in unrealized losses of more than $600bn on US banks’ portfolios of investment securities – see chart.
Unrealized Gains(Losses) On Investment Securities
Source: FDIC
For the US banking sector as a whole, which has assets of almost $23tr, losses of this magnitude should be manageable. However, as the failure of SVB – the second largest bank failure in US history - showed these mark-to-market losses are not evenly distributed throughout the US banking system. After 2018 regulations were relaxed (via the Economic Growth, Regulatory Relief, and Consumer Protection Act ), banks with assets below $250bn did not have to submit to annual stresses tests9and faced more relaxed liquidity rules10. It was in these smaller regional banks – not the big banks – where the problems resided. Having seen SVB fail, depositors in these smaller banks took flight and according to media reports JP Morgan and Wells Fargo can’t open bank accounts fast enough11.
The C Word
This is a powerful reminder that the most important “asset” any bank owns is confidence (what C word did you think I was talking about?). Write that down and you are in a world of pain as Credit Suisse also recently found out when its regulators forced it to be sold to UBS for just 4% of book value12. Despite happening on the other side of the pond, it did little to improve sentiment towards the US banks.
To thwart the risk of contagion, the Fed scrambled to announce the BTFP – a new facility (backstopped by $25bn from the Treasury’s Exchange Stabilization Fund) that allows banks to park assets on the Fed’s balance sheet for up to one year at par – a very generous dispensation given most central banks apply a valuation haircut on pledged assets to provide some additional protection.
So far, and bearing in mind that the BTFP has been in existence only a couple of weeks, the uptake has been fairly modest at some $53.7bn. However, borrowing at the Fed’s discount window (so-named because they typically apply valuation haircuts although this has been waved at present) initially surged to $152bn – a number that exceeds the Great Recession peak (see chart) before dropping to $110bn last week.
Source: Federal Reserve (via Bloomberg)
Like QE, discount window + BTFP borrowing directly inflates the Fed’s balance sheet, and the most recent increase of these two facilities has effectively reversed half of the Fed’s latest QT balance sheet reduction (and in record time).
Printer Go Burr Redux?
At this point I expect to hear howls of protests from monetary nerds – including but not restricted to central bankers – that this is not the same thing. In fact, commenting ahead of the March FOMC policy meeting, former Federal Reserve Bank of New York William Dudley proclaimed that he expected QT to continue because that program is “very separate and different” from the measures the Fed is taking to shore up confidence in the commercial banking system13. While his QT prediction came true (unlike his forecast for an unchanged Fed funds rate) I, and I suspect others, doubt the Fed’s QT programme will be in operation for much longer.
Why?
Regional banks loaded up with (now underwater) Treasuries can pledge them at par with the Fed for the next 12 months. This stops the crystallization of losses on these securities due to customer withdrawals, which may now not even occur. But, this facility is only available for 12 months. By March next year the banks targeted by this programme will be back to square one unless they can either earn sufficient income to offset these losses and/or Treasury bonds rally sharply reducing their losses (hedging duration risk at this point makes absolutely no sense because it simply locks-in the negative valuation).
The former option is problematic because commercial banks face stiff competition from money market funds (MMFs). Recall, in 2014 the Fed announced that it would utilize the ON RRP as an additional tool to help them target the Funds rate. At inception it was envisaged that this would be used “only to the extent necessary and will phase it out when it is no longer needed to help control the federal funds rate”14.
Unlike reserves which are only available to eligible institutions with master accounts at the Fed15, ON RRP is accessible to a wide range of counterparties, including primary dealers, banks, money market funds and government sponsored enterprises. As can be seen in the chart below, contrary to the FOMC’s original expectations – they expected usage would decline as they tightened monetary policy and lifted Treasury yields16- ON RRP has proved to be a very popular product (if product is the right word for a Fed facility). It accounts for around $2.4tr of the Fed’s balance sheet, of which more than half comes from MMFs. This is not much lower than the $3tr in bank reserves.
Fed Balance Sheet – Selected Items
Source: Fred database
Historically, MMFs invest in T bills, but their supply has dwindled over recent years (due to TGA drawdown17). To fill the void they started to tap the Fed’s ON RRP facility. Unlike bank deposit rates which tend to lag the Fed funds target rate, MMF rates track the Fed’s policy rate much more closely. This boosts their relative attractiveness during rate hiking cycles (currently MMF deposits earn their holders 4.55%, compared with commercial banks whose deposit rates are around 0.5%.)
Not only do MMFs offer higher returns, but for account holders wishing to deposit above the FDIC limit exposing them to the credit risk of the bank (in theory at least), T bills and O/N RRP are safer. As any investor will tell you higher returns for lower risk is a slam-dunk. When one also adds the recent bank turmoil to the mix, it is hardly surprising that data from the Investment Company Institute, showed extremely high inflows totalling $239bn to MMFs in the two week period since SVB and Signature Bank were deemed to be no longer going concerns18.
Commercial banks can, of course, offer higher deposit rates to try to compete with MMFs but this will only erode the net interest margins they require to generate the income necessary to offset losses from their Treasury holdings in 12 months time. Whoops.
There is a further complication. As the New York Fed makes clear in its FAQ19when ON RRP transactions are settled, the movement of funds from the clearing bank to the Fed corresponds to a reduction in bank reserves by an equal amount. While the Fed is able to influence bank reserves via QT, the decision to allocate money between the commercial banks and MMFs is determined by the general public. If the Fed continues with QT and there are continued flows out of commercial bank deposits into MMFs – for reasons that do not necessarily have to do with fears over the solvency of regional banks – there is a clear risk bank reserves fall to such an extent that another 2019 “incident” occurs when the Fed was forced to halt its QT program to regain control over its interest rate target20. Double whoops.
If the amount of monetary tightening already implemented results in a sharp drop in inflationary pressures (monetary policy impacts with a lag – something else many people seem to have forgotten – so this scenario cannot be excluded) the inability of central banks to bring down their balance sheets (QT) may not be an immediate issue. Rather they will probably be only too thankful that they have room to respond via orthodox monetary policy tools i.e, cutting interest rates. But that really doesn’t change much in the bigger picture because it would still mean QT is off the agenda.
Hotel California21
What the Fed’s experiences, not to mention that of other central banks, highlights is that unwinding previous QE programmes sounds easy “Oh, we can just sell some bonds back to the market22 and/or let them passively roll off” but in reality it is much more complicated. It might not even be achievable. In a 2021 report by the Economic Affairs Committee in the UK House of Lords, which has former BoE Governor Merv King as a member – someone who clearly knows a thing or two about central bank policy and QE – contained the following paragraph:
“[T]here is an increasing risk that central banks are facing a “no-exit paradigm” from quantitative easing. No central bank has managed successfully to reverse its asset purchases over the medium to long- term, and the key issue facing central banks as they look to halt or reverse quantitative easing is whether it will trigger panic in financial markets that spills over into the real economy. [Ed note: my emphasis]”23
That’s worth repeating. No central bank has managed successfully to reverse its asset purchases over the medium to long-term. I will explain why this matters to everyone, but especially to the crypto industry, in the next research note.
Until next time.
Ryan Shea, crypto economist
Footnotes
1See: https://www.capitaleconomics.com/blog/we-need-to-talk-about-qt
2See: https://www.amazon.co.uk/Balance-Sheet-Recession-Uncharted-Implications/dp/0470821167
3See: https://www.federalreserve.gov/monetarypolicy/reserve-balances.htm
4See: https://www.federalreserve.gov/monetarypolicy/files/fomc_policynormalization.pdf
5The IORB usually serves as an upper limit and the ON RRP the lower limit.
6The goal was to make it as boring as watching paint dry to paraphrase William Nelson, former deputy director of the Division of Monetary Affairs at the Fed Board.
7See: https://www.federalreserve.gov/monetarypolicy/fomcminutes20191030.htm
8See: https://www.newyorkfed.org/research/epr/2021/epr_2021_market-events_afonso.html
9For a more detailed breakdown – see: https://rooseveltinstitute.org/2023/03/15/how-2018-regulatory-rollbacks-set-the-stage-for-the-silicon-valley-bank-collapse-and-how-to-change-course/
10Even though the Fed as SVB’s regulator had discretion to mandate both, they failed to do and following SVB’s failure have launched an investigation into their oversight - see: https://www.federalreserve.gov/newsevents/pressreleases/bcreg20230313a.htm
11See: https://markets.businessinsider.com/news/stocks/svb-collapse-silicon-bank-valley-jpmorgan-wall-street-depositors-2023-3
12See: https://www.cnbc.com/2023/03/20/what-ubs-rescue-of-credit-suisse-cs-means-for-markets-and-banks.html
13See: https://www.bloomberg.com/news/articles/2023-03-20/dudley-sees-fed-sticking-with-qt-even-in-face-of-banking-strains
14See: https://www.federalreserve.gov/monetarypolicy/files/fomc_policynormalization.pdf
15Astonishingly, the Fed does not publish a list of institutions of institutions that have access to master accounts as it considers that confidential business information, although last November it did invite public comment on a proposal to publish a period list of master account holders – see: https://www.federalreserve.gov/newsevents/pressreleases/other20221104a.htm
16See: https://www.ft.com/content/d2444e5c-55c2-41fd-87d5-b90c68b25da6
17This is the Treasury General Account, which is the government’s operating or checking account at the Fed that allows the government to make public money transactions.
18See: https://www.ici.org/research/stats/mmf
19See: https://www.newyorkfed.org/markets/rrp_faq
20To mitigate this effect, in July 2021 the FOMC announced a new policy tool – the standing repo facility.
21A reference to the song by the Eagles that contains the following lyric “You can check-out any time you like, but you can never leave!”
22To-date the only central bank that has adopted a more proactive approach to QT by selling its bonds back to the market is the BoE. This programme, which only began in November last year, was echoing the language of the Fed to be carried out “in a gradual and predictable manner”. I might add the programme is small with only £45bn out of a total £895bn bought via QE pencilled in for 2023. Having failed to get enough demand for its sale in early February it will be interesting to see how they fare, given subsequent market developments.
23See: https://committees.parliament.uk/publications/6725/documents/71894/default/
