Decoding Warsh’s Fed balance sheet plans is far from simple

As Jerome Powell enters his final months at the helm of the Federal Reserve (Fed), market participants have been busy analysing the past speeches of his nominated replacement, Kevin Warsh, with the aim of forming a view as to how his hand is likely to be played. To put it bluntly, we think the usefulness of this process is limited.

For one, economists update their opinions as new data is generated and new papers analysing said evidence are published. It is perfectly reasonable for a serious economist to change camps with respect to a specific development if new evidence suggests the effects of a given policy are broader than previously thought, for example. In addition, it is one thing to criticise policies or call for complete revamps of processes or strategies while you are not the Fed chair. Once in the job, things might be a bit different. There is a political and continuity aspect that is also very important. In the end, a central bank’s biggest asset doesn’t sit on its balance sheet, it is the more intangible asset of credibility.

How could Warsh reduce the balance sheet?

While on the topic of the Fed’s balance sheet, investors seem to have some level of conviction that Warsh might attempt to reduce its size, as he has been an advocate of a smaller balance sheet in the past. Given this could have important consequences for markets, and in particular for monetary policy transmission, we think it is important to understand how this could be delivered in practice. While we do not know at this stage if he will pursue this objective, a detailed analysis highlights the difficulties of doing this at this juncture when the Fed’s balance sheet has already declined substantially.

First, to state the obvious, the current size of the Fed’s balance sheet and the fact that it has grown to the current size of around $6.6tr are not coincidental. In the aftermath of the global financial crisis, as the Fed was injecting liquidity into the system through quantitative easing (QE), we also saw a transition in the way monetary policy transmission works. Sparing the finer details, the new regime (an “abundant reserve” system) requires the Fed to have ample reserves and the transmission of monetary policy operates through the deposit rate offered on these reserves. A radical change to the size of the balance sheet could bring about a change in the way monetary policy transmission works, and potentially be disruptive to the current status quo.

The mechanics of the Fed’s balance sheet

Before we dive further into the recent events, it is worth taking a step back and considering some of the mechanics at play here.

The Fed’s balance sheet has declined markedly since its 2022 peak of around $9tr, but it remains well above pre-Covid levels. Exhibit 1 shows the Fed’s assets, mostly comprised of $6.3tr securities held outright including US Treasuries (USTs) and mortgage-backed securities (MBS). The central bank’s liabilities include currency in circulation, reverse repo balances, the US Treasury general account or TGA, and banking sector reserve balances with the Fed.

QE was restarted in early 2020. From a mechanics point of view, the Fed bought USTs and MBS from the public through the banking sector. Banks would buy a UST from the public, paying the previous holder with a deposit, and then sell the UST to the Fed, for which they would get paid in reserves held at the Fed. In parallel, the US Treasury increased UST issuance enormously to fund unprecedented fiscal stimulus packages. Some of those measures took time to execute, which is why the TGA increased substantially. The TGA is the US Treasury’s current account with the Fed. Just like any person’s current account with their bank is a liability of the bank, the TGA is a liability of the Fed. 

As these monies started getting spent, at the same time that QE was ongoing, the public was left with enormous amounts of liquidity. From the public’s point of view, the Fed was buying their USTs while the government was handing them checks in the post. That cash needs to be invested at some stage in short term deposits or T-bills, which drives short term rates down. This downward pressure in yields was so intense that the Fed relaxed the rules of its reverse repo facility, increasing the amounts available and relaxing the rules for money market funds to participate. The reverse repo facility allowed money market funds to lend money to the Fed at the reverse repo rate, which eased the pressure in short term rates elsewhere, T-bills being the most obvious example, and shows up in Exhibit 1 as an increase in the blue area (Reverse Repurchase Agreements – Domestic). 

When quantitative tightening (QT) started, the Fed passively let part of its UST and MBS holdings mature. The Treasury, therefore, paid the Fed using the TGA balances and then sought to replenish the TGA accordingly by issuing T-bills and USTs. Consequently, as the supply of these increased, money market funds found more alternatives to invest their monies and the use of the reverse repo facility began declining and is today fairly small. In Q4 last year, the Fed stopped its QT programme as the first signs of stress in short term repo markets indicated that liquidity in the system was reaching some sort of equilibrium, i.e. liquidity was no longer abundant. The Fed’s balance sheet has been almost flat since then.

More QT would be easier said than done

After this rather long explanation we can re-ask the question, how exactly could Kevin Warsh look to reduce the size of the Fed’s balance sheet? From the asset side point of view, it’s pretty clear that what would need to come down is the Fed’s UST and MBS holdings. The “Other” category is a collection of small and largely irrelevant assets. A decline in assets must be matched with a decline in liabilities. As the reverse repo facility is now close to zero, a continuation of some form of QT would almost certainly imply a reduction in the reserves that banks hold with the Fed (the dark blue area in liabilities in Exhibit 1). Banks’ demand for reserves is driven by liquidity preferences but more importantly by liquidity requirements. There are liquidity coverage ratios banks need to comply with, which pushes demand for reserves, considering they are the most liquid and safest asset out there. If there was more QT with no changes to these requirements, it’s likely that demand would exceed available supply of short liquid assets, causing spikes in short term rates due to scarcity. Reducing bank liquidity requirements is not without risk though, as less liquid banks’ balance sheets might leave them more exposed to liquidity shortages during any runs, at least in theory. In an analogous manner, allowing slightly less liquid assets to comply with these requirements has similar negative consequences.

Given currency in circulation is largely a function of nominal GDP growth, the only liability left that could in theory be reduced is the TGA, which currently stands at just over $900bn. The TGA is subject to a number of factors that are outside of the Fed’s control. These include government shutdowns, debt ceiling squabbles, and seasonality such as the end of the tax year. While it could be informally agreed that the Treasury targets a certain number for the TGA that is lower than $900bn, large moves in the TGA might impact the availability of liquidity in the system. While these temporary fluctuations in the supply-demand balance for liquidity could in theory be addressed via open market operations, i.e. injecting liquidity when needed, if there is a constant need for liquidity then it might not be the wisest thing to do to reduce the Fed’s balance sheet.