Melissa Davies is chief economist at Rothschild & Co Redburn.
Today the Federal Reserve signalled the return of balance sheet expansion in the form of “reserve management purchases”, designed to keep reserves growing with the pace of the economy and to maintain the “ample reserves” floor system for controlling rates.
As the Fed threw in at the bottom of its statement on the decision to cut interest rates:
The Committee judges that reserve balances have declined to ample levels and will initiate purchases of shorter-term Treasury securities as needed to maintain an ample supply of reserves on an ongoing basis.
The accompanying implementation notice didn’t indicate the size of the coming Treasury bill purchases, but the NY Fed — which actually has to do the central bank’s heavily lifting in markets — said in a separate statement that it would be buying about $40bn a month from December 12 onwards, and taper them at some point next year:
The Desk anticipates that the pace of RMPs will remain elevated for a few months to offset expected large increases in non-reserve liabilities in April. After that, the pace of total purchases will likely be significantly reduced in line with expected seasonal patterns in Federal Reserve liabilities. Purchase amounts will be adjusted as appropriate based on the outlook for reserve supply and market conditions.
Whatever the size, this is potentially a big deal. While the US central bank’s own view is that this is technical issue, entirely divorced from its monetary policy stance, the reality could prove a bit more complicated.
The impact of these RMPs (sorry, everything in finance and economics gets a clunky acronym) on monetary conditions is actually far from clear-cut in today’s complex financial system. That’s especially true given the current floor system for controlling rates, the relatively short duration of overall US debt, and the massive role that money market funds, non-bank financial institutions like hedge funds and the repo market play today.
RMPs vs quantitative easing
Traditionally, the rule of thumb is that QE purchases from the non-bank private sector increase the money supply and has a meaningful economic impact; whereas QE purchases from banks simply increase excess reserves and don’t have any meaningful impact on inflation and private sector asset allocation.
Long duration Treasuries fall into the former category — where the sellers tend to be private sector investors — while short duration bonds fall into the latter category. However, these things unfortunately aren’t quite that straightforward.
Here’s a simple diagram that shows the traditional impact of the Fed buying a T-bill from a bank, creating reserves but not money (ie deposits). To the extent this preserves an ample reserve floor system for setting rates, it can have a meaningful impact in preserving financial stability, even if it doesn’t boost the money supply directly or lead to the famous “portfolio rebalancing” that the Fed has advocated for.
However, what does it look like when the Fed buys a bill from a money-market fund?
Well, this in practice boosts cash in the US central bank’s ON RRP (Overnight Reverse Repurchase) facility, which indirectly relieves pressure on the repo market.
Now let’s see what happens when the money market fund puts this newly created Fed cash to work in the private repo market instead of depositing it in the ON RRP facility.
This cash can then be used by a hedge fund to finance the purchase of a newly issued Treasury, boosting cash held by the government in its Treasury General Account at the Fed. If the government then spends this cash, it would create a new private sector deposit (money), along with new reserves, that were indirectly financed by the Fed’s original T-bill purchase.
So why even shift the Fed’s portfolio towards Treasury bills, if the effect of holding a big portfolio of bonds is potentially roughly the same? Well, a shorter duration bond portfolio has several advantages for the Fed.
Firstly, buying bills means that the Fed can argue it is engaging in a technical operation, rather than a growth or inflation-stimulating intervention in the bond market. Secondly, it gives the Fed the option to “term out” purchases in the future to add stimulus to the economy but without needing to increase the size of the portfolio. Thirdly, short duration purchases reduce the duration mismatch on the Fed’s balance sheet, and therefore the probability of future losses.
It also better reflects the actual Treasury market today. T-bills only comprise about 3.5 per cent of the Fed’s current bond holdings, but the share of T-bills in the US government bond market is currently about 22-23 per cent — and potentially on its way back to the ca 25 per cent-plus share last seen persistently during the late-1990s.
Even bringing the Fed’s share of total assets up to 25 per cen would take a long time, even as longer-duration Treasuries and MBS are allowed to mature and fall out of the portfolio. This obviously creates a strong incentive for the Treasury to skew issuance towards the short end of the curve, which is where the bulk of the demand will now be.
In which case, where will the major sources of demand for longer-dated bonds come from? Fortunately there are a few more pillars of support:
1) There’s been a flurry of bank and stablecoin regulatory measures designed to support demand for Treasuries;
2) Increased MMF liquidity will support hedge fund trades which increase demand for long-term Treasuries.
3) Overseas demand from official sector buyers and private institutions, which just can’t seem to be able to wean themselves off Treasuries.
4) Ongoing Treasury buybacks, which utilise the Treasury General Account cash to buy long-duration debt which is then replaced by more short-term debt issuance.
So what does this mean?
To the extent that Fed buying is skewed towards short duration debt, allowing long duration bond holdings to mature — and to the extent that the Treasury both issues shorter duration debt and buys back longer-dated debt — duration will shift to the private sector in relative terms.
However, it’s questionable whether it’s actually desirable for increased duration risk to reside with the private sector and not the government, especially if this duration risk is increasingly financed by repo funding that is vulnerable to Fed policy errors.
Government liabilities becoming increasingly more ‘money-like’ will probably also put upward pressure on inflation and downward pressure on the dollar, in the absence of healthy interest rate adjustment in the bond market.
The complexity of the current US monetary policy operating system and its inherent risks are not lost on current policymakers, including Treasury secretary Scott Bessent. But to unwind the Fed’s large market footprint would require a twin approach — both boosting domestic non-Fed demand for bonds, and also significantly reducing bond supply by at least stabilising debt-to-GDP, if not running an outright government budget surplus. Clearly, that’s not on the cards.
Fundamentally, the merry-go-round of short-term debt issuance, short-term debt purchases and Treasury buybacks may not have the monetary impact of QE, but will probably show up in higher inflation, a weaker dollar and growing US financial stability risks over time — along with a diminishing ability of the Fed to deliver on its price stability mandate.
