What will happen when the Reverse Repo Facility gets depleted?
What will happen when the Reverse Repo Facility gets depleted?
Not much attention is being paid to what’s happening with the Fed’s reverse repo facility, which has been the main source of liquidity for the financial system and source of demand for the flood of Tbill issuance from the Treasury this year. Its rapid depletion will have consequences for the functioning of Treasury markets and the Fed’s balance sheet in 2024.
First, some context on the purpose of the reverse repo facility (RRP) is and its place in the plumbing of the financial markets. It will take too many words to explain the mechanics of how the RRP works, but you can find some info here. The RRP was created after 2008 as a way for the Fed to control short-term interest rates. It’s a facility where institutions can lend to each other at an overnight interest rate set by the Fed, in line with where the Fed has set the Fed funds rate. The existence of the facility prevents short-term interest rates from deviating too far from Fed funds, because any deviation can be easily arbitraged in size to bring the market back to where the Fed has set the repo rate.
The borrowers from the RRP are primary dealers, who use the RRP to finance their purchase of bonds at Treasury auctions, and hedge funds, who use the RRP to finance their massively leveraged holdings of cash Treasuries while shorting futures against them, making money when the prices of cash and futures converge (aka the Treasury basis trade). The lenders on the other side are money market funds (MMFs) and other yield-seeking counterparties that aim to earn a yield equal to the Fed’s overnight reverse repo rate. This is important to know because without the large balance in the RRP, the available liquidity for primary dealers and hedge funds to purchase and hold Treasuries would be severely diminished.
As you can see from the chart below, the size of the RRP exploded from almost zero in Q1 2021 to $2.5T by Q1 2023. The cause for this explosion was two-fold. First, there was more demand for Tbills than available supply, resulting in Tbill yields being less competitive to the yields institutions and MMFs could get in the RRP.
Second, the total assets in money market funds increased from $4.76T in Q4 2020 to $5.69T in Q1 2023, increasing the total amount of capital that was lent into the RRP.
What happened in Q3 of 2023, when the size of the RRP took a sudden nosedive? That’s when the debt ceiling was resolved, and the Treasury started to flood the market with Tbills to refill the Treasury General Account. Supply of Tbills overwhelmed demand and made Tbill yields more competitive relative to RRP yields, thus incentivizing capital to leave the RRP to invest in the Tbill market.
The Treasury issued roughly $829B in Tbills in Q3, while the balance in RRP declined by $544B, suggesting that two-thirds of the demand for Tbills came out of the RRP. The estimated net issuance for Q4 2023 is $513B in Tbills and $338B in bonds and notes.
Based on the above estimates, we may see the RRP draw down further from today’s balance of $1.265T to around $900B by the end of Q4. The RRP might be completely depleted by Q3 2024, thus removing a source of demand for Tbills and a source of liquidity for primary dealers and hedge funds to absorb the bonds that the Treasury is selling.
The repo crisis of 2019 offers a hint of what happens next. From the Fed’s website:
In mid-September 2019, overnight money market rates spiked and exhibited significant volatility, amid a large drop in reserves due to the corporate tax date and increases in net Treasury issuance. Although some upward pressure on money market rates due to these seasonal factors was expected, the extent of the increase in both the level and volatility of rates in secured and unsecured markets was surprising…
The moves in both secured and unsecured rates on September 16 and 17 were much larger than any of those observed over the past few years. Figure 1 shows the effective federal funds rate (EFFR) and the secured overnight financing rate (SOFR), a broad measure of the cost of borrowing cash overnight collateralized by Treasury securities, since December 2015.
Why did this happen?
Two widely cited factors exerted upward pressure on overnight funding rates in mid-September. First, quarterly corporate tax payments that were due on September 16 were withdrawn from bank and money market mutual fund (MMF) accounts and went to the Treasury's account at the Federal Reserve (Fed). Second, $54 billion of long-term Treasury debt settled on September 16, which increased the Treasury holdings of primary dealers that purchase these securities at auctions and finance them through the repo market. As tax payments and the settlement of Treasury auctions drained a large amount of cash, reserves in the banking system declined by about $120 billion over two business days. In the repo market, there were more Treasury securities to be financed in the market that day with relatively less cash. The increase in the repo rates on September 16 seemed to stem from a demand-supply mismatch in the market.Going into mid-September, although some upward pressure on money market rates was expected in response to tax payments and Treasury auction settlement, the realized rate movements were quite large by historical standards, as shown in Figure 1.
How did the Fed respond? It reversed its reduction in its balance sheet by injecting funds into the RRP and buying $60B in Tbills per month. They insisted it was not QE, but because the Fed’s balance sheet reversed from contraction to expansion, it effectively had the same impact on asset prices as QE did.
The RRP is essentially the pool of liquidity keeping the Treasury market afloat, allowing primary dealers to do their job of supporting Treasury auctions and allowing hedge funds to provide buy-side liquidity on a leveraged basis. A depleted RRP would have disastrous consequences for the Treasury markets and risk assets in general.
I see two potential scenarios for how this will play out in 2024. The first scenario, where the Fed does nothing as the RRP gets depleted, would result in an even more dysfunctional market than the 2019 repo crisis. Buy-side liquidity in Tbills and bonds would evaporate quickly, tightening financial conditions and triggering a risk-off response. In the worst-case scenario, hedge funds would be unable to finance their holdings of cash Treasuries, and margin calls would force them to unwind their basis trades. The Fed would be forced to react by rapidly expanding their balance sheet like they did in March 2020. I envision a V-shaped selloff and subsequent recovery.
The second scenario, where the Fed proactively ends QT and injects liquidity back into the RRP and the banking system, would be a bullish catalyst without the downside whiplash. I anticipate that as the balance in RRP falls below $500B, we’ll start hearing conversations about ending QT to put a floor under how much liquidity is in the financial system.
For now, I’ll be keeping a close eye on how quickly the RRP gets depleted as Tbill issuance continues to hit the market. A faster depletion will suggest waning demand for Tbills and notes, forcing the RRP to do the heavy lifting in absorbing the supply.
Great summary Geo. Ironically, Fed and the authorities in general have created multitude of these new facilities to reduce the various tail risks in market since GFC. They did it to suppress vol so as to increase risk taking (lending, animal spirits etc). Fed swap lines to other central banks are similar instruments. But due to all these additional facilities removing or reducing various tail risks, this time around they are finding that when they want to reduce risk taking (lending to broader economy, or asset market buying by financial players), it’s become harder to put the brakes on the economy.
Becoming more and more relevant as we approach March