The Fed sets interest rates as it sees fit, but it is losing control of its balance sheet policy to the growing fiscal deficit. Fed officials look to repo rates for signs of reserve scarcity, which influences their decision to shrink or expand their balance sheet. But repo rates are determined by supply and demand and not simply the level of reserves. Demand for repo financing continues to grow rapidly from the increasing fiscal deficit, while the supply of repo financing can only keep up with the help of the Fed. Stable repo rates would ultimately require persistent Fed balance sheet expansion and effectively cede control of its balance sheet size to the fiscal authorities. This post reviews the supply and demand factors that have driven repo rates above interest on reserves and suggests that balance sheet expansion will resume shortly after QT.
Infinite Repo Demand
The demand for repo financing is insatiable because the $2t fiscal deficit is in part financed by leveraged investors. Treasury repo volumes remained steady through most of 2022, but began a steady ascent as the supply of Treasuries to the private sector grew rapidly. Both the Fed’s QT program and continued net issuance increased the supply of Treasury securities to private investors. Some of the investors purchased Treasuries with cash, but others financed their purchase with a repo loan. These leveraged investors often bought cash Treasuries as part of a cash futures basis trade, which has grown over a trillion in size. Demand for repo financing will continue to grow with the growing fiscal deficit, which is expected to be about $2t annually for the foreseeable future.
Robust Repo Supply
The available amount of repo financing is enormous, but it is not enough to keep up with the growing demand. The primary repo lenders are the money market funds and commercial banks, who each have different opportunity costs. The largest pool of cash is the $7t money market fund complex, whose opportunity cost is the RRP offering rate. MMFs place cash in the RRP facility when they have nowhere else to invest, so a high RRP balance implies excess cash that could be used to meet repo financing demand. RRP balances have gradually declined to zero as MMFs deployed cash to fund repo loans and purchases of Treasury bills. As RRP balances dwindled, repo rates steadily rose to entice the next marginal lender.
The next marginal lenders are the prime money market funds and Federal Home Loan Banks, whose opportunity cost is the overnight unsecured rate. Prime funds and FHLBs lend in the overnight unsecured market because they can usually earn a little bit more return by taking on unsecured credit risk. The borrowers in the overnight unsecured market largely just park the proceeds at the Fed to earn interest on reserves, so they are not willing to pay up for funding. Repo borrowers can bid away cash from these lenders, who are happy to earn higher returns with lower risk. Volumes in overnight unsecured markets declined by $100b as repo rates steadily moved above overnight unsecured rates.
The final layer of repo lending is the commercial banks, whose opportunity cost is interest on reserves. Commercial banks hold about $3t in reserves, which are deposits at the Fed that earn IOR. A Treasury backed overnight repo loan is equivalent to reserves under a range of regulations, so banks feel comfortable lending in repo when it is trading at a spread to IOR. The amount of cash they have available to lend fluctuates with their business needs, so their lending is less consistent and can lead to greater rate volatility. Note that banks are net lenders in repo rather than borrowers, so higher repo rates do not imply banks lack reserves.
Reserve Management Purchases
The Fed is expected to stop QT by the end of the year in part due to rising repo rates, but repo rates will continue to increase even after QT. The Fed perceives higher repo rates as a sign that reserves have become scarce and are eager to avoid a repeat of the September 2019 repo spike, which they link to over doing QT. Reserve levels can impact repo rates through the bank lending channel, where lower reserve levels mean banks have less cash to lend into repo. But repo rates are rising in large part due to an increasing demand for repo financing, so just looking at the supply of cash is incomplete. Without a change in fiscal policy, any amount of cash will ultimately be exhausted.
A steady expansion of the Fed’s balance sheet is the most likely way to meet the growing financing demands of the Treasury. The Fed could do this through its standing repo facility, where it lends according to the needs of repo borrowers. The facility has so far seen limited take-up and does not appear to constrain repo rates. They could also buy Treasury bills to add reserves into the banking system, who could then lend them into the repo market. It is also possible that commercial banks could increase their lending in repo by expanding their balance sheet and not just shifting the composition of their assets, but that is not likely under current regulations and spreads. The end of QT is coming soon, and further growth in the Fed’s balance sheet will necessarily follow shortly.
59 comments On Balance Sheet Dominance
Hi Fedguy, so can we say this balance sheet expansion is one form of not-qe QE?
From simple rate cuts to a marginal shift toward quantitative easing.
I don’t think it’s QE because it’s not buying longer dated Treasuries, which would put downward pressure on long rates.
Hi Joseph, isn’t it effectively putting downward pressure on long dated yield by facilitating Treasury department to borrow at short end so Treasury never has to issue long dated debt any more?
Isn’t it just money printing though? thus bullish for markets
There are people who look at it that way, but I am not one of them.
i’m sure many of us here would like to understand your view on this, why you don’t think this type of money printing isn’t bullish for markets, thanks a lot Joseph!
can you please explain why this is not bullish?
Seems obvious to me that SOFR lending has been used to purchase equities and market has exploded using this lending. Obvious solution from my perspective is to just allow equities to sell off. That would generate the cash the market seems to demand. The SOFR rate should be allowed to rise to such a level that the cost of borrowing/lending exceeds the expected return on the equity trade.
Just because there is more demand for repo due to the basis trade, why does anyone including the FED need to respond? Asked another way, why should the FED get involved with the markets desire to put on the basis trade? Doesn’t facilitating the basis trade (especially if levered) add systemic risk?
They care about higher repo rates. That impacts their ability to control short term interest rates.
do you see balance sheet expansion leading to inflation again?
No, I don’t think there is a link between balance sheet expansion and inflation. Remember, the Fed’s balance sheet expanded massively for over a decade after 2008 and nothing happened.
what happened was asset inflation though right? including for real estate and rent which are part of CPI
When I look at history it seems assets go up, with or without QE.
Joseph, you need to read Milton Friedman. Only one way inflation occurs is the printing press (balance sheet expansion). Period. The more you print the more inflation one will get.
China’s emergence as a manufacturing power and entry into the WTO really kept a lid on goods prices for two decades. Pretty much every household good or piece of hardware simply got better yet cheaper and cheaper. Now that de-coupling is increasingly a thing along with tariffs and re-shoring, I believe that further Covid style balance sheet inflation will actually lead to goods inflation because the increase in the supply of goods this time will not outpace the increase in the supply of money.
Yes that sounds reasonable. One thing to note is that inflation is very unpopular, so the Administration could further roll back tariffs even on China to boost public support.
All in all, it’s money printing whether you call it QT, or by some other name. BTW, what kind of issuance does it take to close the arb?
The arb exists because lots of investors prefer to own futures rather than cash Treasuries. It has less to do with the composition of issuance.
Joseph, is it safe to say that there was indeed inflation through the 2010s, just only in the financial markets (not main street) because no fiscal mechanism pushed that extra liquidity to the consumer? Obviosly, Covid changed that…
Financial assets grew in the 2010s, but so did the economy. I think the difference post-pandemic is the fiscal spending. That is different from monetary policy, which just buys financial assets. Fiscal spending buys real goods and services.
hi thanks for the insight
Commercial banks hold about $3t in reserves, which are deposits at the Fed that earn IOR
here I think that the total amount of reserve stays same unless there is QE or QT
so can the reserve be the potential supply for the repo?
there were some papers from fed that when the balance is tight, the banks wont let go of this reserve and transaction between other banks as late as they can if i got the idea right.
what is ur thought for this?
Yes when repo rates trade above IOR banks will lend reserves into repo. That will help but ultimately the extra reserves banks hold will be exhausted.
Can you make a post about equities? Where you think the market is going.
Thank buddy! Love you stuff!
I think we’re in a mania, so all bets are off. We could go higher than any one expects, but then we’ll decline sharply.
Hi Joseph do you still think your weak dollar policy that leads to capital flight thesis will eventually play out? or have you changed your mind about this
The mini crisis was Sept 2019 not 2018 for the record. That WE the houthis bombed a Saudi oil facility and a large lender decided not to lend on fears of big grttting his bank’s funds back. The reserve shortage was a part of the excuse for the inflated repo rate. Too early to think QE is around the corner and the Fed could just do system repos when tightness occurs. Nice piece otherwise.
Thanks! will edit.
Fiscal also leads to more profits for companies which increases cash pools ultimately – do you think that once the government uses the TGA balance, the money will be back in the system calming this down?
Running down the TGA balance would help, but at the moment Treasury is happy to maintain their cash balance at its target of 850b.
Very informative post. I was under the impression that commercial banks could only lend their cash reserves to other commercial banks, but you indicate that they can lend cash to, or borrow cash from, anyone as long as the loan (repo) is collateralized by T-bills. Did I state that correctly?
Also, I am intrigued by your closing comment that the commercial banks could create money (expand their balance sheet) to participate in the repo market. Would you expand on why they do not do this already?
Thanks.
Because of the Supplementary Leverage Ratio (SLR), They can not expand their balance sheet unless they have more regulatory capital
Hi Joseph
I’m guessing you still like gold then?
Yes I will buy some back in the market view portfolio this week.
Banks can lend to anyone, but they ultimately settle in reserves with the borrower’s bank. When commercial banks lend in repo, they send reserves to the repo borrower’s bank. The repo borrower’s bank will then have a reserve asset balanced by a deposit liability to the repo borrower.
Manuel is correct on the SLR creating costs for balance sheet expansion.
Thanks for the post. I am confused why the Fed has been doing QT shrinking the balance sheet to a level that to carry out QE soon will be desperately needed. Couldn’t the Fed keep tbe balance sheet a sustainable level?
The demand for liquidity in the financial system grows over time, so the Fed has to adjust its balance sheet to meet that demand.
Thank you Joseph,
Will balance sheet expansion increase market liquidty or just helps banks ?
Second, budget deficit thus goverment borrowing needs are increasing sharply.. Will bond vigilents be happy with this expansion ?
Regards
It adds more cash to the financial system, so it should help everything. Or rather, just make sure everything still works. I don’t think Fed expanding its balance sheet to add cash is inflationary (remember we had a decade of QE!), so I don’t think the bond market cares.
Joseph, the repo lending chart is very interesting, but doesnt match my data which I get from FFUNDS (which reports depository institution repos at 565bn). . Do you mind sharing where your data set is from?
Joseph, I’m sorry I got it confused. I see institutional repo assets at 765bn and liabilities at 565bn…my error. thanks
Suppose the Fed ends up having to buy bills perpetually to inject more reserves in order to satiate the demand for repo borrowing, which is a function of the high budget deficit. Doesn’t that basically mean the Fed is monetising the deficit? I understand that it’s different to reducing duration (i.e. if they buy t-notes and t-bonds) but they don’t need to if the Treasury continues to favour issuing bills over t-notes and t-bonds.
Indirectly I guess, but the Fed has been monetizing the deficit via QE over the past decade pretty frequently. In this case it’s a little different, rather than buy Treasuries outright the Fed is indirectly lending money to an investor that buys Treasuries. So the interest rate risk stays within the private sector.
The Federal Reserve stopped QT+balance sheet restructuring and increased its holdings of T-Bills+SRF to maintain the upper limit of interest rates in the reverse repo market.
As for the relaxation of eSLR, it may not be so urgent?
The above is what I am currently observing the Fed.
Why does the fed HAVE to add liquidity to meet the demand for cash? This isn’t fiscal dominance or monetary dominance, it’s NDFI/hedge fund dominance. Who is in the driver’s seat here? Why is it out of the question for the Fed to let the market equilibriate on its own? Has the Fed always behaved this way? Why can’t the Fed let the cost of SOFR borrowing rise to such a level that demand just naturally tapers off? It might require a bit of a sell off in equities’ treasuries gold or bitcoin but so what? Markets can free up a trillion tomorrow with a 3-5% pullback in those assets.
Fed needs to control short-term interest rates, which is how they implement monetary policy. If they can’t control short rates, they can’t impact the economy. Short rates are not market determined but set by the government – this is true in all countries.
The Fed is not the government.
Great piece as always! The themes of higher deficits, higher issuance, increased repo activity and sovereign debt demand from leverage investors is playing out not just in US but other developed markets (ofc on different scales). Given that some CBs have moved to a demand driven frameworks (also hinted at by Logan for Fed), how would you tweak the CBs lending facilities to foster more bank intermediation in repo ? Central clearing of operations has been mentioned but be interested to hear on other features. Do you think can we learn something from the BOE’s STR facility given its persistent usage?
Thanks alot
There’s no easy fix because even if we tweak the bank regulations, the banks don’t have access to cheap enough funding to entice them to step into the market. Don’t think there is a solution other than hoping the fiscal deficit gets smaller. STR facility sounds like the standing repo facility – I guess they could lower the rate there and then end up being a big repo lender. Don’t think the Fed wants to do that.
Hi Joseph
I’m guessing you still like gold then?
Great piece as always! the themes you mention ( higher deficit, debt issuance, repo borrowing and sovereign debt purchases by leveraged funds) are also prevalent in other jurisdictions where there has been shifts to more demand driven frameworks which will rely more on lending facilities. Therefore if sizeable asset purchases are the first port of call, what design features do you think are needed to help repo market and I guess facilitate bank intermediation ? Central clearing is often mentioned but be interested your thought on other options. The BOEs STR facility seems to be getting decent demand so are there any insights from that? Thanks
“ if sizeable asset purchases are not the first port of call,
This isn’t my area of expertise but I did notice looking at the fed data that the daily SOFR volume in September/October appears to have eclipsed the amount of commercial bank reserves for the first time. Bank reserves fell to about 3T while at the same time SOFR volumes hit 3T. Is that the primary bottleneck here?
Why did bank reserves fall in recent months?
The two are not directly related, because commercial banks are not the only lenders in repo. Bank reserves have been falling in recent months monthly because the Treasury has been rebuilding its cash account at the Fed – the Treasury General Account. It grew several hundred billion to around $1t on Thursday. That sucked out a ton of cash of the financial system.
Very insightful thank you. You say that “The primary repo lenders are the money market funds and commercial banks” and then “The final layer of repo lending is the commercial banks” so, what is the weight of commercial banks in the repo market ? I would tend to say that they don’t play a huge role since IOR is usually above repo rates, but at the same time we can read everywhere that the stress in the funding market is (partly or primarily) due to reserves shrinking.
They are not a big player in the repo market – lending in repo is not a big part of their business. But they have lots of cash (reserves) and are willing to deploy that when repo trades above IOR. The key reason repo rates are rising is because demand grows faster than available financing – reserve levels play a role only in that more reserves would give banks more cash to lend.
Very insightful. Would be great to chat if you can email me to arrange. Andrew
Perhaps this is why an alternative to the FED absorbing issuance is being sought i.e. monetization of asset side of US the balance sheet. Namely, exiting FNMA/FMCC and/or re-valuing gold. We have an insanely large maturity wall coming up in <1 year