This is still hands-down the most valuable place for me to learn about markets and finance. I can't thank you enough! The visual representations here were perfect, well done!
My only question is, when RRP rates start going down again along with fed funds rates, this "absorbed" liquidity will make it to the markets again pushing prices up again, and pushing inflation up again?
Because if a recession hits, supply side will be weaker, and if you inject thus liquidity again, prices would soar. What am I missing?
If dollars come out of the RRP for any reason it will be liquidity enhancing (think QE). This is probably more meaningful for asset prices and financial markets than inflation (though the two are obviously still connected). Today, bank deposits are probably a better measure of real-world consumer spending power, and those deposits have not shrank even with reduced liquidity in the financial system. For more details on that, I recommend The Furnace: https://thelastbearstanding.substack.com/p/the-furnace
Thank you for taking the time to write this out; your elegant analogy answers a host of questions that I’ve had for months about the role of the RRP. One question which remains for me: how is one to interpret the significance of the continual increase in funds parked in the RRP?
I guess it depends on what you mean by significance.
If the question is, why is so much money there - the answer is that $2.2t is the amount of dollar liquidity that needed to be removed in order for the Fed to bring short term rates up to their target (305bps today). This is backsolved - the $2.2t was not a known figure up front, rather it is the "variable" parameter to hit a "fixed" overnight rate target.
If the question is, what is the effect of the large RRP uptake - the implication (in my opinion) is a reduction in market liquidity and dollar availability. This manifests in increased volatility and price declines across asset prices and well as the strength of the dollar vs. foreign currency.
Well, it certainly controls the front end of the curve. The term YCC as far as I understand it typically is referring to the long end. In other words, BOJ setting a yield target on 10yr JGBs is outright yield control. QE/QT/RPR could maybe be called "YCI" - Yield Curve Influence. They aren't setting outright targets for the long end that they defend, but they are using tools to push long yields one way or another
Wow, you knocked it outta the park with this one! That simple illustration you used makes so much bloody sense. I'm sitting here with my coffee staring out the window and thinking it all over. Ty
Another cool analogy! (The furnace one was absolutely brilliant BTW, it's stuck in my head.)
You deserve credit for this ON-RRP thing, you are the first one I saw spotlighting it as a factor for asset price devaluation. After you I found @maxjanderson on Twitter running a very compelling graph of Fed Balance Sheet ex-ON-RRP vs SPX, you should check it out.
At first I was skeptical, I didn't see a reason why ON-RRP would be any different than Bank Reserves at the Fed, it was just supposed to be a reserve account substitute for non-banks. The same way as banks put money with the Fed in their reserve accounts, MMFs parks them in the ON-RRP. Why would one have more impact on asset prices than the other?
What seems to be happening is that Reserves are a "higher octane" form of liquidity than ON-RRP. For a fixed size of the Fed Balance Sheet, any transfer of liabilities from Reserves to ON-RRP leads to muting of economic and financial transaction activity. It all makes sense when you see things from the POV of the actual economic actors, who are the liabilities counterparties to the Banks' and MMFs' assets (Reserves and ON-RRP, respectively). Bank deposits yield ~0.2% while MMFs are around 2.5%.
So what is happening is people (and other economic entities) are transferring money from Banks to MMFs, and keep them there. These MMF deposits are simply sleepier, lower-velocity money, they are less likely to chase stocks or cryptos than bank deposits.
It's certainly an interesting observation and gives the Fed another lever to control economic activity and asset valuations, by simply changing the delta between the interest on Reserves and ON-RRP. If they want to stimulate they can lower the ON-RRP rate relative to Reserve Interest (the latter I think has to equal FFR), and vice versa.
Thanks Krassen. I have seen @maxjandeson's work, which is similar to the framework that I laid out for understanding the balance sheet in Down the Drain (and of course many other people have written about as well). The S&P500s correlation to bank "net liquidity" has certainly been uncanny in over the past year, but would also caution against such a tight reading of a "fair value" as max draws in his work. For example, liquidity was draining from 2017-2019 while stocks rose (though with bouts of volatility). Similarly the Fed opened the liquidity bazooka in late February 2020 and throughout the entire March 2020 COVID crash yet the market still fell >30% in a month (though rebounded strongly obviously). It will be interesting to watch the correlation going forward.
Your comment on bank reserves vs. ONRRP is an interesting counterpoint to consider. I think that the liquidity effect of the RRP is different than just low-velocity reserves though. Low velocity bank reserves are still private sector liquidity and support financial markets and lending/borrowing. When the RRP is utilized it effectively takes a dollar out of circulation in funding markets whereas bank reserves may just be inert but they are still accessible. Of course the two are not independent either. Increased RRP usage removes bank reserves (other things equal), and vice versa.
On your last point - the Fed may have more levers but at the same time they are limited by the construct they have created to control short term rates. The RRP and IOR are key parts to the O/N "corridor" that they are targeting and so changing the spread between the two can throw off that dynamic. Plus, its not clear that a minor increase in IOR would pull cash out of the RRP. Banks still pay far less interest on deposits than MMFs pay in interest today even though the IOR is higher than the ONRRP rate - so depositors/investors still have incentive to park money in a MMF today.
Bear, would you expand on this technical point of "Increased RRP usage removes bank reserves (other things equal), and vice versa"? When my deposit moves from MyBank to MMFbank, the reserves move over. Then what happens next to the reserves? does it depend who the MMF purchases bills from? Could they be transferred or destroyed?
Thanks TLBS. On Bank Reserves vs RRP, it took me awhile to figure out the money flow hydraulics, but I think I got it right.
Both of these are overnight money. When the Fed injects liquidity via QE at the end of every business day the money they injected needs to show up on the Fed's balance sheet as reserves, and ONRRP, and Treasury money, otherwise the Fed's balance sheet would not balance (Physical cash is a bit of a different story....). So the QE liquidity "sleeps" at the Fed. The question is what it does during the day.
Both bank reserves and ONRRP reflect private money: bank deposits and MMF deposits. But with MMF deposits yielding ~2.5% these private money don't transact and the MMFs just roll over their repo night after night. The bank deposits seem to be more transactional.
People sometimes confuse stocks and flows. The stock of money injected by the Fed ends up inevitably on its balance sheet. Asset prices and economic activity are the flows of these money. Suppose Fed bought Treasuries from me via QE. What do I do with the money? I can put them in an MMF, collecting 2.5%, and the MMF would just repo the money to the Fed via ONRRP and roll it over day after day collecting 3%+ . That's a very short flowpath, with minimal economic "irrigation". However, if I deposit the money in a bank, it's more likely I will buy a new car or AMC stock or some NFT, etc. instead of letting inflation destroy the purchasing power, and so the money is generating economic activity or pumping up asset prices....
A lot of the more astute bloggers have said this: "The rate hikes will continue until something breaks" not until inflation goes down. That breakage will probably be a currency crisis like in the UK or Japan but much worse.
Yes, this post provides a bit more specifics on why a "break" might occur, and how to think about the effects across assets. Currency markets are indeed more interesting today than they have been since I've been alive (or attuned).
I'm not sure it is the "preferred" mechanism, rather that its the more effective enforcement to keep short rates at the Fed's target.
I think the Fed would prefer if they QT could be sized with such precision as to achieve that goal but its just not the case. Is $120bn /month of QE or $95/mo of QT precisely calculated numbers to hit a yield target? No because its impossible for the Fed to know exactly what effect a fixed quantity of QT/QE will have. In other words its a best guess as to what is appropriate to hit their targets. By contrast RRP is yield targeting with no fixed volumes - so it works more effectively but the ramification for liquidity are then not in the hands of the Fed.
Amazing analogy, one can remember something for life if that something is explained to them as if they were 5. Thanks Bear!
"No because its impossible for the Fed to know exactly what effect a fixed quantity of QT/QE will have." --> I suppose it is impossible for multiple reasons. For one, the effect would be the sum of billions of actors, rational or irrational, with different risk profiles, different biases etc. But apart from that, the Fed has a rather good view of the amount of dollars issued by the US government. But to what extent would they know what is the breaking point of the Eurodollar market? I imagine that in a largely financialized world, full of financial first and second order derivatives that is something very hard to know.
Regarding the RRP rate, you have mentioned a couple of times that the Fed "chose" to start paying out a yield that would increase in tandem with its FFR. Can the Fed from a legal perspective simply decide to turn that switch off, and adopt a new policy saying for example "the RRP rate will always be FFR - 50 bps" (hypothetical example). Are there any reasons why they would want to do that instead of simply letting short term bonds mature and simply let money find its own way? Also, would they be able to do such a thing or would they be perceived as untrustworthy? I suppose what I am asking is a) whether such a move would be feasible and b) your view on why they might want or might not want to consider such a move.
I think the Fed has good info into US dollar markets, and less insight into its various derivatives around the world. But even with "good info" they have still proven to make big mistakes, like significantly underestimating the required reserves of the banking sector in previous round of QT.
To your last question - the simple answer it that the Fed can adjust the RRP however it wants, whenever it wants, to meet its goal. However, the Fed can just as easily re-start QE and lower rates tomorrow if it wanted to as well. Lowering the RRP award rate would have the same effect as lowering the FFR. They COULD do this tomorrow, but its the opposite of what they want to do from a policy perspetive.
IMO, it is more likely that such a pivot would be in response to a "crisis" rather than to prevent one.
see, this is a bit misleading. The goal of QE was not to inject money yielding below FFR. Sure, the excess liquidity was 0% yield, but that was also the policy rate. Once FFR started climbing, all liquid liabilities of the FED (Reserves + RRP) were re-rated to the FFR. Only exception is cash in circulation.
QE intent was to swap higher yielding securities (Treasuries and MBS) for FFR yielding liquidity (Reserves and RRP).
"target yield" perhaps was misleading on my part. But directionally, QE's intent is to push down rates on the long end even if there isn't a specific "target" further out on the curve.
Thank you LBS, your observation about liquidity explains in part why the Vix has been sloping upward since de third quarter of 2021. It appears market participants recognize risk is increasing incrementally as the FED proceeds withdrawing liquidity. In your opinion, which option timing would make sense at this moment. A call on the Vix expiring this November or somewhat longer expiring January 2023. Please consider relating this article about liquidity to your 2021 article about volatility.
My conclusion was that we shouldn't expect a vol spike until the vol-of-vol turns around. VVIX fell dramatically through the summer but has been on the rise since mid-august...
Thanks for your kind reply. Exactly, I remember the 5 legged race description and per your observation, VVIX started cooperating.
Best
Chris
Vivid, crystal-clear analogy!
Its one that has been bouncing around in my head for a while and finally got put on the page. Glad you enjoyed, and thanks for reading, Edmund!
This is still hands-down the most valuable place for me to learn about markets and finance. I can't thank you enough! The visual representations here were perfect, well done!
I can't believe this is free content.
Outstanding.
Thank you so so so much.
My only question is, when RRP rates start going down again along with fed funds rates, this "absorbed" liquidity will make it to the markets again pushing prices up again, and pushing inflation up again?
Because if a recession hits, supply side will be weaker, and if you inject thus liquidity again, prices would soar. What am I missing?
If dollars come out of the RRP for any reason it will be liquidity enhancing (think QE). This is probably more meaningful for asset prices and financial markets than inflation (though the two are obviously still connected). Today, bank deposits are probably a better measure of real-world consumer spending power, and those deposits have not shrank even with reduced liquidity in the financial system. For more details on that, I recommend The Furnace: https://thelastbearstanding.substack.com/p/the-furnace
Thanks again for reading!!
Thank you for taking the time to write this out; your elegant analogy answers a host of questions that I’ve had for months about the role of the RRP. One question which remains for me: how is one to interpret the significance of the continual increase in funds parked in the RRP?
Hi Ernest,
I guess it depends on what you mean by significance.
If the question is, why is so much money there - the answer is that $2.2t is the amount of dollar liquidity that needed to be removed in order for the Fed to bring short term rates up to their target (305bps today). This is backsolved - the $2.2t was not a known figure up front, rather it is the "variable" parameter to hit a "fixed" overnight rate target.
If the question is, what is the effect of the large RRP uptake - the implication (in my opinion) is a reduction in market liquidity and dollar availability. This manifests in increased volatility and price declines across asset prices and well as the strength of the dollar vs. foreign currency.
I’ve been thinking about your reply, and I can’t help but ask if the RRP is not yield curve control by another name?
Well, it certainly controls the front end of the curve. The term YCC as far as I understand it typically is referring to the long end. In other words, BOJ setting a yield target on 10yr JGBs is outright yield control. QE/QT/RPR could maybe be called "YCI" - Yield Curve Influence. They aren't setting outright targets for the long end that they defend, but they are using tools to push long yields one way or another
Hindsight produces elaborate and beautiful narratives.
Something about 20/20.... though to be fair, many of the challenges today were foreseeable and raised by plenty of observers
Extremely well done analogy. Congratulations. You're a really Talented writer.
Thanks Sam! glad that you enjoyed.
Wow, you knocked it outta the park with this one! That simple illustration you used makes so much bloody sense. I'm sitting here with my coffee staring out the window and thinking it all over. Ty
Thanks! If i can make you sit and think i've done my job.
Another cool analogy! (The furnace one was absolutely brilliant BTW, it's stuck in my head.)
You deserve credit for this ON-RRP thing, you are the first one I saw spotlighting it as a factor for asset price devaluation. After you I found @maxjanderson on Twitter running a very compelling graph of Fed Balance Sheet ex-ON-RRP vs SPX, you should check it out.
At first I was skeptical, I didn't see a reason why ON-RRP would be any different than Bank Reserves at the Fed, it was just supposed to be a reserve account substitute for non-banks. The same way as banks put money with the Fed in their reserve accounts, MMFs parks them in the ON-RRP. Why would one have more impact on asset prices than the other?
What seems to be happening is that Reserves are a "higher octane" form of liquidity than ON-RRP. For a fixed size of the Fed Balance Sheet, any transfer of liabilities from Reserves to ON-RRP leads to muting of economic and financial transaction activity. It all makes sense when you see things from the POV of the actual economic actors, who are the liabilities counterparties to the Banks' and MMFs' assets (Reserves and ON-RRP, respectively). Bank deposits yield ~0.2% while MMFs are around 2.5%.
So what is happening is people (and other economic entities) are transferring money from Banks to MMFs, and keep them there. These MMF deposits are simply sleepier, lower-velocity money, they are less likely to chase stocks or cryptos than bank deposits.
It's certainly an interesting observation and gives the Fed another lever to control economic activity and asset valuations, by simply changing the delta between the interest on Reserves and ON-RRP. If they want to stimulate they can lower the ON-RRP rate relative to Reserve Interest (the latter I think has to equal FFR), and vice versa.
Thanks Krassen. I have seen @maxjandeson's work, which is similar to the framework that I laid out for understanding the balance sheet in Down the Drain (and of course many other people have written about as well). The S&P500s correlation to bank "net liquidity" has certainly been uncanny in over the past year, but would also caution against such a tight reading of a "fair value" as max draws in his work. For example, liquidity was draining from 2017-2019 while stocks rose (though with bouts of volatility). Similarly the Fed opened the liquidity bazooka in late February 2020 and throughout the entire March 2020 COVID crash yet the market still fell >30% in a month (though rebounded strongly obviously). It will be interesting to watch the correlation going forward.
Your comment on bank reserves vs. ONRRP is an interesting counterpoint to consider. I think that the liquidity effect of the RRP is different than just low-velocity reserves though. Low velocity bank reserves are still private sector liquidity and support financial markets and lending/borrowing. When the RRP is utilized it effectively takes a dollar out of circulation in funding markets whereas bank reserves may just be inert but they are still accessible. Of course the two are not independent either. Increased RRP usage removes bank reserves (other things equal), and vice versa.
On your last point - the Fed may have more levers but at the same time they are limited by the construct they have created to control short term rates. The RRP and IOR are key parts to the O/N "corridor" that they are targeting and so changing the spread between the two can throw off that dynamic. Plus, its not clear that a minor increase in IOR would pull cash out of the RRP. Banks still pay far less interest on deposits than MMFs pay in interest today even though the IOR is higher than the ONRRP rate - so depositors/investors still have incentive to park money in a MMF today.
Bear, would you expand on this technical point of "Increased RRP usage removes bank reserves (other things equal), and vice versa"? When my deposit moves from MyBank to MMFbank, the reserves move over. Then what happens next to the reserves? does it depend who the MMF purchases bills from? Could they be transferred or destroyed?
Thanks TLBS. On Bank Reserves vs RRP, it took me awhile to figure out the money flow hydraulics, but I think I got it right.
Both of these are overnight money. When the Fed injects liquidity via QE at the end of every business day the money they injected needs to show up on the Fed's balance sheet as reserves, and ONRRP, and Treasury money, otherwise the Fed's balance sheet would not balance (Physical cash is a bit of a different story....). So the QE liquidity "sleeps" at the Fed. The question is what it does during the day.
Both bank reserves and ONRRP reflect private money: bank deposits and MMF deposits. But with MMF deposits yielding ~2.5% these private money don't transact and the MMFs just roll over their repo night after night. The bank deposits seem to be more transactional.
People sometimes confuse stocks and flows. The stock of money injected by the Fed ends up inevitably on its balance sheet. Asset prices and economic activity are the flows of these money. Suppose Fed bought Treasuries from me via QE. What do I do with the money? I can put them in an MMF, collecting 2.5%, and the MMF would just repo the money to the Fed via ONRRP and roll it over day after day collecting 3%+ . That's a very short flowpath, with minimal economic "irrigation". However, if I deposit the money in a bank, it's more likely I will buy a new car or AMC stock or some NFT, etc. instead of letting inflation destroy the purchasing power, and so the money is generating economic activity or pumping up asset prices....
Nice article and absolutely brilliant graphic. It does a fabulous job conveying the concepts.
Thanks John! Always appreciate your feedback.
A lot of the more astute bloggers have said this: "The rate hikes will continue until something breaks" not until inflation goes down. That breakage will probably be a currency crisis like in the UK or Japan but much worse.
Yes, this post provides a bit more specifics on why a "break" might occur, and how to think about the effects across assets. Currency markets are indeed more interesting today than they have been since I've been alive (or attuned).
Curious, why is the Fed's preferred tightening mechanism RRP rather than QT?
I'm not sure it is the "preferred" mechanism, rather that its the more effective enforcement to keep short rates at the Fed's target.
I think the Fed would prefer if they QT could be sized with such precision as to achieve that goal but its just not the case. Is $120bn /month of QE or $95/mo of QT precisely calculated numbers to hit a yield target? No because its impossible for the Fed to know exactly what effect a fixed quantity of QT/QE will have. In other words its a best guess as to what is appropriate to hit their targets. By contrast RRP is yield targeting with no fixed volumes - so it works more effectively but the ramification for liquidity are then not in the hands of the Fed.
Amazing analogy, one can remember something for life if that something is explained to them as if they were 5. Thanks Bear!
"No because its impossible for the Fed to know exactly what effect a fixed quantity of QT/QE will have." --> I suppose it is impossible for multiple reasons. For one, the effect would be the sum of billions of actors, rational or irrational, with different risk profiles, different biases etc. But apart from that, the Fed has a rather good view of the amount of dollars issued by the US government. But to what extent would they know what is the breaking point of the Eurodollar market? I imagine that in a largely financialized world, full of financial first and second order derivatives that is something very hard to know.
Regarding the RRP rate, you have mentioned a couple of times that the Fed "chose" to start paying out a yield that would increase in tandem with its FFR. Can the Fed from a legal perspective simply decide to turn that switch off, and adopt a new policy saying for example "the RRP rate will always be FFR - 50 bps" (hypothetical example). Are there any reasons why they would want to do that instead of simply letting short term bonds mature and simply let money find its own way? Also, would they be able to do such a thing or would they be perceived as untrustworthy? I suppose what I am asking is a) whether such a move would be feasible and b) your view on why they might want or might not want to consider such a move.
Hello Mr. Cat!
I think the Fed has good info into US dollar markets, and less insight into its various derivatives around the world. But even with "good info" they have still proven to make big mistakes, like significantly underestimating the required reserves of the banking sector in previous round of QT.
To your last question - the simple answer it that the Fed can adjust the RRP however it wants, whenever it wants, to meet its goal. However, the Fed can just as easily re-start QE and lower rates tomorrow if it wanted to as well. Lowering the RRP award rate would have the same effect as lowering the FFR. They COULD do this tomorrow, but its the opposite of what they want to do from a policy perspetive.
IMO, it is more likely that such a pivot would be in response to a "crisis" rather than to prevent one.
see, this is a bit misleading. The goal of QE was not to inject money yielding below FFR. Sure, the excess liquidity was 0% yield, but that was also the policy rate. Once FFR started climbing, all liquid liabilities of the FED (Reserves + RRP) were re-rated to the FFR. Only exception is cash in circulation.
QE intent was to swap higher yielding securities (Treasuries and MBS) for FFR yielding liquidity (Reserves and RRP).
"target yield" perhaps was misleading on my part. But directionally, QE's intent is to push down rates on the long end even if there isn't a specific "target" further out on the curve.
exactly!
Thank you LBS, your observation about liquidity explains in part why the Vix has been sloping upward since de third quarter of 2021. It appears market participants recognize risk is increasing incrementally as the FED proceeds withdrawing liquidity. In your opinion, which option timing would make sense at this moment. A call on the Vix expiring this November or somewhat longer expiring January 2023. Please consider relating this article about liquidity to your 2021 article about volatility.
Thanks
Chris
This is a good idea - and I was tempted to get into volatility here in this post. But agreed - a follow up on vol is in order. The last post I've written on the topic was in May: https://thelastbearstanding.substack.com/p/the-volatility-squeeze-part-3
My conclusion was that we shouldn't expect a vol spike until the vol-of-vol turns around. VVIX fell dramatically through the summer but has been on the rise since mid-august...