I wish I could have sat in that meeting, watching the bewildered faces of Fed officials as they got hourly updates on repo rates blowing out.
The FOMC meeting minutes released this afternoon – instead of being stuffed with mind-numbing language – were spiked with a practically riveting account of the repo fiasco as it was unfolding over September 17 and 18 while the FOMC was meeting.
The account lays out some reasons behind the repo fiasco, the Federal Reserve’s reaction to it, and the changes it implemented and is going to implement to prevent the repo fiasco from spiraling further out of control.
Deep trouble in the repo market that caused “volatility in unsecured rates” had already percolated into the minutes of the July 30-31 FOMC meeting. But apparently nothing had prepared Fed officials for what would happen in the repo market on September 16, the day before their meeting, and on September 17 and 18, as they were meeting: Repo rates blew out.
I wish I could have sat in that meeting, watching the bewildered faces of Fed officials, Fed staff, and other participants as they were getting hourly updates on where repo rates were at the moment.
Day 1 of the meeting: repo rates blew out, Fed responded with $75 billion repo. From the minutes:
“Money markets were stable over most of the period [since the last meeting], and the reduction in the interest on excess reserves (IOER) rate following the July FOMC meeting fully passed through to money market rates.
“However, money markets became highly volatile just before the September meeting, apparently spurred partly by large corporate tax payments and Treasury settlements, and remained so through the time of the meeting.
“In an environment of greater perceived uncertainty about potential outflows related to the corporate tax payment date, typical lenders in money markets were less willing to accommodate increased dealer demand for funding.
“Moreover, some banks maintained reserve levels significantly above those reported in the Senior Financial Officer Survey about their lowest comfortable level of reserves rather than lend in repo markets [and so they didn’t lend to the repo market].
“Money market mutual funds reportedly also held back some liquidity in order to cushion against potential outflows.
“Rates on overnight Treasury repurchase agreements rose to over 5 percent on September 16 and above 8 percent on September 17. Highly elevated repo rates passed through to rates in unsecured markets.”
“Federal Home Loan Banks reportedly scaled back their lending in the federal funds market in order to maintain some liquidity in anticipation of higher demand for advances from their members and to shift more of their overnight funding into repo.
“In this environment, the effective federal funds rate (EFFR) rose to the top of the target range on September 16.
“The following morning, in accordance with the FOMC’s directive to the Desk to foster conditions to maintain the EFFR in the target range, the Desk conducted overnight repurchase operations for up to $75 billion. After the operation, rates in secured and unsecured markets declined sharply. Rates in secured markets were trading around 2.5 percent after the operation.
“Market participants reportedly expected that additional temporary open market operations would be necessary both over subsequent days and around the end of the quarter. Many also reportedly expected another 5 basis point technical adjustment of the IOER rate.”
Day 2 of meeting: second repo, from later in the minutes:
“Open market operations conducted on the previous day had helped to ease strains in money markets, but the EFFR had nonetheless printed 5 basis points above the top of the target range. With significant pressures still evident in repo markets and the federal funds market, and in accordance with the FOMC’s directive to maintain the federal funds rate within the target range, the Desk conducted another repo operation on the morning of the second day of the meeting.
“The staff presented a proposal to lower the IOER rate and the overnight reverse repurchase agreement rate by 5 basis points, relative to the target range for the federal funds rate, in order to foster trading of federal funds within the target range.”
Repo-rate blowout threw Fed’s policy into turmoil.
With overnight rates and the EFFR blowing out and escaping the Fed’s control, something wasn’t working as planned, and a new policy with different elements is now on the front burner to deal with these issues, particularly:
To increase balances of excess reserves that banks keep at the Fed – the “ample reserves” – so that banks would be incentivized to provide liquidity to the repo market, which they failed to do during those tumultuous days.
This means the Fed starts buying Treasury securities again – and as it pointed out, only short-term Treasury bills. But this is not QE, the minutes say emphatically, and the idea that this is not QE needs to be communicated to the markets, the minutes also say emphatically.
To dust off its “standing repo facility,” which is what the Fed used to have before September 2008, when it abandoned it in favor of QE and zero-interest-rate policy.
From the minutes, underscore added:
“Participants agreed that developments in money markets over recent days implied that the Committee should soon discuss the appropriate level of reserve balances sufficient to support efficient and effective implementation of monetary policy in the context of the ample-reserves regime that the Committee had chosen.
“A few participants noted the possibility of resuming trend growth of the balance sheet [“trend growth” was about 4% per year before 2008] to help stabilize the level of reserves in the banking system.
“Participants agreed that any Committee decision regarding the trend pace of balance sheet expansion necessary to maintain a level of reserve balances appropriate to facilitate policy implementation should be clearly distinguished from past large-scale asset purchase programs that were aimed at altering the size and composition of the Federal Reserve’s asset holdings in order to provide monetary policy accommodation and ease overall financial conditions.”
“Several participants suggested that such a discussion could benefit from also considering the merits of introducing a standing repurchase agreement facility as part of the framework for implementing monetary policy.”
Fed Chair Jerome Powell supplied some additional tidbits.
Yesterday, Powell came out and explained how the Fed would undertake these asset purchases: The Fed would buy short-term Treasury bills only. Longer-term Treasury securities would not be purchased under this program. By buying only short-term Treasury bills, long-term rates should not be impacted, and therefore these purchases should not act as a stimulus, Powell’s logic went – though buying short-term Treasury bills is precisely what the Fed had done during the early phases of QE.
But what brought long-term rates down during QE was Operation Twist, under which the Fed shed its short-term securities and replaced them with long-term securities; and QE-3, which was entirely focused on longer-term Treasury securities and Mortgage Backed Securities. By the time QE ended, the entire Fed portfolio was composed of longer-term securities with the average maturity exceeding 8 years.
Under the new-new plan, the average maturity would decrease at a faster rate than under the old-new plan outlined earlier this year, which should eventually contribute to a steepening of the yield curve.
In an initial reaction, since Powell’s explanation yesterday, and supported by the minutes this afternoon, all Treasury yields of 2 years and longer rose, with the 10-year Treasury yield up 7 basis points to close today at 1.59%.
The hullabaloo in the repo market torpedoed the function of Interest on Excess Reserves and forced the Fed to go back to the future. Read... Fed Admits Failure of ‘Plan A’ to Control Money Market Rates, Shifts Back to Repos (which was ‘Plan A’ till 2008)
Enjoy reading WOLF STREET and want to support it? You can donate. I appreciate it immensely. Click on the beer and iced-tea mug to find out how:
Would you like to be notified via email when WOLF STREET publishes a new article? Sign up here.
Speaking of secured and UNsecured markets, I never understood why any bank would lend reserves to another bank overnight without having the loan secured by specific collateral. The Fed certainly only this repo, which by definition is secured.
(Try again with brain engaged)
Speaking of secured and UNsecured lending, I never understood why any bank would lend reserves to another bank overnight without having the loan secured by specific collateral. The Fed certainly only does repo, which by definition is secured by specific collateral.
Morover, it baffles me that the OBFR overnight bank funding rate (unsecured) today was 1.82%, whereas the SOFR (secured overnight funding rate( (yes, SECURED) was HIGHER at 1.85%.
The only explanation I can come up with is a combination of two factors: That the banks that borrow in the OBFR market are “known good banks”, and that some of the collateral offered in the SOFR market wass/is less than pristine.
Reference: New York Fed market dashboard (google it).
When the fed is the lender they charge more and require more to discourage the banks from coming to the fed window, the extra charge is a penalty. When the banks lend to each other it is suppose to be cheaper to encourage them to maintain(now reestablish) a repo market.
OBFR and SOFR are both non-Fed loans and rates. The questions stands why the unsecured loan was LESS expensive than the secured loan.
The Fed has become the lender of choice when no other banks, being risk-averse, step up.
Is the Fed becoming more like Dutch dike boy who plugged the leaks with his fingers? Could this be the canary in the coal mine and portend liquidity problems ahead? Will DB blow-up and unraveling of CDOs be bigger than AIG/Lehman/Bear Stearns combined?
This is stealth debt monetizing by the Fed. First the balance sheet in the trillions that is never going to be unwound as promised by Bernank and now temporary repos that seem to be anything but temporary, a QE that is notQE. It is obvious by now that any attempt to unwind the balance sheet will send the market in a tailspin and as the Fed’s raison d’être is the continuous pumping of the stock market and assets, money printing will continue until morale improves.
Eventually this will come crashing down for reasons that no one can tell today, but look for Dow to hit 30k in the near future first.
I’m speculating that all the so-called “excess” reserves have leaked into the private (interbank, non-Fed) repo market and enabled huge new issuance of corporate and mortgage debt.
SOFRVOL (look it up at FRED) went from the 700-900B overnight balance in 2018 to a peak of 1284B (yes, that is 1.3T USD) on 2019-09-04. At that point, all the banks holding “excess reserves” got spooked and the reserve shortage ensued. How the Fed could not have noticed this, and clamped down on it already, is unfathomable
Zerohedge had an article 1-2 days ago by someone at Curvature Capital that described how, for example, mortgage REITs, were using rolling overnight repo to fund their mortgage purchases. Basically, funding long-term debt with cheap short-term (overnight!) repo borrowing operations.
How the Fed could not have noticed this, and clamped down on it already, is unfathomable. I guess they do not look under the hood at banks that have “excess reserves”? But shouldn’t they, being the main regulator and all? It’s scandal. Shadow banking rides again.
NARmageddon,
Borrowing money in the short-term or overnight market and investing it in Agency MBS is the oldest game in town. It’s almost free profit. For example, American Capital Agency Corp. has been doing this for a decade. Except suddenly it — borrowing short-term, lending long-term — tends to blow up.
BTW, in this business — borrowing short-term, lending long-term — banks are king of the hill.
But the question is why s the Fed not clamped down on this dangerous activity?
Because the FED is smaller
than he market. Surprise? You think FED is God?
Here is my fake news kind of guess of what’s happening to illustrate why FED can’t do shit about repo.
Repo is like pawn shop market for securities. One comes in the pawn shop saying I have this mortgage paper, I can give this to the shop in exchange of money with a promise to buy back the security with interest. The mortgage paper is the collateral. FED says:” All pawn shops! Do it with interest rate at 1.5%! If you don’t do it, I will print to do it and you make no money!” For a long time, the pawn shops obeyed. Now the customer is spoiled, he come in to the pawn shop with a piece of shit and asking for 1.5% rate loan. Fed says:” pawn shops! Do it”. Pawn shop says:” FED, we don’t do it, you do it”. So the FED printed money and take in shit as collateral. Why FED is NOT God? Because FED can’t make people eat shit. Now they have to shove in shit into their own throat. What is that shit? It is bad collaterals. It is NOT treasury paper. It is NOT good mortgages. It is a paper with somebody else promising some mortgage promised by some other guy who is promised by some other guy. This is just a bad example of what “shit” collateral means, don’t take it as realistic case since I am NOT in the industry to know the details. But I do know when everything turns into shit, FEF has to eat it themselves.
JPM ultimatly decide that the Bears deal did not taste very good, and gave it big indigestion for a long time, perhaps this indigestion memory has tempered its appitite for certain types of activity.
Like short-term lending to desperate counter-parties and helping friends in need. The FED at the time being a BIG friend with a BIG need.
If so, is this a bad thing?
Many keep,pointing the finger at DB for being the needy party, if it is. I would not want ti be in that lending pool. As today the Bundesbank can not “Print” Db into solvency.
JZ,
I love your shitty explanation.
@Memento mori , so, to get back to your comment, it is not stealth monetizing the debt, but shadow banks misappropriating the excess reserves for speculative purposes.
Well, the fed balance sheet grew by close to 200billions in the last weeks.
The Fed didn’t sell some other securities and use the proceeds to buy those short term treasuries or whatever garbage they bought but freshly printed 200billions instead.
Until those transactions are unwound that is stealth monetizing of debt by printing money.
If any thing , this balance sheet will grow more, it will never be unwound in our life time.
@NARmageddon i’m shocked, shocked to find that gambling is going on here!
i’m with memento mori on this one. it’s official, the debt will be monetized savers and the pension system be damned.
Okay, I agree. It is a (so far temporary ;)) monetization of some USG debt, performed in order to paper over the problem caused by private entities that have created too much debt. Again, the real problem is that the Fed did not prevent the excessive private debt creation.
Does anyone believe the guff about corporations draining the money supply because they had to suddenly pay taxes at the end of the year? I suspect a more likely explanation is a whole class of CEO’s and shareholders went rushing for the exits at the same time and wanted their money lickety-split. I can’t believe the Reserve is manufacturing more money so these scoundrels can leave before everything implodes.
See my comment about SOFRVOL above when it comes out of moderation.
“Does anyone believe the guff about corporations draining the money supply because they had to suddenly pay taxes at the end of the year?”
1. They’re not draining “money supply,” they’re drawing money out of the “money market” where they borrow short term.
2. Yes, quarterly tax payments by corporations are a large amount of money that suddenly — within a few days at the end of the quarter — gets moved from the private sector to the US Treasury. This is well known and was expected. But several other things happened on top of it, and some of them were unexpected — for example a major bank did not lend its reserves into the money market for whatever reason — and suddenly the whole thing froze.
Maybe I missed this, but do we know WHICH “major bank” didn’t lend its reserves?
Seems rather important, and reminiscent, not of ’08, but of ’07.
Who didn’t lend is not the important point, they all have reserves. The fact that they would rather sit on cash than chase extremely high yields is the most important fact. The banks chose cash over other types of collateral, they already have enough treasuries thanks to the govt deficit, they don’t need more. So the most likely reason is they don’t want to hold MBS under any circumstances.
No one officially names names. But JPM has been named through the rumor mill as one of them that didn’t lend.
My guess about the repo freeze is banks decided why should they risk in the repo market? The repo market makes them nervous right now. They know how opaque their own finances are so therefore they know their brethren are at least as questionable. Though they fondle each other daily in dark pool grottos they have limits as to their trust of each in the sunlight. So, they turned to welfare from the government. The Fed acquiesced, systemic risk, too big to fail etc., etc.
Wolf,
Now that the ‘market’ is pricing in a 75% chance of rate cut in October Fed meeting, wouldn’t a reduction down to a 1.75%-1.50% Fed Funds rate exacerbate the problem and potentially require increased liquidity injections to suppress them or am I missing something? If so, do you think this will make them hesitate to cut rates until they feel the coast is clear?
Sure expand the balance sheet by a couple hundred billion, but “This is not QE. In no sense is this QE” When did Powell learn the Jedi Mind Trick. Like you said sort term is what QE started out as.
As the saying goes: “Don’t believe anything until it is officially denied.”
Perhaps the Fed considers QE to be the buying of long term treasuries and MBS. Therefore, buying short term treasuries is a more normal operation designed to stabilize the banking industry in a fractional reserve banking system. So, the Fed is back to doing what it originally was meant to do and is getting out of the business of manipulating the whole yield curve (and buying MBS). This would be positive. Why the repo market blew up is another matter. Given how much press it has gotten I would expect the Fed to start hiding the repo market data soon. Why we continue to blow credit bubbles to the point of having them blow up, and who is the force behind this practice, may actually be a different matter as well. The Fed clearly has been drawn into the credit bubble blowing fold, but the perps are perhaps not the Fed but a larger force that is unlikely to be accountable. I’m not defending the Fed, but I think we do a little too much Fed-watching and not enough of keeping an eye on legislation that allows the financial industry to get away with high crimes and misdemeanors.
Sorry, but Wolf had it right. The Fed was buy short term until Operation Twist in late 2011 and calling it QE.
Zero Hedge had an informative piece by an experienced Repo expert entitled “Panic at the Repo”. Besides a look at the mechanics of repo operations the takeaway was that cash lenders made extraordinary profits by withholding cash during the early hours of trading forcing borrowers to scramble and bid up rates to get their cash.
As usual, the dealers actually plugged into the market saw their opportunity and they took it. The Fed, as usual, was unaware of what was going on even as it happened and had to intervene to stop the ‘gaming’ of the repo market.
On Saturday Sept. 14th Yemeni Shiite Muslims attacked Saudi oil facilities knocking out 5 million barrels of oil a day. There was a massive spike in the repo rate after markets opened on Monday the 16th. There were large transfers of capital from money markets to oil commodity futures and oil equities.
https://www.bloomberg.com/news/articles/2019-09-16/repo-market-chaos-drives-overnight-rate-up-by-most-in-months
Nah, that was a one-day event that might have exacerbated the repo problem for a couple of days. But the repo problem continues to this day. The one-day price-spike of oil totally collapsed over the next few days and dropped below where it was before the attack. This isn’t 2007 anymore, with the US and world awash in oil:
https://wolfstreet.com/2019/09/16/drama-in-the-oil-markets-but-this-isnt-2007-anymore/
The repo rate normalized within a few days. Oil outages were quickly restored. It showed short term liquidity was thin.
There are more problems coming. They turned out the lights in California to avoid wildfire lawsuits.
David Hall:
Ultimately aprox 750,000 utility (PG&E) customers were left without power. Many were without power for more than four days.
I am waiting patiently for someone to explain how the fed “fixed” this problem .
After listening to everyone here I am inclined to think Doug Noland has it right, No Coincidences.. http://creditbubblebulletin.blogspot.com/2019/09/weekly-commentary-no-coincidences.html Nominal rates is not the issue, it’s volatility.
Bookmarked the nice link – thanks.
Volatility may portend below the surface liquidity challenges as 1 key takeaway from 2008 is how seemingly plentiful liquidity dried out rather quickly once excrements hit the fan. Afraid history will repeat as CDO/CDS will rear its ugly head as the CBs instead of quashing it fanned it further to bigger bubble…
and there is certainly more liquidity in the system now, the situation coalesces when players leverage up for decreasing returns, (corporate junk selling at historic lows, the picking up dimes in front of a bulldozer metaphor). Even Noland is chagrined that spreads and CDS are behaving this well, but this time around the monetary spigot is running overtime ahead of the recession, which in terms of conventional economic policy results in bankruptcy and liquidation when the slowdown occurs. He also makes the point that the system cannot deleverage, and that leads to one conclusion, an inflationary recession.
Debt is either paid off or defaulted and kicking the debt can down the road into the eternity with cheaper and taking on more ZIRP debt by governments, folks living off their means and corporations (companies using debt not for expansion/working capital but for buyback to juice up return and management bonus) portend the HUGE debt crisis sometime in the future.
Then lo and behold a global leader appears in the horizon to give solve the debt problems by giving debt jubilees to all and new world digital currencies where one has to have some kind of biometrics (AKA mark of the beast) to buy and sell. Then takes over the world controlling everything in police state like what is happening in China. All prophseized in the Book of Revealtion…
Here’s another interesting viewpoint on “what happened” in the Repo market.
https://www.theinstitutionalriskanalyst.com/single-post/2019/09/23/Nationalizing-the-Federal-Funds-Market
This planned program of suppressing market price discovery is looking more and more like a game of wack a mole. Just how many mallets is it going to take to keep all the “moles(rates)” down?
With the US debt quickly closing in on 23 T$, the side game of ‘thumbs and mallets’ will soon fail to entertain “US”.
As of Sep 2019 there are in reality no more of the so-called excess reserves, because pretty much all the “excess” reserves are tied up in overnight lending. How do you know? Well, you compare the daily balance of Secured Overnight Financing (SOFRVOL) with the Excess Reserves of Depository Institutions (EXCSRESNS).
Look at the following graphs and compare the numbers around Sep 2019.
You will see that SOFRVOL has increased up to the level of the excess reserves.
https://fred.stlouisfed.org/series/SOFRVOL
https://fred.stlouisfed.org/series/EXCSRESNS
NARmageddon,
When a bank lends out its excess reserves, it takes the cash from its account at the Fed and transfers it to whoever it lends that money to. In other words, if a bank lends its excess reserves to the repo market to earn some interest there, that amount disappears from the Fed’s books and is no longer counted as excess reserves (the second chart you listed).
The money that is lent to the repo market cannot be simultaneous be sitting on the Fed’s books. It is in one place or the other, but not both.
I agree with Wolf’s above description of the reserve accounting, and retract the proposition that it was the crossing of SOFRVOL and EXCSRESNS that *caused* the interbank lending (including repo) freeze.
I had been hunting high and low for an explanation of why interbank lending froze, and the fact that SOFRVOL and EXCSRESNS reached the same level exactly as the crisis unfolded was such a juicy of coincidence that I thought I had found something. I suspended my faith in proper accounting for a moment there :).
The above mea culpa affects some of the remarks I have made in this thread before the present time, but not all of them. Thanks for listening.
If the Fed genuinely wants to know how many actual Excess Reserves there are, and get them out into the economy, it should continue to pay less and less interest on them. Best to phase the concept out altogether since mathematically it’s purely welfare for the banks, and empirically it’s a failure. A return to the pre-2008 repo system is long overdue.
The recent fiasco indicates that the Excess Reserves system is being gamed to provide banks with free lunches: interest on what are allegedly “excess reserves”, while simultaneously those reserves aren’t excess at all because they are being used somewhere else… The situation reeks of bank accounting fraud at taxpayer expense.
>> while simultaneously those reserves aren’t excess at all because they are being used somewhere else…
EXACTLTY!! It looks like excess reserves are being counted as excess (and interest-paying??) AT THE SAME TIME as the reserves are actually being lent out overnight in the private (meaning: non-Fed) repo market. Otherwise how can SOFRVOL be as high as the EXCSRESNS measure (look them up on FRED, my comment on this topic is above but still in moderation).
Here’s your link plotting Excess Reserves (converted to dollars) and Secured Overnight Financing Volume (also converted to dollars).
The rise in SOFV mirrors the fall in Excess Res.
It’s a pity the SOFV dataset doesn’t go back before 2018. Is there an earlier equivalent series?
https://fred.stlouisfed.org/graph/fredgraph.png?g=p8xy
They both dropped in late 2019. So, drop of excess reserves does not necessarily mean increased SOFR volume.
Reserves leak out with currency in circulation or Treasury purchases paid to Treasury TGA. Payments between banks generally shuffle the reserves between banks.
GFC repo is cash for collateral (usually Treasury or Agency MBS). The purchase of Treasuries by dealers is a simple coincidence whether there is excess reserves or not. Repo simply finances the cash to pay for Treasuries.
Repo plumbing knowledge can help us all.
“In an environment of greater perceived uncertainty about potential outflows related to the corporate tax payment date”
Too bad someone can’t ask the Fed about what the heck they mean by this. Why on earth would there be a ‘perceived uncertainty’ about corporate tax payments? That number would almost certainly be very well known. It might vary by a fraction of a percent based on economic performance in the quarter, but otherwise it is a very well forecasted number.
So what, were a group of major corporations threatening a tax strike and holding back their payments? I could imagine something like that creating large uncertainty about these quarterly tax payments. But otherwise, this sentence sure reads like BS from the Fed.
The theory that a major wall street player was having big money problems makes a lot more sense. That’s what made the whole system freeze up on ’08, as everyone knew (except the public) that Citibank was in trouble, but the other players didn’t know who else was exposed in a web of derivatives. Thus, the banks were afraid to lend to other banks in case they might be left holding the bag when the whole thing collapsed.
Someone or someones are in trouble. Nobody knows how far the contagion might spread. Thus, they all want to keep their money closer to home, thus they stopped the lending to other banks just like they did the last time. That makes a lot more sense than “perceived uncertainty” in the amount of corporate tax payments.
Deutsche Bank is the primary lender who publicly appears to be in the most trouble. JP Morgan has pulled a huge amount ($150B) from the Fed Reserve in the last year.
Note also that the ‘temporary’ operations have now extended well beyond the ‘end of the quarter’ which was supposed to be the cause of all of this.
Personal Observations:
It does not take too much to key in Treasury Auction Results, NY Fed Primary Dealer Weekly Statistics, H.4.1 Factors Affecting Reserve Balances, SIFMA US Marketable Treasury Issuance and Outstanding, SIFMA US Trading Volume, TICDATA and DTCC GCF trading data; crunch them for similar dates, then make some observations. It beats listening to some of the clowns in Bloomberg and CNBC. Actually, you might get some fresh and interesting views like, for example, who is the new marginal buyer of treasuries and how do you think they finance their positions.
Some key observations:
1.) Primary Dealer Inventory of long term Treasuries dropped to about $190b in the week ending 9/18. That used to be a high of more than $253b in the week of 5/29. Since you know, from TICDATA, that foreigners were not that gung-ho on Treasuries, then who bought $60b worth of long-term notes and bonds from the hands of the dealers? These were non-dealers, non-foreigners.
2.) If you add up all the monthly auctions of Treasury Notes and Bonds for 2018 and 2019 (up to Sept) you will find that DIRECTS have been cavalier buyers at auctions for 2Y, 3Y, 5Y, 7Y, 10Y and 30Y. (It is no wonder the yield curve inverted.) Make sure you understand that Directs are NOT Primary Dealers. They probably are not banks with excess reserves at the Fed (a great buffer for liquidity needs). In 2019 alone, directs had a HIGHER monthly going rate (compared to 2018) of about $14.5 billion. Remember that is per month.
3.) How do you think Non Dealers (aka Directs) fund their (1) and (2) purchases? Well, repo of course. Coincidentally, the DTCC expanded the list of members that can trade and net Treasuries at the FICC. See http://www.dtcc.com/news/2019/april/01/dtcc-to-expand-sponsored-service-to-bring-greater-capacity-to-the-repo-market
While these funds were probably previously net-lenders, they now have become net-borrowers at repo. Meanwhile, the dealers did not have to do that much more to swallow new Treasury Notes and Bonds. The natural demand left from foreigners and the new demand from Directs took care of that.
4.) In the last two weeks of Sept 3-13, long term treasury yields spiked:
2y: from 1.47 to 1.79 or 32 bps = 21.8%
3y: from 1.38 to 1.76 or 38 bps = 27.5%
5y: from 1.35 to 1.75 or 40 bps = 29.6%
7y: from 1.42 to 1.83 or 41 bps = 28.9%
10y: from 1.47 to 1.90 or 43 bps = 29.3%
This meant that the prices of long-term Treasuries likely dropped. In the overnight repo market, when trades are unwound daily, this would mean than one had to raise more collateral, or post cash for the difference. If you do not have access to cash, like bank reserves, this could be a tight squeeze.
5.) Even if your collateral is pristine (e.g. Treasuries) the price fluctuates. Repo collateral is designed NOT to be volatile. You don’t like the price to move on you while you repo. Unfortunately, it did that week. The Fed did what it was supposed to do – provide a backstop. The lender of last resort.
The above are only my opinions. Thank you for reading.
Good stuff. I have not digested it all yet but keep these insights coming! And as you said above, “Repo plumbing knowledge can help us all”. Very true.