3
$\begingroup$

I think this will be an easy question for practitioners but going through several websites & papers is causing more confusion than anything else. So here goes: The effective federal funds rate is defined as the rate at which depository institutions lend reserve balances which are held at the Fed to each other overnight. This is uncollateralized. In the repo market, banks can also make overnight borrowing to each other, but here a financial instrument is delivered as collateral. So the question is, why is repo nowadays higher than the effective fed funds rate? The counterparty is the same, but repo is collateralized, so I would expect it to be lower. Is there a guarantee on lendings of reserve balances by the Fed made to these loans?

Thanks a lot for any help understanding this, the internet seems to be vague or contradictory!

$\endgroup$
2
$\begingroup$

In general, the Fed Funds rate is below the repo rate. That is because of a few things:

  1. With the introduction of interest on excess overnight reserves (IEOR) banks can park their money at the Fed and get paid for them. They have less of a need to move their money. This means that banks with idle reserves no longer need to go to the Fed Funds market to lend them out. They can get IEOR by default.

  2. There are far fewer participants in the Fed Funds market than are in the repo market. The credit quality is the highest for these participants.

  3. The agencies were large participants in this market - now they have shrunk their portfolio and with it their participation.

  4. The largest participants now in the Fed Funds markets are the few entities like the Fed Farm bank that can't get IOER - that entices them to lend dollars(reserves) via Fed Funds - that's better than getting nothing. There's just not as much activity in this market - making it a less meaningful benchmark compared to repo trading and IEOR. The agencies (Freddie and Fannie) are also Fed Funds participants, but they no longer have the same level of funds to lend or borrow as they have smaller core portfolios.

  5. The people who lend Fed Funds usually can't trade repo (I think)

  6. There's some sort of credit enhancement to a Fed Funds trade (I think - I'm trying to get more info from our repo desk at work about this - I'll update this once I find out). I believe you have an enhanced position in the Fed Funds market compared to regular inter-bank lending.

There are also nuances to how different loans/repos settle. I'll try to get some more detail, but I think tri-party repo is EOD and bi-lateral repo is NOW. They all send cash over FedWire, but the timing makes a difference. I'm not sure when a general Fed Funds loan settles.

Also, in terms of counter-party access, often a person, like a money market fund, cannot lend to a user like a hedge fund, who needs to repo out their positions. This need to intermediate through adds friction to the process of arriving at one true rate.

These frictions make it hard for a Fed Fund lender, like the FHLB's to take advantage of the higher repo rate. They are lending into a smaller, more captive pool. Now if this went of for a few more days, then it probably would bleed in.

One other detail, who are the people with money to lend out for repo? That would be: banks with deposits, funds who have cash put them into money market funds that run repo programs (instead of just letting the cash sit at their bank or prime broker), and corporate treasury departments usually via money market funds.

One important thing to remember is that the total value of reserves and the total value of repo are really somewhat independent. There is about \$3.8t of Fed Balance sheet. I think there's about \$1.7t in cash out there somewhere (although I'm sure some of that is lost/destroyed). That gives you about \$2t of actual reserves. (Including foreign and Treasury accounts and the Fed). Now, you could have no one trading any repo, or people trading \$20t of repo. The amount of reserves doesn't directly correlate.

Another interesting thing. Users of the Fed' Foreign Reverse Repo facility usually have access to the general repo market. But, they were parking their money at the Fed in return for SOMA securities (the Fed's stash of Treasuries) instead of earning the higher rates. That's about \$300b of collateral - multiple times the \$75b that the Fed added this week.

$\endgroup$
  • $\begingroup$ Thank you, those seem indeed valuable reasons! For the credit enhancement, I'd be interested in hearing the update once you find out, thanks a lot :) $\endgroup$ – blah_crusader Sep 22 '19 at 9:01
  • $\begingroup$ Thank you! I have a few additional questions. 1. Why do banks care about less need to move money given IOER if they can find higher return with fed fund rate? 2. Who are the participants in the fed fund market, is it just depository institutions and GSEs? When you refer to agencies, are you talking about the likes of Fannie Mae & Freddie Mac? 3. On the comment for Fed Farm bank participating in fed fund, are you referring to lending side? $\endgroup$ – gtr32x Sep 22 '19 at 10:15
  • $\begingroup$ 4. Who are the primary suppliers of money side in the repo market? I'm thinking money market fund, and banks and large fund? 5. I'm curious on the claim where fed fund lender cannot trade repo. E.g. FRED data shows that excess reserve recently amounts to roughly \$1.3T, and repo market accumulates to \$1.2T, which signals to me that much of the fund for repo market come from excess reserves, or is that a simple happy coincidence? Thank you so much! $\endgroup$ – gtr32x Sep 22 '19 at 10:40
  • $\begingroup$ Hi, good questions. I'll edit the post to clarify. $\endgroup$ – JoshK Sep 22 '19 at 14:18
  • $\begingroup$ The Economist has a great article on the subject as well last week: economist.com/finance-and-economics/2019/09/19/… . Probably not availble for read without subscription unfortunately.. $\endgroup$ – blah_crusader Sep 23 '19 at 8:54
1
$\begingroup$

Overnight repo rates have indeed tended to be slightly higher than Fed funds rates in the last few months. When this difference is small (say 5-10bp) I would say that most banks that have access to both markets would regard this as too small to exploit (not worth deploying scarce balance sheet for that small a return).

However there have been some instances where the difference has been much larger, such as earlier this week when repo rates traded as high as 9% while Fed funds only reached around 3%. There are a few possibilities here (I) the Repo market participants include non -banks who do not have access to Fed funds, so non banks could have driven up repo rates. (2) the Fed funds market has fallen into disuse since the post -QE era of excess bank reserves , and volumes in that market have declined substantially as noted by a previous answer. Hence there could simply not be enough financing available in that market at short notice.

The third explanation for this week’s anomaly is that indeed it was an anomaly that some participants were able to take advantage of.

$\endgroup$
0
$\begingroup$

I've also looked for this for the past several days and have not come up with much similar to you. Though I would like to throw out a speculation that there are different degree of creditworthiness attributed to overnight lending entities/reasons and most aren't justified in the unsecured lending scene, causing much volume in the repo market.

If we look at the FRED data, it appears that while the federal funding volume actually dropped from \$300B+ in 2016 to just barely over \$100B a day, the secured funding volume has in fact gone from barely \$800B to \$1.2T within the past year. This is signalling to me that there is an overall risk aversion towards secured lending vs unsecured lending.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.