# Replacing USD OIS discounting based on FED Funds Rate with SOFR discounting

Slightly related to my other question (The exact mechanics of USD OIS Swaps: replacement of USD Libor by SOFR) but nonetheless, this is a separate topic:

US banks fund themselves via EFFR (Effective Federal Funds Rate), as well as the Secured Overnight Financing Rate (SOFR). There are OIS swaps on both rates, albeit the EFFR OIS swaps are much more liquid (at the moment): as per the answers to my other question.

Does anyone understand why the FED, alongside ISDA, decided to - going forward - drive a shift from EFFR OIS discounting to SOFR OIS discounting? The fact that EFFR OIS swaps tend to be much more liquid tells us that banks utilize funding via EFFR more than via SOFR. Why would the FED (and ISDA) encourage a shift in discounting towards a less liquid curve? Intuitively, I cannot really make sense of it: it would only make sense to me as a "marketing" or "political" move in an attempt to make SOFR a more prominent rate (as part of LIBOR cessation).

• Well, your intuition is right. Sometimes you need to force people to do something, otherwise they might just stick to the status quo. In particular there has been some resistance to SOFR, evidenced especially during the March-April stress, because it's worse at capturing credit risk (as opposed to LIBOR), hence on banks' client side SOFR is actually not a perfect substitute for LIBOR as it does not allow to hedge some credit-sensitive exposures. There are discussions going on about complementary reference rates which might be better at capturing credit risk, search for Ameribor. Sep 30, 2020 at 16:36
• Thanks @DaneelOlivaw: good points. I can understand the drive to phase out LIBOR with SOFR, but I do not understand the drive to replace EFFR OIS with SOFR OIS. Do you have a view on that also? Sep 30, 2020 at 16:46
• I am not familiar with a push from regulators to shift from EFFR to SOFR $-$ not saying it's not the case though. Major exchanges are gradually shifting to SOFR for collateral interest though, which is an incentive to trade in SOFR swaps. Post-2008 discounting (theoretically grounded in Piterbarg's 2012 article) has been driven by the rate at which collateral is remunerated, so I speculate that's the primary drive behind a shift from EFFR to SOFR. Sep 30, 2020 at 16:57
• @DaneelOlivaw: I think you nailed it there with the comment that discounting is driven by collateral remuneration (as indeed per Piterbarg's article). I am not so familiar with what rate USD-denominated collateral is usually remunerated these days: would that be SOFR, rather than EFFR? Sep 30, 2020 at 17:11
• The clearing houses (LCH, Eurex etc) will transition to SOFR in mid October; CSAs will likely follow. Sep 30, 2020 at 18:12

The Fed convened the ARRC (Alternative Reference Rate Committee) in I think 2015 to begin the process of transitioning the financial markets away from Libor. Why? Because Libor had been manipulated , on account of the fact it was based on a poll rather than being sampled from a large , liquid market. The Committee, whose minutes are public, selected SOFR as a replacement rate. Why did they select SoFR instead of EFFR? Because the volume of overnight Treasury repo transactions that are the basis of SOFR far exceeds the volume of overnight Fed Funds transactions, which are overnight loans of funds between banks.

Having selected SOFR as the replacement rate , the Fed wanted to encourage the development of liquidity in longer dated derivative transactions based on the rate. One way of doing this is to change the discounting rate of cleared Swaps to SOFR instead of EFFR. Hence the main exchanges including the CME obliged and the change in discount rate is scheduled for mid October 2020.

• That's a great answer @dm63, exactly what I was after: thank you! PS: if the transaction on SOFR far exceed those on EFFR, does that mean that US banks primarily fund themselves via SOFR? It then begs the question: why wasn't SOFR used for discounting (rather than EFFR) already? Because no OIS swaps existed on SOFR previously? Oct 1, 2020 at 5:45
• @dm63 do you know if regulators are using any legal mean to force this change in discounting at clearing houses, or merely exercising pressure for them to do so? Oct 1, 2020 at 8:09
• @janstuller yes, SOFR didn’t even exist as an index until 2yrs ago so there were no basis swaps on it. Basis on EFFR has been liquid for decades.
– dm63
Oct 1, 2020 at 10:30
• @daneel Olivaw the clearing houses are members of the ARRC so have been involved in the process all along. They were not legally forced to change the discount rate.
– dm63
Oct 1, 2020 at 10:34
• @dm63 Ok thanks. Oct 1, 2020 at 11:09

"US banks fund themselves via EFFR (Effective Federal Funds Rate), as well as the Secured Overnight Financing Rate (SOFR)"

Bank funding is only partly via Fed Funds - there were many important structural changes to that market post-GFC which I believe reduced bank participation in the market. See, for example, this paper: https://www.clevelandfed.org/en/newsroom-and-events/publications/economic-commentary/2017-economic-commentaries/ec-201707-the-federal-funds-market-since-the-financial-crisis.aspx

Perhaps more representative to say banks fund via repo but only a fraction of that is overnight. Funding desks will want to have a significant portion of their funding via term rather than O/N.