Law360 (March 27, 2020, 5:45 PM EDT) --
Libor currently administered by the Intercontinental Exchange Inc. is scheduled to end by 2021, mainly in response to the demand for a better benchmark that is more robust and less subject to potential manipulation. Some jurisdictions like the U.K. are even attempting to accelerate the transition and have advocated that market participants switch from Libor to a replacement benchmark as early as this month.
In light of the extreme financial market turbulence brought about by the novel coronavirus that causes COVID-19, this article focuses on the recent behavior of the new Secured Overnight Financing Rate, or SOFR, benchmark replacement for U.S. dollar Libor.
In the past few weeks, SOFR has moved sharply in the opposite direction of USD Libor, an aberration that does not appear to be attributed to random noise or drift, but rather to deeper fundamental differences in the way each benchmark tracks financial risk.
The significant divergence in the two benchmarks comes when market participants have grown to rely on and most need a reliable replacement to Libor — during a time of extreme financial market stress, as is now being experienced.
SOFR as a Replacement to Libor
In the U.S., the replacement benchmark to USD Libor is based on SOFR. As intended, SOFR offers several advantages over USD Libor, which include being based on actual transactions in the wholesale interbank funding market and being drawn from a wide range of market participants that includes dealer-to-dealer trades and dealer-to-customer trades.
These features broaden the appeal of SOFR and satisfy the architects' objective of devising a benchmark that is less subjective and less prone to manipulation. A brief background about how SOFR is constructed will illuminate the way it is performing in the current crisis period.
SOFR is compiled daily from rates on hundreds of billions of dollars in trades occurring in the U.S. overnight repurchase agreement, or repo, market. The type of repos that the U.S. Federal Reserve tracks to derive daily SOFR are overnight and collateralized by U.S. Treasury securities.
This is an important feature of SOFR: it is an overnight rate backed by the most widely traded and trusted debt instrument in the world. In times of economic turmoil, investors flee to the safety of U.S. Treasury securities. Indeed, since the start of the coronavirus crisis, global investors have revealed their strong preference for the safety of U.S. Treasury securities in a flight to quality.
The way this has impacted the behavior of SOFR and Libor differentially during a crisis period is economically significant, as discussed below.
In contrast to SOFR, the most prevalent form of Libor used by investors is term Libor, meaning it is a rate associated with a loan expected to be repaid in the future, such as three or six months out. A second important distinction is that Libor represents an unsecured borrowing rate; that is, a rate not predicated on backing the loan by collateral.
As such, one risk element embodied in LIBOR rates is the risk of the financial health of the banks that must repay their interbank, unsecured loans. In short, Libor communicates information about banking system risk that financial market participants have come to know and rely on.
The Relationship Between SOFR and Libor Benchmarks in Response to the COVID-19 Crisis
The chart below plots the rates over time on USD 3-month Libor and the corresponding 3-month average SOFR, or 3-month SOFR, from the first available date of 3-month SOFR data through March 20. Three-month SOFR is an average of the previous 90 days of daily SOFR rates, and is updated and reported daily by the U.S. Federal Reserve as the intended benchmark replacement to USD 3-month Libor.
The Calm Before the Storm: The Two Benchmarks During Relative Economic Stability
During the first year of the launch of 3-month SOFR, USD 3-month Libor rates were always higher. The chart shows a positive spread through the first half of 2019. A positive spread is not surprising, given investors require a default risk premium in Libor, an unsecured rate, compared to repos backed by high-quality collateral.
An interesting difference in the benchmarks is that around late 2018, USD 3-month Libor peaked, while 3-month SOFR continued trending upward. Three-month SOFR eventually reversed course and followed LIBOR downward during 2019, albeit gradually and with a lag. This is mainly due to the ninety-day averaging effect embedded in 3-month SOFR.
By late 2019, 3-month SOFR caught up with and fell below Libor, resuming the expected spread differential and continuing a parallel downward trend into early 2020. Thus, from the middle of 2018 through early 2020, both benchmarks generally followed the same inverted U-shape trend.
Onset of the Novel Coronavirus Upends the Benchmarks
The two benchmarks have moved in surprisingly different directions since the onset of the novel coronavirus. The benchmarks' substantially divergent paths in such a very short period can be expected to have profoundly material impacts on the portfolios and decisions of market participants who were unprepared for a storm of this magnitude.
The shaded region in the right side of the chart shows the patterns in USD 3-month Libor and 3-month SOFR since February, around the time the COVID-19 virus became a public emergency in the U.S. The 3-month SOFR briefly flattened, then turned sharply downward in March.
By contrast, while USD 3-month Libor initially fell substantially and quickly moved below 3-month SOFR, it then abruptly reversed course and began a sharp climb. Every day since March 12, the two benchmarks have moved in opposite directions. The increasing USD 3-month Libor is likely responding to some degree to perceived elevated banking system risk.
On the other hand, 3-month SOFR appears to be responding to a combination of U.S. Federal Reserve liquidity intervention and the flight to safe-haven U.S. Treasury securities.
The pattern of divergence in terms of the magnitudes of daily rate movements is also quite significant. From March 12 through March 20, USD 3-month Libor rose a total of 46 basis points — nearly one-half of a percent. Over that same time, 3-month SOFR declined by roughly 7.7 basis points.
Moreover, the underlying SOFR rate, which is measured daily, declined over this eight-day period by over 100 basis points.
In part because of its construction, the new 3-month SOFR benchmark never had a chance to closely follow the fast-moving pace of 3-month Libor, as the financial system veered course on a daily basis. Finding an orderly path to transition from current Libor to the new SOFR benchmark is likely to be a far more challenging task, having now experienced this extremely turbulent episode.
Notes: The correlations shown are based on USD 3-month Libor and 3-month SOFR.
Last, if one were to consider the pre-novel coronavirus period as "normal" and the onset of the novel coronavirus as "highly uncertain," the statistical correlation between the two benchmarks can be calculated.
Using as the onset of the crisis period the date on which the World Health Organization declared the novel coronavirus a pandemic, March 11,  and the time period prior to that date as the normal period, the chart above depicts how different these time frames are in terms of the relationship of 3-month SOFR to USD 3-month Libor.
The correlation during normalcy is positive 0.6 and highly significant, while the correlation during the highly uncertain period is negative 0.97 and highly significant.
SOFR has been referred to as the risk-free rate replacement to USD Libor, a characterization that the architects appear to use to suggest the superiority of SOFR over Libor. There are many valid reasons for the advantages of SOFR, including: being rooted in hundreds of billions of dollars of actual interbank borrowing transactions; representing a wide cross-section of the interbank funding and money markets; and the transparency in its definition, construction, and reporting.
SOFR is positioned to become the next short-term interest rate benchmark referenced by financial contracts in lieu of USD Libor, and organizations like ISDA, banks, clearing houses, end users and regulatory bodies have worked together to expedite the smooth transition.
But the dilemmas that the new SOFR-based benchmarks potentially pose have also not gone unnoticed. The general counsel of the Federal Reserve Bank of New York commented that without the proper preparations to facilitate an orderly and planned changeover in 2021, the situation invites potential litigation on a "massive scale."
As the global financial market's response to the novel coronavirus outbreak demonstrates, significant shortcomings in the transition may be a very likely scenario.
Jeffrey S. Armstrong, Ph.D., is a director at Berkeley Research Group LLC.
The opinions expressed are those of the author(s) and do not necessarily reflect the views of the firm, its clients, or Portfolio Media Inc., or any of its or their respective affiliates. This article is for general information purposes and is not intended to be and should not be taken as legal advice.
 See Bank of England, "FCA and Bank of England encourage switch from Libor to SONIA for sterling interest rate swaps from Spring 2020" (Jan. 16, 2020), available at: https://www.bankofengland.co.uk/news/2020/january/fca-and-boe-joint-statement-on-sonia-sterling-interest-rate-swaps-in-q1-2020.
 A repo is an arrangement where an entity in possession of cash seeks to lend to a borrower by doing the following: the borrower sells a security to the lender in return for cash, and then returns the borrowed cash (plus interest) by repurchasing the security from the lender on the maturity date. While the maturities of repos can vary, a large proportion of repos are for overnight loans. Repos are thus a form of collateralized short-term borrowing. The repo transactions that underlie the daily SOFR regularly exceed $800 billion. See Alternative Reference Rates Committee, "Transition from Libor," available at: https://www.newyorkfed.org/arrc/sofr-transition.
 The 3-month average SOFR shown in the chart is the "90-day Average SOFR" reported by the Federal Reserve Bank of New York. Data sources for historical Libor and average SOFR are the U.S. Federal Reserve Banks of St. Louis and New York, and Bloomberg. See Federal Reserve Bank of New York, "Secured Overnight Financing Rate Data," available at: https://apps.newyorkfed.org/markets/autorates/sofr; and Federal Reserve Bank of St. Louis, "3-Month London Interbank Offered Rate (Libor), based on U.S. Dollar," available at: https://fred.stlouisfed.org/series/USD3MTD156N.
 See Federal Reserve Bank of New York, "Statement Regarding Publication of SOFR Averages and a SOFR Index" (Feb. 12, 2020), available at: https://www.newyorkfed.org/markets/opolicy/operating_policy_200212.
 The overnight SOFR rate has fallen by roughly 114 basis points since March 12, 2020. See Federal Reserve Bank of New York, "Secured Overnight Financing Rate Data," available at: https://apps.newyorkfed.org/markets/autorates/SOFR.
 See World Health Organization, "Rolling updates on coronavirus disease (COVID-19)" (update March 23, 2020), available at: https://www.who.int/emergencies/diseases/novel-coronavirus-2019/events-as-they-happen.
 The t-statistics are 17.1 and −10.1, respectively. As the trend chart above shows, there are other subperiods where the two benchmarks have qualitatively different correlations; for example, from late 2018 to mid-2019, the correlation is negative. But as discussed already, this appears to be an artifact of the averaging effect in 3-month SOFR.
 See Alternative Reference Rates Committee, "Frequently Asked Questions" (Sept. 19, 2019), available at: https://www.newyorkfed.org/medialibrary/Microsites/arrc/files/ARRC-faq.pdf.
 Op cit.
 See speech by Michael Held, Federal Reserve Bank of New York, "SOFR and the Transition from Libor" (Feb. 26, 2019), available at: https://www.newyorkfed.org/newsevents/speeches/2019/hel190226.
For a reprint of this article, please contact email@example.com.