BLOG

Until two days ago, the critical level for both the Libor-OIS and FRA-OIS spread was the "psychological level" of 50bps. This, however, was breached on Wednesday when as we reported Libor pushed significantly higher without a matching move in swaps. And yet, despite the sharp push wider, both spreads remained below the peak levels observed during the European sovereign debt crisis of 2011/2012, with some speculating that open central bank swap lines at OIS+50bps would limit the move wider.

That changed this morning when the day's 3M USD Libor fixing jumped higher for the 27th consecutive session, rising to 2.2018% from 2.1775%, and the highest since December 2008. And, as has been the case for the past two months, the move was again not matched by OIS, resulting in the Libor-OIS spread jumping to 51.4bp, surpassing the 2011/2012 highs and the widest level since May 2009.


At the same time, the FRA-OIS also spread spiked to a new multi-year high of 53.3bps, the highest in years.



Commenting on the move, NatWest Markets strategist Blake Gwinn urgent clients "don't fade FRA/OIS’ recommendation is still in effect, but certainly on watch", adding that the most frequently asked question this week has been "where will Libor stop?"

While the clear answer - at least for now - is "not here", Gwinn repeated what we said on March 14, noting that the Fed’s central bank swap lines should "theoretically put a cap on USD funding rates" as the banks are authorized to offer terms out to three months at OIS+50bp, and also echoed BofA's comments on the topic, noting that among the impediments are haircuts that add roughly another 10bp to the effective rate, the stigma of going to central banks for funding, and lack of availability of swap lines.

And yet, should Libor keep pushing wider, the Fed will have to notice.

As we said two days ago, the Federal Reserve is increasingly monitoring the rise in LIBOR and is trying to understand what exactly is driving it. In fact, in the most recent dealer survey the Fed specifically asked about the 3m L-OIS spread widening.

As a quick reminder, we previously laid out the 6 possible reasons for the ongoing spread blow out: while this is a simplification of the various catalysts behind the spike in Libor-OIS, here is a quick summary of what is going on - the expansion of Libor-OIS and basis swaps have been impacted by a complex array of factors, which include:

  • an increase in short-term bond (T-bill) issuance
  • rising outflow pressures on dollar deposits in the US owing to rising short-term rates
  • repatriation to cope with US Tax Cuts and Jobs Act (TCJA) and new trade policies, and concerns on dollar liquidity outside the US
  • risk premium for uncertainty of US monetary policy
  • recently elevated credit spreads (CDS) of banks
  • demand for funds in preparation for market stress

Those who say "don't panic" have been focusing entirely on the first three, while traders are increasingly concerned that the answer may be found in the latter, and far more concerning, three explanations.

To be sure, the wider the spread blows out, the more likely it is that something far less benign is causing the move than merely the surge in T-Bill issuance (which the market is well aware will subside soon, and can price it in), or simple supply/demand mechanics for CP/CD.

In fact, should Libor keep rising and LIBOR-OIS blowing out, it won't matter what is causing it as banks will suddenly find themselves in a severe funding shortage, because just like the VIX-market "tail wags dog" relationship - which took the market about 3 years to figure out despite repeated warnings by this site - so the tighter financial conditions become, the wider the spread will move in a similarly reflexive move.

Unless the Fed intervenes of course... which with Libor blowing out so aggressively, is precisely what traders are wondering may happen.

The Fed is certainly aware that the around 35bp increase in 3m LIBOR-OIS since mid-November equates to roughly 1.4 hikes. However, according to BofA, the Fed is probably monitoring LIBOR in the context of broader financial conditions and is not yet sufficiently concerned by the recent tightening to adjust policy. Indeed, the Chicago Fed financial conditions index has tightened, but still shows conditions are much easier vs when the Fed started tightening policy in 2015.

In other words, while the Fed is aware of the blowing out Libor spread, it is unlikely to intervene - i.e. cut rates - unless stocks finally notice and tumble as a result of the sharply tighter monetary conditions on the front end.

* * *

What about the Fed's existing swap lines at OIS+50bps: will they provide a ceiling? According to a recent research report by Citi, the answer is no.  As Citi rates analyst Ruslan Bikbov writes, "we have received a number of questions on whether central banks currency swap lines may be tapped and form an effective ceiling for LIBOR/OIS at 50bp."

First, some background: in October 2013 the Fed converted temporary liquidity lines to standing arrangements with the following five central banks: BOC, BOE, ECB, SNB, and BOJ. A dollar FX swap between the Fed and a foreign central bank involves two transactions. At the start of the swap, the foreign central bank sells local currency to the  Fed in exchange for USD at the market exchange rate. The dollar amount is held at the foreign central bank’s account at NY Fed, while the foreign currency is held at the Fed’s account at the foreign central bank. The foreign central bank is bound to buy the local currency back at the same exchange rate and to pay interest to the Fed at the termination of the swap.

Since December 2011 - the peak of the European sovereign debt crisis - the required interest was reduced from OIS+100bp to OIS+50bp. There is no limit on the size of swaps. Foreign central banks offer swapped USD to corresponding domestic banks at pre-announced fixed rate tenders with full allotment, most often with the term of 7 days. These loans are secured by accepted collateral denominated in local currencies.

As such, the OIS+50bp funding offered by swap lines suggests LIBOR/OIS may see a ceiling at 50bp. The fact that 3m LIBOR/OIS didn’t go above 50bp in late 2011-earlier 2012 may support this idea. However, according to Citi, this simplistic argument misses the following considerations:

  • The term of the loan. The terms of the loans are ECB/BoJ’s discretion. As discussed, a typical term of the loan is 7 days. While central banks offered longer-term (3-month) before, this was done during crisis times, such as in 2011, when solvency of the banking system was in question. ECB and BoJ also provided longer funding (2-3 weeks) over year-ends to reduce dollar funding pressures of their banks. Given ECB/BoJ’s inaction during the US MMF reform, we do not expect them to extend their tenor beyond 1-week. Even if they decide to act, we still expect OIS+50bp to work as a “soft” ceiling than the “hard” ceiling, given the factors discussed below.
  • Stigma. Reaching to FX swap lines may be associated with non-trivial reputation risks. We hear anecdotal reports of foreign central banks dissuading their banks to use the facility as it may signal systemic risks
  • Collateral. While Libor represents the cost of an uncollateralized CP/CD funding, FX swap lines are secured by accepted collateral denominated in local currencies with rather restrictive haircuts. For example, the BOJ requires a  13% haircut for yen-denominated collateral, while the ECB requires a 12% haircut in addition to any usual haircuts applied to ECB loans.

Furthermore, the fact that both FRA and L-OIS are now above 50bps confirms that these lines are indeed not being used (unless they are, in which case we will get a confirmation when the Fed and other central banks report their weekly usage on swap lines next).

But don't hold your breath: the restrictive collateral and stigma explain why the usage of dollar swap lines remains minimal despite the fact that 3m FX OIS for EUR, JPY and CHF already are trading at levels below -50bp for 3M.

Meanwhile, over the last few years, even 1w FX OIS has been exceeding -50bp at month ends, implying only a “soft” ceiling, despite the availability of weekly liquidity offerings. Here Citi notes that FX OIS were trading at much wider levels back in 2016, yet the usage of FX swap lines also remained minimal at that time.

Citi's conclusion: "do not expect much usage of FX swap lines in this environment either, although we will be watching out for ECB/BoJ’s longer-dated swap lines."

Which brings us back to the underlying crisis. As we discussed first last October, the main reason for the recent widening of FRA/OIS is uncertainty about corporate cash holdings following deemed reparation of offshore cash. As a result of this uncertainty, the demand for CP/CDs with maturities longer than 1m-3m from offshore funds and corporate Treasuries has evaporated, which is evident from the contraction of average maturity of outstanding CPs.

As such, FRA/OIS levels may continue to blow out until this persistent uncertainty is resolved, which is unlikely to happen before corporate earnings announcements in Q2.

The problem is what happens should Libor (and associated spreads) blow out so much in the next month until Q1 earnings season begins, that corporations and banks finally throw in the towel as a result of the collapse in funding, and demand a central bank bailout, one which will only be greenlit in the old-fashioned way: with a market crash.