As the ECB remains the liquidity provider of last and only resort, we suspect the oh-so-transparent central bank is causing some banks to avoid it and look to the cross-currency basis swap market to fund themselves in USD as the 3 month EUR-USD swap reaches 126bps (-6bps more today). These levels are the lowest (widest and most USD desperate) since December 2008 and perhaps, away from the SMP-driven sovereign spread markets, are the cleanest and least interfered with market view of the extraordinary USD funding crisis that is occurring. These stresses are just as evident in the GC repo markets and Goldman agrees with us that this crisis is escalating and offers a primer on why the GC repo / Libor markets are dysfunctional currently.
And in an effort to comprehend Libor's movements, Goldman Sachs takes a deeper dive into FF/Libor spreads and GC repo markets:
There has understandably been a heightened focus on funding and Libor over the past few months, but it's reached a crescendo in recent days as Europe has seemingly taken a turn for the worse. Globally we've seen FF/Libor begin to steadily move wider as funding pressures become more evident and market participants seek to hedge downside risks.
WHERE SHOULD DOLLAR LIBOR BE?
This is a finger-in-the-air exercise – particularly in an environment like this – but I think the US debt ceiling "scare" in early August provides a useful template / floor. Prior to August, money funds and other short end investors faced a massive collateral shortage and chased front end rates to near-zero levels. But once the debt ceiling debate really came to the forefront and investors sensed the potential for near-term bi-modal event risk, short end rates briefly spiked higher before the situation was eventually resolved.
On Monday July 25th, Investor XYZ was seemingly desperate for yield and comfortable lending cash overnight at ~5bps and in the return accepting collateral in the form of Agency MBS (essentially an overnight claim on the "effectively" government-guaranteed GSEs and itself collateralized with real estate). The following Monday, August 1st, that same repo exchange was trading at 40bps as Investor XYZ decided he would rather park cash in an FDIC-insured bank account earning zero rather than take overnight collateralized ultra-high-grade credit risk. (See the chart below for the levels of overnight GC for UST, Agency Debt, and Agency MBS repo during that period.)
When there's obvious event risk looming, investors – particularly those in the short end – have very little incentive to take this sort of bi-modal "headache" risk. In the example above, Investor XYZ was reluctant to lend cash on an overnight basis in exchange for agency collateral that is "effectively guaranteed" by the US government – 40bps may have been the effective breakeven clearing level. Conservative short end money is rightfully more concerned with the return OF capital rather than the return ON capital. So what does this mean for USD Libor?
3m Libor is supposed to represent some trimmed-mean average of where a panel of 19 global banks (11 of which are European) are able to borrow on an unsecured basis for three months, but in reality, only a handful of banks are actively raising cash for that term (mostly issuance from the likes of Aussie / Canadian / Scandi names who are underrepresented on the panel). Libor has for years been an "interesting" figure (I'm sure to get some strong comments on this), but it's probably especially so at the moment given the obviously large distinction between the "haves" and the "have nots" and the lack of overall data points in money market space.
So if only three months ago, Investor XYZ was afraid to lend ** overnight collateralized ** at 40bps, what's the likelihood that XYZ would be willing to lend ** uncollateralized 3-month term ** for anything even close to 40bps? As 2008 taught many of us, there's a clearing level for everything, but with Libor having set today at ~46.5bps, that 6.5bp extra "compensation" seems especially paltry. Granted, Libor is priced to increased to about 60bps over the next month, but even after reaching that level, will it be fair? A Belgium (AA+) bill auction this morning itself tailed by 60bps... 60bps.
I write the above knowing that many will respond and again remind me that Libor is only a survey-based number that may not always accurately represent universal funding reality; I can appreciate that. But I also write the above to point out that with the current volatility in the world, the "right" level for Libor is likely significantly higher than where it's currently setting and very likely where it's currently set to go. So the current drip-drop increase in Libor can probably go quite a bit further and pick up the pace if one assumes the volatility persists and if one assumes that Libor is very slowly converging to the "right" clearing level. Yes, the banking system as a whole has become better capitalized and has an improved liquidity profile, but in an environment punctuated by an overall questioning of previously risk-free sovereigns, this sort of systemic re-pricing shouldn't be unexpected.
WHERE MUST IT STOP?
Libor should in theory be capped given the outstanding FX swap lines that the Fed has in place with various central banks. Specifically, the Fed offers unlimited USD liquidity (up to 3m term) to foreign central banks at OIS+100 who then pass these funds to foreign banks. Currently, OIS+100bps equates to about 1.08%, but as central banks tend to only lend against haircutted collateral, the effective cost of funds will in practice be even higher. A colleague in Europe – Bernhard Rzymelka – did an analysis in September and found that this equated to a ~35bp charge at the ECB, so all in, the actual cost of funds via the FX swap facility may closer to the 1.40-1.50% vicinity. Hence, that should set the upper bound on libor. (Meanwhile, domestic banks – 3 of them on the Libor panel – are able to borrow collateralized from the Fed's discount window at 75bps.)
WHAT CAN BE DONE TO STOP / SLOW IT?
We see several different possibilities:
1/ Fed cuts the cost of the FX swap line to, say, OIS+50bps. The aforementioned "upper bound" would in theory then be reduced by the amount of the cut. While the effective cost of funding would still be near 1%, it would succeed in essentially cutting off a third of the negative tail of the distribution function (1.00-1.50%) and should do much to contain any further rise in Libor. That said, this option has been on the table for quite some time and the Fed has yet to go down this route, and in an environment where such an action would appear to be a "subsidy" to foreign banks, it would likely encounter some strong domestic resistance.
2/ Europe re-introduces bank debt guarantees. While the effect may be blunted given the obvious stress on sovereigns (the ultimate guarantors), this would on the margin make it easier for banks to obtain short-term funding.
3/ Fed cuts IOER. I'm sure I'll get pushback on this (people love to hate IOER), but if the Fed can push very short rates to zero or even negative levels, investors would likely feel more inclined to entertain the possibility of investing in short-term bank paper.
4/ Fed ends Operation Twist and switches to outright QE. As mentioned earlier, the front end of the curve has suffered from a dearth of collateral for months. By pledging to sell 400bn in <3yr securities, the Fed intends to quench a large portion of that thirst. Unfortunately, the Fed also succeeded in effectively "crowding out" other issuers in the front end. If the Fed isn't raising rates until mid-2013 and I can "risklessly" earn 22bps by buying UST 6/13s, I'm probably going to feel less inclined to buy much-riskier 3m bank paper at the current Libor set (~46.5bps). Thus, if the Fed were to curtail its selling of shorter-dated securities, I suspect short-term Treasury rates would crash lower and bank funding conditions may at the margin become easier.
5/ Europe fixes itself and tensions relax. In this instance, banks may feel less compelled to submit higher Libor sets even though their actual cost of funding may yet still be higher. But I think most investors would consider this outcome unlikely...
(h/t David H)