Crisis Note 2008-2: What's LIBOR Got To Do With It?
LIBOR has been making the news over the past month or so: first it was over concerns it was artificially low due to banks not reporting their true borrowing rates to BBA (a lie, as Bloomberg called it), and more recently, about creating an American version of LIBOR.
So the big question is: why do we care.
The flippant answer would be because numerous financial instruments are now indexed off LIBOR. However, that's an incomplete answer. I'll explain why:
I've been writing about and following the TED spread for my entire career, since 1987. You will find multiple links on this page to my writings on this subject.
This is a link to an excerpt from Research piece I published in 1988. This is very basic and identifies some of the forces that drive LIBOR, mainly demand and supply of Treasuries, and perceived risks of the banking system.
One of the things that people have never fully appreciated is the role of US T-bill SUPPLY in determining the TED spread. During the low volatility years leading up to the LTCM blowup, the TED spread stayed very tight. This was, in my opinion, because of the RTC thrift bailout and resultant T-Bill supply expansion to meet the working capital needs of financing the balance sheets of the bankrupt thrifts.Thus, during this era, no-one really cared about LIBOR or the TED spread (appx 40bps, in from a historical avg of appx 100bps), and asset spreads were generally benchmarked to US Treasuries.
In 1997, Mr. Clinton's Treasury finally started reducing the size of the T-Bill Auctions, a long time after the need for the RTC financing had past (certainly helped with the mid 90s recovery too!). This let to the widening in the Ted spread, and made people aware of the need to hedge with LIBOR. Post LTCM, everyone finally started hedging and benchmarking assets to LIBOR. I warned my clients about this in multiple Bloomberg posts in 1997, one of which is here. I am quite sure that I'm the first person to write about using LIBOR as a benchmark for asset pricing, for both MBS and ABS in 1990. The first paragraph of that article lays out the most important concept, and explains why LIBOR matters: Asset valuations are dependant on the financing costs or hurdle rates of the marginal buyer.
This should remind you of undergraduate microeconomics: the equilibrium price is determined when marginal cost = marginal revenue, with arbitrage opportunities ensuring movements in the demand and supply curves till this is achieved.
During the periods when depositories or other levered institutions were the marginal buyers of assets, asset prices were related to the costs offunds of these buyers - namely LIBOR. Assets with positive LIBOR OASes were cheap (after adjusting for credit), and vice versa.
During other periods, such as when FNMA & FHLMC were the marginal (and largest) buyer of assets, LIBOR OASes for MBS actually went negative, reflecting the sub LIBOR financing that the agencies have. This did not mean that MBS were rich - just that there were more powerful buyers than banks in the market.
This brings us to today's pricing dilemma. Are ABS and MBS rich or cheap?
I will argue that the market has bifurcated. With most large banks and other levered institutions functionally bankrupt, they are no longer the marginal buyer of most assets, nor are the Federal agencies.
There are 2 main sources for funding assets today: 1) The FED and ECB, via the Discount Window, TAF and TSLC, and 2) newly seeded (typically private-equity-funded) hedge funds, thus creating 2 different funding/hurdle rates for assets.
The first, borrowers from the Fed/ECB, are trading/purchasing bonds that can be financed with the Fed/ECB: investment grade debt securities, AAA private label RMBS, CMBS, agency CMOs and 'other ABS'. Their hurdle rate is the TLFC/TAF rate adjusted for the haircut, and also for the asset/liability term mismatch. Lets call this 2% for passthroughs (can be funded at fed funds), 2.50-2.75% for the very short paper such as CMOs, and with rapidly rising term-mismatch premium as WALs increase (explaining why passthroughs are clearing at 5 LOAS, and Agency CMOs are clearing around 50-70 LOAS). Lets call these Category A assets.
A joke I heard recently: what's the difference between an investment bank and a hedge fund? Answer: the investment bank is more leveraged.
The second, newly seeded unleveraged hedge funds, should have hurdle rates in the low teens - the number I've heard most of 13.5%. This explains why downgraded ABS, that are not eligible for future taxpayer ownership (call them Category B assets), trade at high yields, even though they might be very short, and 'money good'.
In between these 2 owner categories lie the original unlevered 'real-money' accounts - money managers and insurance companies. These accounts seem willing to compete for assets with buyers in the second category, and I would argue have deeper pockets than the new hedge funds, (while being on the other hand more cautious having recently had their hands scalded), which has resulted in the recent tightening of front pay downgraded ABS to sub 10% yields, leaving only the much riskier assets for new hedge funds.
What are the implications and risks of this new market structure?
a) Within Category A, I'd stick to Agency bonds, which can get bepurchased by the agency portfolios. In fact, I'd recommend longer duration Agency (7-10yr wal) bonds, as the agencies (FN & FRE) will usually be able to issue debt to buy them, especially if their balance sheet mandates rise (and capital requirements decline) in order to 'bailout the homeowners'. Not that much widening risk.
b) There is tremendous downgrade risk in non agency Category A assets. If an asset drops out of Category A, and cannot be funded by the Fed, it will automatically widen to hedge fund hurdle rates to clear! CMBS, for example, would fall into this category. I would avoid all longer duration non-agency AAAs as a result, unless they had sufficient enhancement to avoid downgrades under drastic scenarios. The risk-reward is just not there, IMHO. In the non-agency space I'd only buy short paper and heavily credit-enhanced bonds as a result. Certainly not newly issued CLOs backed by stale levered loans!
c) The new floor on discount rates is 13-15% (assuming realistically conservative scenarios). Due to competition between the new hedge funds, I don't think discount rates widen any more. Pricing scenarios might change, however, as assets deteriorate further. It might thus be safer to buy riskier Category B assets close to these yields than "low risk" high grade assets.
d) Within Category B, I only like short front pays, and certain types of LCFs and Mezz bonds with low $ prices, where severe cumulative losses have already been priced in.