Lots of confused chatter in the bond community as to why the negative swap spread story (anywhere between 7Y and 30Y) is being largely ignored by the media. After all, the associated market, which according to the BIS was roughly $154 TRillion in June 30 makes the Greek bond debacle and various sovereign CDS discussions in the media pale in comparison. As several bond traders pointed out, the likelihood of negative spreads having been modelled out by the TBTFs is very low, if any, meaning that unhedged bank IR-swap exposure is suffering massively, and is likely to surpass all record past prop desk losses. In fact, rumors abound that a few of the desks having placed leveraged bets on spread divergence over the past months and years are currently in critical condition, yet nobody is discussing this for fear of another round of bank run concerns among the TBTF banks. What is odd, is that the Primary Credit borrowings are now at almost financial crisis lows of just under $9 billion, leading many to speculate that banks now satisfy all of their short-term funding needs via the fungibility of excess reserves (and indicating once again that the Fed's discount rate hike was the most irrelevant action in a history of irrelevant actions). And just in case there is still confusion as to what negative swap spreads mean, here is a useful primer.
From Morgan Stanley:
The big news is that 10y swap spreads (Swap Spread = Libor Rate - UST Yield; e.g. 10yr swap spread = 10yr Libor Rate - UST 10yr yield. This spread has always been positive, today it is negative implying that UST 10y yields have risen above 10y Libor rates) have fallen below zero (Exhibit 1). But the bigger questions are how do we define value in spreads and how much more inverted can 10y spreads go? At the height of the crisis in 4Q08, 30y swap spreads got to -41bps and have remained inverted ever since. The inversion of 30y spreads had more to do with technical, exotic and hedge-related factors. But those elements are missing from the inversion of 10y spreads. That's what makes it interesting. Today we view the inversion in 10y swap spreads as a harbinger of the massive supply of UST debt that will ultimately drive yields higher.
What drives swap spreads?
Back in the late 1980's and early 1990's, swaps were a ground-breaking innovation. It was a “magic formula” to turn long-term liabilities into short-term liabilities. ABC Company would issue (that is, pay an interest rate) on, say, 10yr debt in the market, then receive a 10yr fixed interest rate in the Libor market while simultaneously paying a 3-month floating Libor rate (Exhibit 2). And voila, ABC Company 'swapped' a long-term fixed liability for a cheaper and easier to manage short-term liability. Magic! But things were simpler in days of yore. Swaps were a less volatile vehicle used by corporations to manage cash flows on their debt. They used swaps because they could customize the end dates of their fixed cash flows, rather than relying on US Treasuries with their pesky issuance cycles and inherent technicalities. For example, 10y spreads from 1992–1997 trade in a range from 28bps to 41bps, pretty narrow. But this simplicity allowed us to understand what drove swap spreads. These factors were primarily:
- The spread between Libor and repo rates
- The slope of the yield curve
- The US deficit and UST supply
The first point argued that Libor rates should be higher than UST rates because there was a 'repo-specialness' premium between UST’s and Libor. The second point refers to corporate issuance: if the curve steepens, then corporations are more likely to swap their long-term liabilities at higher rate levels into short-term liabilities at lower rate levels. The last point refers to the relative level of UST issuance. If the US economy slows and goes into a deficit, then the US will issue more Treasuries to fund itself, therefore narrowing swap spreads (Exhibit 3). Currently, we have a high 9.9% deficit to GDP ratio and correspondingly, a $2.4Tr gross issuance of UST’s in 2010.
What has changed?
Later in the 1990s and over the last decade, the swaps market took on increased importance not just as a hedging vehicle but also as a speculative vehicle. What drove swaps over the past 10 years has been hedging the interest rate sensitivity in mortgages. But today, a case can be made that mortgages are less interest rate sensitive than in the past (i.e., less negatively convex). The reason is that households may be less able to refinance their mortgages for a given change in interest rates, because refinancing is more related to FICO scores, credit availability, loan-to-value, incomes, etc. We believe swap spreads will narrow and remain inverted as interest rates have stayed low and stable, volatility has fallen, spread products have been narrowing, and there is little hedge-related need to pay fixed in swap. Add to that the old-time reasons, which are becoming more popular drivers of swap spreads today, of reduced specialness premiums, tighter repo-Libor spreads, steeper yield curves and monstrous US Treasury supply. All of which become the contributors to 10y swap spreads moving negative. Oddly, the tight level of swap spreads is a look back into the future. But the inversion of spreads is the new twist.
Morgan Stanley's conclusion is that, independently of our concerns, US Treasury rates are about to spike. To be sure, MS has been pushing for high rates and major curve steepending for a while: recall it is their call that the 10 Year will hit 5.5% this year.
The issuance of UST debt is dwarfing Libor-related issuance. For example, we expect UST net issuance to be $1.7Tr and net issuance of MBS to be zero. Thus, the relative issuance of UST’s vs. Libor-based products mainly accounts for the inversion in swap spreads. This is a first sign of stress leading to higher UST yields and is not to be missed.
And back to our question: is there currently a massive P&L hit to some or all of the Big 5 US banks as a result of this very much unexpected inversion? While surely the full $154 trillion or so amount is not applicable to the 7Y+ inversion, the OCC as of its most recent report does indicate that there is $27 Trillion in Interest Rate swaps outsanding with a maturity greater than 5 years.
And when looking at IR holdings by bank, it would seem that JPM, Goldman, Bofa, and Citi are most impacted. While JPM, GS, BofA, Citi and Wells have about $131 billion in IR swaps among them, more relevantly, JPM, Goldman and BofA have $9, $7 and $5 trillion in >5 year IR swaps. This is very troubling.
Maybe some of those fantastic financial analysts who were telling the general public to buy Lehman a few days before its bankruptcy, and are now saying financial companies will quadruple over the next few years, can do something useful for a change and ask the executive teams of the above mentioned banks 1) how big their exposure to negative swap spreads is and 2) what the negative P&L impact as a result of this unprecedented spread inversion is?