Interest rate swaps have become ubiquitous in almost every corner of credit markets. In their purest form, they serve to hedge borrower costs against rising interest rates. An obligor often seeks to reduce debt service by issuing debt at the lowest possible rate, and that usually means (like banks) some form of short-term, floating rate security. Given the fixed nature of obligor cash flows and assets, that doesn’t fit the typical risk profile. Enter the swap.
John’s Hopkins Hospital (JHH) is one of the best-endowed health service systems in the country, yet it often relies on debt financing for capital costs and budget shortfalls/enhancements. As of March 31, 2013, JHH had a separately incorporated endowment trust with $593 million in net assets, and cash on hand in the hospital system of $2.23 billion. Yet, there is about $1.7 billion in outstanding debt (as of Mar 31), with 45% issued at floating rates of all types and flavors (including auction rate).
The hospital first entangled with rate swaps in 2004, initiating a fixed payor swap with JP Morgan. JHH is paying JPM 3.329% in exchange for 67% of 1-month LIBOR, with a maturity of July 2023. In April 2006, JHH followed that with two fixed payor swaps with Goldman Sachs, structured in much the same way. JHH is paying 3.911% and receiving 67% of 1-month LIBOR on the first, and paying 3.922% to receive 67% of 1-month LIBOR again on the second.
The notional amounts of the three swaps were together about $350 million. JHH added a couple more, with impeccable timing, in July 2007. As of December 31, 2012, JHH was receiving 0.16% as 67% of 1-month LIBOR while paying north of 3.3% on all of them. From JHH’s perspective, even though they appear to be on the losing end of the trades, it is nothing more than the cost of “insuring” or hedging their floating rate debt.
However, there is a complication here. JHH had to post cash collateral at the inception of the cash payment streams, totaling about $20 million. The collateral posting is marked daily, and cash must be deposited according to the value of the swap contracts. When interest rates started to fall in 2007, further in 2009 thanks to ZIRP, JHH was hit with collateral calls on its cash. By the end of 2011, JHH’s total cash collateral postings were $54.4 million, on a -$99 million “fair value” for the swap contracts.
The cash continues to be owned by JHH including the ability to earn interest, but it remains in possession of the swap counterparty bank (in this case both JPM and Goldman). In terms of hospital system operations, the cash is fully restricted and unavailable for anything other than collateral.
In 2012, the swap contracts were a particular problem as interest rates fell further, in full part as QE 3 moved from rumor to reality. The net result for JHH: come up with another $86.1 million in cash to place in the restricted collateral accounts! By the end of 2012, JHH had cash collateral posted of $140.5 million on a -$206.7 million fair value.
Across town, the University of Maryland Medical System (UMMS) has been in the same position. Despite a much smaller asset base, and swapping a separately funded endowment for state aid, the corporation has also been interest rate swap trading. At the end of 2012, UMMS had a notional of $610.1 million in swaps with JPM and Bank of America (likely Merrill Lynch). At one point last year, UMMS had cash collateral posting (restricted against their use) of nearly $200 million.
The hospital system had $732.4 million in cash and unrestricted short-term investments at the end of the year, and $102 million in collateral postings outstanding (which they borrowed from an existing credit line to fund). Like JHH, UMMS has about 43% of its $1.4 billion in total debt in floaters.
The good news for UMMS, despite a lot of recent criticism about the swaps, is that collateral requirements have moved in the hospital’s favor. From the end of 2012 through the end of Q1 2013, swap rates have, as we well know, moved upward changing the fair value calculations on the contracts, and thus the collateral requirements. As of March 31, collateral postings from UMMS have declined to about $93 million. Over the same period at JHH, collateral balances have fallen from that $140 million all the way to $103 million.
And that was just at the end of March – we will have to wait to see how far their collateral lockup has fallen now that interest rates have sold off so dramatically in the interim. The “fair value” of the swap contracts has no doubt moved very much in both UMMS and JHH’s collective favor, freeing up cash collateral that actually flows through their income statements as positive “income”.
The flipside to that process is what I am most interested in, namely the big 4 derivative banks. The four largest derivative dealers in the US, JPM, Bank of America Merrill Lynch, Citibank and Goldman Sachs, account for 93.9% of total notional swaps (Q1 2013), amounting to $129.97 trillion out of $138.36 trillion total. As in the cases of UMMS and JHH swaps, the collateral process which had once been so very favorable has now moved against them.
For much of the past five years, particularly 2012, the banks have been net receivers of collateral, particularly cash, which they can use in their own business operations due to accounting for dealer balance sheets. Any decline in cash postings to those banks places collateral pressures on other operations and trades, meaning the banks themselves start getting collateral squeezes.
I believe this explains a good deal of the collateral issues we have seen this year. In many ways this was like AIG’s credit default swap business pre-crisis (AIG thought it would be “take the premiums and never have to pay out”), where, given expectations for unlimited QE, the derivative dealers were very much willing to receive fixed on rate swaps because they thought it was both “easy money” and, as I have shown above, a steady source of positive collateral into their systems. Now they are entangled on the wrong side of the trades given taper, including strains on collateral.
When looking at the derivatives “market” (can it really be a market with only 4 real dealers?), more than half of the total notional amount of contracts outstanding are to each other.
The reason for that comes from the way in which dealers manage total book risk. It would be easy for a dealer to enter a swap agreement with JHH, for example, and then lay off all the risk by entering the opposite position of equal terms with another counterparty. The derivative book would be perfectly balanced, save for some small spread created from the bid or ask to midpoint as a dealer “fee”. But that is not how books are run, particularly as these derivative positions have magnified and grown.
The dealers instead total up their expected exposures from each derivative trade to estimate total exposure across all kinds of risks. They will then hedge those risks in the aggregate, making it more economical from cost and management standpoints. These hedge positions change constantly with estimations of interest rate direction (delta), intensity (gamma), and, most of all, implied volatility across all book positions (called vega).
What you end up with is what the chart above shows in the aggregate. JPM does a rate swap with JHH and hundreds of other counterparties to create exposure, and then swaps or hedges with Goldman to reduce or hedge that exposure. Except Goldman has already swapped with UMMS and even more counterparties, and needs Merrill Lynch to offload some exposure; and then Merrill reduces that risk by swapping back with Goldman and then laying off some back at JPM. Together they all hedge other modeled exposures with Citigroup, who has its own layer of exposure from the contracts it is taking on. At the end of the day, all these exposures and hedges are “netted” to some modeled specificity on those risk parameters (particularly vega).
The netting results in mathematical representations of risk that are seemingly defused in the aggregate, but only if you ignore that most of that exposure is simply tied in a Gordian Knot of cross exposures among only four domestic banks (and their megabank counterparties in Europe, Deutsche Bank in particular). What happens, then, when these exposures turn from a source of collateral into a need for it?
The total amount of collateral, at least as defined as a percentage of net current credit exposure (an estimated measure of a bank’s potential liability; it’s not perfect, but it gives us an anchor to at least measure relative collateral changes), has been rising steadily for the past year and a half. But that is somewhat misleading since it does not account for types of counterparties (remember that the corporate sector was responsible for most of that in 2012, being on the “wrong” end of the swaps).
To get a better sense of this collateral inflection, we need only look at collateral from the banks/dealers themselves.
We have turned back toward the high level of collateral engagement/needs from 2009. While the Office of the Comptroller of the Currency (OCC) does not give specific information on collateral by type of contract, we can reasonably assume (by total notionals in swaps, and the behavior of swap rates over that period) that it is almost exclusively interest rate swaps driving the marginal changes shown above.
This gives us a sense of illiquidity and collateral pressures that are not readily apparent in these opaque markets. There were $296.8 trillion in notional derivative contracts outstanding as of March 31. Out of that total, 94.6% were OTC.
I’ll have more to say about collateral and derivatives in an upcoming post, but these are pressures in the system that nobody seems to account for until it rips open markets. Taper is a dangerous game; in terms of systemic processes it is not unlike the effect the “unexpected” inflection in home prices had on mortgage structures beginning in late 2006.
Click here to sign up for our free weekly e-newsletter.
“Wealth preservation and accumulation through thoughtful investing.”
For information on Alhambra Investment Partners’ money management services and global portfolio approach to capital preservation, contact us at: jhudak@4kb.d43.myftpupload.com
Stay In Touch