By Brian P. Lancaster and Mark Feldman, managing directors, Bear, Stearns & Co.
The linkage between the commercial mortgage-backed securities market and interest rate swaps has grown dramatically over the past few years. Yet based on the number of calls and questions we have received, it is apparent that for many, swaps remain a mystery.
Therefore, this report is devoted to clarifying the relationship between the CMBS and swap markets. The first section covers the basics of the swap market. We then discuss the impact of the swaps market on the CMBS market. Finally we provide a quantitative analysis which shows that the 10-year swap is the best predictor of 10-year triple-A CMBS spreads vs. Treasurys since January 1998. Based on this model, 10-year triple-A CMBS as of Aug. 13 was six basis point cheap versus the 10-year Treasury given the 10-year swap spread on that day.
What are Interest Rate Swaps?
A standard or vanilla interest rate swap is merely a contractual agreement between two parties to exchange two sets of cash flows usually based on a fixed rate the other based on a floating rate (usually LIBOR). The cash flows are typically interest-only based upon a notional amount of principal from which the payments are calculated.
To illustrate, counterparty A agrees to pay counterparty B a fixed rate cash flow stream of the 10 year Treasury yield plus 106.5 basis points (the spread one typically hears quoted when asked "where are swaps?") based on $1 million of notional principal in exchange for receiving three-month LIBOR flat. So with a 10-year Treasury yield of 6.02% and a three-month LIBOR rate of 5.00%, counterparty A would have an annual liability of $70,850 or $35,425 semi-annually. In exchange, counterparty B would have a quarterly liability of $50,000/4 = $12,500.
Why Do Swap Spreads Influence CMBS Spreads?
The primary reason swap spreads influence CMBS spreads is that the movement in swap spreads is viewed as a benchmark for the movement of double-A/single-A "credit spreads" in general. The average credit rating of the counterparties exchanging the cash flows is about AA/A. Since a number of investors cross between CMBS and corporates, 10-year triple-A CMBS spreads tend to move with 10-year single-A spreads.
Thus as swap spreads widen and tighten versus Treasurys, so too do 10-year triple-A CMBS spreads. As supply, credit or liquidity events push out these corporate bond spreads; swap and CMBS spreads widen as well. In addition, the use of swaps to hedge CMBS positions has grown increasingly prevalent which has strengthened the relationship. The link between swap spreads and the corporate bond market is one reason why during the fall of 98 as the bond markets "panicked" about a possible global financial meltdown and liquidity dropped, swap spreads widened to such historically wide levels.
Ironically, during this period, 10-year triple-A CMBS spreads decoupled from the swap market and became incredibly cheap versus swaps. This was due to technical factors specific to the CMBS market including the forced liquidation of CMBS holdings due to the large dominance of leveraged hedge funds in the CMBS market. In addition, a number of Street firm's CMBS traders were constrained from making liquid markets in CMBS, since their conduits already had large positions of commercial mortgages waiting to be securitized.
The Relationship Between Swap Spreads and CMBS: The Proof
There is compelling evidence that since January 1998, 10-year swap spreads have been the dominant factor in determining the spread between 10-year triple-A CMBS and the 10-year Treasury. While over time the exact values of the regression multipliers will somewhat vary and significant movements of other variables may have an impact, swap spreads will continue to play a central role in explaining the 10-year triple-A CMBS spreads. Using weekly data with the 10-year swap spread as the only explanatory variable, our estimated regression equations are:
CMBS Spread (t) = 11.74 + 1.356*SW10(t) + e(t),
and e(t)=0.645 * e(t-1) + u(t),
where the week is indexed by t, SW10 is the 10-year swap spread and e(t) is the residual. According to this equation, if the 10-year swap spread widens 10 basis points, the basis is predicted to widen on average, 13.56 basis points. For example, if on Aug. 13, the 10-year swap spread is 105 basis points, then the model predicts that 10-year triple-A CMBS on Aug. 13, should be 154.15 basis points.
If the actual 10 year triple-A CMBS spread is 160 basis points, then the model says that 10 year triple-A CMBS are 5.85 basis points cheap. In terms of timing, the model is also saying that if the swap spread remains unchanged at 105 basis points, then the 10 year, about 35% of this cheapness should "go away" in one week (i.e., in one week, 10 year triple-A CMBS should tighten about 2basis points. (35% X 5.85 basis points).
The Importance of Swaps to the CMBS Market
Because swap spreads and CMBS spreads are highly correlated, CMBS market participants (conduits, investors, traders etc.) can use them to hedge positions and determine relative value. If CMBS and swap spreads widen, the CMBS investor loses money on CMBS and makes money on the swap and vice versa. If done with the correct ratio, one can substantially mitigate losses due to spread volatility. For example, based on Bear Stearns CMBS-Swaps model, we expect that for every 1 basis-point of widening in the 10 year swap spread, 10-year triple-A CMBS spreads would widen 1.35 basis points.
If the fixed rate spread received from the CMBS is greater than the fixed rate spread paid on the swap (which is currently the case) then the participant has "net net" locked in a profitable spread over Libor while hedging himself against spread widening. Of course, if the spread between CMBS and swaps is not stable then this can reduce the profitability of this position.
For example, if we create a synthetic floater at LIBOR + 55 basis points by purchasing a 10-year triple-A CMBS and receiving LIBOR in a swap, as long as the position is maintained, it pays Libor + 55 basis points.
If the spread between 10- year triple-A CMBS and the 10-year interest rate swap widened to say 70 basis points, our position would lose money upon liquidation but gain if the spread tightened.
The above portfolio is hedged against changes in the overall level of interest rates. That's because the participant has through the above transaction in effect created a synthetic floater. Since uncapped floating rate instruments tend to have negligible durations, the value of the transaction will be less sensitive to changes in interest rates.
Because swaps can help hedge against spread widening as well as rising rates they have become a valuable tool for CMBS conduits to remain profitable or at least to minimize losses under current volatile market conditions. Indeed for those conduits that currently use them effectively they can be an effective competitive weapon as market volatility shakes out those less sophisticated players who do not.
Recent Market Conditions
While factors such as Treasury financing rates, credit spreads, and liquidity play a major role in determining swap spreads, during the last few months "rising interest rates" have also been a factor. Since swaps can change the duration of one's assets and liabilities, the market's overall "demand for duration" is increasingly being reflected in swap spreads.
Given the bearish tone of the market over the last few months and the market's expectation that the Federal Reserve might raise rates, more market participants want to pay fixed and receive floating which is contributing to wider swap spreads. Because CMBS and swaps are linked, swap spreads have taken CMBS spreads with them even though little has changed in the real estate markets.
The situation of too many market participants wanting to pay fixed has been exacerbated by two factors in the last few months. First, the "Street" including corporate bond traders, residential mortgage traders, CMBS and ABS traders as well as CMBS conduits not to mention many investors, have learned the lessons of last fall and the strong relationship between swaps and bond spreads. As a result they are to a greater extent than ever before using swaps (i.e. paying fixed) to neutralize the price volatility of their inventories.
Second, many corporate Treasurers, concerned about a potential rate hike and the possibility that Y2K issues could disrupt capital markets in the fourth quarter, have pushed their corporate, ABS and CMBS issuance forward to the present. This has increased the need of the Street to pay fixed to hedge these under-writings, while at the same time forcing out spreads on double-A/single-A credits which also pushes out swap spreads.
The growth and use of swaps over the last year has been a positive to the CMBS sector in that it has given the market away to partially hedge against spread volatility. This gives broker/dealers better ability to provide liquidity in difficult market environments.
However, this increased usage may have the negative side effect of partially causing increased short term swap spreads (and therefore CMBS spread) volatility.
Increased usage may have also created problems for the institutions and dealers on the other sides of these transactions. For example, in just one day recently, ten-year swap spreads blew out 8 bps and swaption volatility soared 1.7% reportedly due to "trouble" at one or more major players in the swaps markets.
It is readily apparent that swaps will continue to play an important role in both hedging and determining relative value in the CMBS market. For this we have developed a valuation model which shows that 10-year triple-A CMBS are currently 6 basis points cheap to 10-year swaps.
However they have also linked the CMBS market more tightly than ever before to the capital markets and pulled it away from what is happening in the real estate markets. We would agree that current wider triple-A CMBS spreads are in part reflective of anticipated Fed tightening and the concomitant potential rise in commercial real estate risk.
But, it is by no means apparent that all of the current widening is appropriate for expected increased risk levels. Indeed we suspect that a considerable portion is due merely to short term technicals not only in CMBS but also in the corporate and ABS markets which could reverse themselves in the coming months.