IRS portfolio compression is built on a simple idea. As dealer firms continue to trade swaps with each other, trades can begin to be removed without impacting the interest rate risk profile of the swap portfolio.
For example, Dealer A may have written a swap with Dealer B such that it is receiving fixed rate cash flows from Dealer B with a maturity slightly longer or shorter than five years. Suppose a new swap is written for the same notional amount where Dealer A pays fixed rate cash flows to Dealer B for five years. In this case, the macro interest rate risks have been largely offset. Dealer B is paying fixed in the first swap while Dealer A is paying fixed in the second swap. The notional principal is the same. All the dealers need to do is determine the difference in value between the two swaps caused by the mismatch in dates and in the fixed rate payments. If the parameters for valuing these mismatches are agreed, there is no reason why the swaps cannot literally be torn up – this is what is meant by the term compression in a bilateral context.
This type of exercise had been going on in an ad hoc manner for years between dealers as a means of reducing notionals, cleaning up portfolios and reducing operational costs. ISDA reports that an exercise similar to compression was used in 1999 to tear up nearly $1 trillion of interest rate swaps at Long Term Capital Management.
The examples used in this appendix require dealers to tear up and rewrite swap contracts. It may sound simple in concept but it involves considerably more negotiation among dealers than the successful process currently in use managed by TriOptima.
Multi-Lateral Compression
The benefits of compression between two dealers are obvious. To achieve meaningful reductions in notional outstanding, however, bilateral compression is a cumbersome exercise. If applied across the dealer community, meaningful results for the industry would involve literally hundreds of compression exercises per currency. Multi-lateral compression, on the other hand, can produce tremendous results in a very efficient manner provided there is widespread acceptance by market participants. This is the approach adopted by TriOptima in its triReduce service.
Consider, for example, a closed world where there are only four dealers: A, B, C and D. Consider as well their interdealer swaps in five years are as follows
As can be seen, the total Net Amounts of Rs are 100 and the total Net Amounts of Ps are 100. This is a closed system. All the swaps among the four have to net to zero. In a bilateral compression world, the dealers will not be able to compress any trades because in this simple world they only have one swap with each other dealer.
In a multi-lateral compression world, all the work can be done at once. Dealer A needs to receive fixed for 25 and Dealer B needs to receive fixed for 75. C and D need to pay fixed for 50 each. The compression will result in A receiving 25 from C and B receiving 25 from C and 50 from D. We started with 400 of notional and are down to 100.
This was a very simple example but it shows the value of increasing participants in compression.
