While working out the possible impact of the SEC’s proposal to require central clearing of triparty repurchase agreements, we realized that we short-changed the analysis of multilateral netting in our last post. Our explanation of the SEC’s example focused on just the cash side of the trades, which is to say the amounts to be paid. To appreciate multilateral netting fully, we need to consider the security side of the trades, what is to be delivered, as well. This post seeks to rectify our oversight.
Tale of Two Tables
Recall that the SEC provided the following example of trades that could be subject to multilateral netting.
Firm A is buying $90 million in U.S. Treasury securities from Firm B, Firm B is buying $80 million in the same U.S. Treasury securities from Firm C, and Firm C is buying $100 million in the same U.S. Treasury securities from Firm A.”
We used the following table to determine the net amount the FICC owed to (and was entitled to receive from) each Firm after novation and netting.
|Dollar Amount due from (to) FICC (in millions)|
|First||Second||Third||Net Amount Due|
We omitted the corresponding principal amount of securities due to (from) the FICC in each trade. The example did not provide the price of each trade (which is likely to be different), so we cannot tell what these amounts would be. But making the simplifying assumption that all the trades were at par produces a mirror image of the first table.
|Principal Amount due from (to) FICC (in millions)|
|First||Second||Third||Net Amount Due|
Hence, after giving effect to multilateral netting, Firms A and B must each deliver $10 million principal amount of the Treasury Security to the FICC in exchange for $10 million, and Firm C must pay the FICC $20 million in exchange for $20 million principal amount of the Treasury Security.
Why the Security Side Matters
In our analysis, we assumed that Firm C became insolvent and could not settle its trade. Without novation, Firm A would still owe Firm B $90 million, but would not receive from Firm C the $100 million it needs to pay this. What we overlooked was that Firm A would still have $100 million principal amount of the Treasury Security it was supposed to deliver to Firm C. Firm A could close out and cover its defaulted trade with Firm C by selling this Treasury Security to another dealer and using the proceeds to pay Firm B. There might be a problem with timing if the cover trade doesn’t settle on the same day as the trade with Firm B. In that event, Firm A could probably get a clearing loan to cover the payment to Firm B until the payment is received on the cover trade.
With novation, Firm A and Firm B would settle their net trades with the FICC, each receiving $10 million and transferring $10 million principal amount of the Treasury Security to the FICC. The FICC could close out its defaulted trade with Firm C and sell the Treasury Security to cover the $20 million due from Firm C. This would reduce the FICC’s exposure to Firm C from $20 million to any deficiency remaining after selling the Treasury Securities. If the FICC sells the Treasury Securities for 99.5% of par, for example, its claim against Firm C would be only $100,000.
Taking both sides of the trade into consideration, this example demonstrates how multilateral netting can have the following salutary effects:
- The FICC’s sale of $20 million principal amount of the Treasury Security may have less effect on its price than Firm A’s sale of $100 million principal amount;
- The FICC has liquidity reserves it can use to bridge the payments to Firms A and B and the receipt of proceeds from selling the Treasury Security;
- The capacity to bridge payments may give the FICC time to find a better offer for the Treasury Securities than Firm A might obtain; and
- Any residual claim against Firm C may be covered by margin held by the FICC.
Having more completely accounted for the benefits of multilateral netting, we will continue with trying to work out the impact of the SEC’s proposal.