Relying On Systems And Mechanisms

Clearing Corporations

Q. Mr Narain, if a few members take very large positions then how will 250 crore or so suffice? How is the distinct clearing corporation any different from clearing within the exchange at such times? After all, at the end, the clearing corporation will also be unwinding trades.

A. If your concern is: can members take positions huge enough to sink the clearing corporation, the answer is Yes! It is obviously possible to dream up a position from which bankruptcy by a few members would wipe out any settlement fund.

The key here, and I cannot emphasize this enough, is the risk containment system which must prevent such situations from developing. The settlement fund is not intended to deal with any contingency taken in isolation - no settlement fund could be strong enough for this. The settlement fund is designed to cope, with a generous margin for error, with the crises which manage to slip past a stringent risk containment system.

 

Could this be done within the exchange? Sure, it could be. The reasons for separating this out into a distinct entity is to provide a clear focus upon the distinct objectives of the clearing corporation, and to separate out the balance sheet. Why would unwinding still be required? It would only come into the picture where the entire settlement fund is used up. Apart from such a disaster scenario, no unwinding of trades would take place.

Q. Mr Narain, how did you come up with this number of

Rs 270 crore? How much of it is liquid?

A. We looked at NSCC and OCC in the US very closely, and looked at the ratio between their risk at any point and their settlement fund. As business grows, the settlement fund will also grow commensurately. Roughly Rs 70 crore is cash. There is nothing sacrosanct about the breakup.

Q. Mr Narain, theres a member who fails to pay Rs 5 crore on Monday. The clearing corporation will step in and cover the liability. When will this position close out?

A. This isnt any different with the clearing corporation as compared with the way its done today. The short position will be auctioned on Wednesday as is the case today. Nothing will change.

Q. Suppose a default takes place, and you make a call -

A. Thats not correct. The issues of the fund size are distinct and separate from the handling of a default event. You are linking those two in an immediacy which is not present. Please dont think: a default occurs, a payout is made, and a call is made to members.

How the fund is managed, what size it is kept at, is a function of a number of parameters. The size of business, the experiences in handling difficult situations, the evolution of the risk containment system, all these will influence the size of the fund that is considered prudent. It may be six months or a year before a call is made, or it could never be made. How would I know today? We have never experienced a default on the NSE yet.

My point is just this: the management of a default situation, and the management and size of the settlement fund, are unlinked issues.

Q. Mr Mathur, what are BSEs plans about a clearing corporation?

A. We are trying to work out an insurance cover which will cover day-to-day risks of members (likes theft or fire). Insurance for default is not done by insurance companies, and we havent thought of doing a clearing corporation which guarantees trades so far.

Q. Mr Narain, how would the clearing corporation deal with a situation like MS Shoes, the biggest default ever on the BSE?

A. MS Shoes has been a very important source of motivation for our thinking about the clearing corporation. If the risk containment systems allow an MS Shoes to take place, and there is still a loss after using the margins, the clearing corporation will indeed pay the money required to prevent a cascading effect upon of the members.

Subsequent to that, the clearing corporation will then move into the networth of the members who have defaulted and take all recourse available to try to replenish the fund to the extent that it can. The key principle is: the contracting parties will be totally untouched.

Q. Mr Chandrasekaran, what do you think about NSEs plans for the clearing corporation?

A. I have long advocated not separating the clearing corporation from the depository. If part of the function is done in the clearing corporation, you cant have perfect DVP. Thus, I disagree with NSEs design of a clearing corporation which is distinct from the depository. We seem to have agreed upon allowing multiple depositories, which is inefficient, but multiple clearing corporations are also inefficient. If UTI has operations in 23 markets, it will have to deal with 23 clearing corporations. If we had exactly one depository + clearing corporation, things would be so much simpler.

Q. Mr Narain, would you like to react to this?

A. For the first point, there is no question in our minds that the clearing corporation and the depository cannot be unified. The simplest of the many reasons for this is that for many years from now, we will have to deal with physical paper in our country. There is no way in which this can be done using a unified clearing corporation and depository. They have to be kept distinct.

Q. Mr Chandrasekaran, what is your view of the risk adequacy of NSCC?

A. It is claimed that the clearing corporation will guarantee trades. How can it deal with daily volumes of Rs 1,500 crore using a fund of Rs 300 crore? Who will believe that the clearing corporation is actually capable of living up to the claims?

We need a great deal of paranoia about risk management here. If the top two membersfail, will the system cope? Im also concerned about the precise legal liabilities of members to the clearing corporation. If the clearing corporation takes my money and gives me bad delivery, that sounds unfair. Why should the clearing corporation take my money, give me bad paper, and take no responsibility for it?

Q. Mr Narain, would you like to respond to this?

A. On the subject of risk containment, it is a well-known fact that a good risk containment system can do away with the fund altogether. The clearing corporation knows its participants, and it has powerful controls over them. It is possible to build a completely secure clearing corporation using a tiny fraction of the funds that seem to be required for the purpose at first glance. Its a question of the sophistication and intelligence of the risk containment system which is built. The daily trading volume is a huge over-estimate of the true exposure of the clearing corporation, given that risk-containment procedures are in place.

The concern was raised: why would anyone believe you on this? As is the international practice, we will be getting the clearing corporation rated by independent credit rating agencies. That will be a clear signal about the strength of the risk containment system backed-up by the settlement fund.

Let me put all this in perspective. We did simulations where the top 20 brokers simultaneously default. Note that with the clearing corporation, the chains of payments which bind brokers and generate cascading bankruptcies, are cut at each point. Yet, we assumed that 20 brokers default. And, we assumed 20-40 per cent price movements in the adverse direction. We did not even reach 40 per cent of our startup settlement fund. This is the foundation of my confident assertions in this regard.

Technical Infrastructure

Q. Mr Rawalgaonkar, your software was developed on Unix. That means it can run on just about any Unix system?

A. Yes, our software will work on essentially any Unix system. In principle, you could buy a PC for Rs 1 lakhhe chains of payments which bind brokers and generate cascading bankruptcies, are cut at each point.