Derivative transactions governing system and network.
The invention relates in general to a system for a network based, automated process to decrease the number of outstanding financial derivative transactions between counterparties with the aim of lowering the gross and net credit exposures between counterparts and/or to enhance risk-control and/or lower operational costs and/or -risks for each counterpart.
State of the art
Introduction to financial derivatives
Financial derivatives are contracts of which the price/value of the contract varies with the value of an underlying instrument. The underlying instrument may be a bond, an equity, an interest rate, a currency exchange rate, commodity price or, in a recent development, even a credit risk.
Financial derivatives can either be standardised contracts traded on a recognised Exchange or OTC traded (see below). In the first case one of the counterparties in a trade is an Exchange. As the contracts are standardised there is no difference between the net and gross outstanding volume for each market participant. Buying on day 1 and selling on day 2 leads to a total disengagement of the position. Futures and exchange- traded options are common instruments in this class.
As the prices of these instruments are objectively known at any point in time they can fulfil the function of "benchmark instruments" for the calibration of other prices. This "benchmark" function can as well be fulfilled by some liquid underlying instruments, like government bonds, stocks of companies with a sufficient high market capitalisation or frequently traded currency-pairs.
The other possibility is that the contracts are individually negotiated and tailor- made between 2 counterparties (so called "over the counter (OTC) transactions"). In this case entering into a contract on day 1 and entering into a similar contract with the opposite sign on day 2 may lead to a market risk position of close to zero, but still to 2 contracts which have to be maintained until their designated final maturity. Instruments in this class are swaps, swaptions, FRA's, caps, floors, FX-forwards, all kind of options
on stocks, bonds, interest rates or currencies, as well as more exotic varieties and credit derivatives. Since the OTC contracts are individually negotiated, no objective market price exists. The parties need to do their own valuation of the contracts both when dealing as well as later on when establishing the market value of their contracts during their life. This valuation is normally based on pricing formulas using so called valuation parameters for input. There are different models that can be used, depending on the preferences of a party. The valuation parameters are either estimated or derived from benchmark instruments using other theoretical models. This subjective method of establishing the value of a contract means that the parties do get to different results. The differences are small in simple products but can be quite substantial in more complex products.
Active players in these instruments may run books that comprise of 10's of 1000's individual transactions, (the largest dealers even have more than 200.000 outstanding contracts) . The management of these positions is no longer possible on the level of the individual transactions, but only on an aggregated level. A common way of risk management is to sort the OTC-instruments by underlying instrument, and then express the market value and some risk-characteristics in time-buckets.
Frequently used risk-characteristics or sensitivities to the valuation parameters are for example the so-called greeks, delta, gamma, vega and theta. Delta refers to the first differential from the value of the trade for a change in value of the underlying instrument, thus enabling to judge the risk-content of the position expressed in units of the underlying instrument. Gamma refers to the second differential from the value of the trade for a change in value of the underlying instrument, or the change in delta for each subsequent change in the value of the underlying instrument. Gamma can be extremely non-linear, even for individual trades, and certainly for portfolios of many trades. Vega is the change in value of the instrument for a small change in the volatility of the instrument. Theta is the change in value of the instrument from 1 day to the next, all other parameters held constant. In case the contracts have a foreign currency component, the net present value of the foreign currency amount is sensitive to changes in the FX-rates. Hence, these net present values should also be considered a risk- characteristic. Where the text later on refers to the greeks, all sensitivities to the valuation parameters should be included. The greeks from individual trades can be aggregated for a portfolio of trades to form the corresponding greeks of the portfolio. A
set of trades is defined as being risk-neutral when the aggregated greeks are close to zero. Hence, the aggregated value of a risk-neutral set of trades is quite insensitive to changes in the valuation parameters .
The institutions active in the OTC markets use computer systems for storing the details of the transactions they have done. These systems normally have the capability to revalue all transactions and calculate the greeks of all transactions on a unilateral basis. A normal procedure is also to do a daily mark-to-market, which means re-valuation of all transactions according to the prevailing market conditions. The mark-to- market values as of close of business are stored in databases for calculating profit and loss, risk exposure, credit exposure etc. The greeks or the risk parameters can be calculated in different ways using different units or different scaling factors. Semantically, the risk parameters contain the same information although the numerical values may differ, depending on the methodology used by the risk-management system.
The professionally dealing participants in the OTC markets are banks, investment banks and other financial institutions (later referred to as dealers or market-makers). ISDA, the International Swap and Derivatives Association has roughly 180 primary members. A substantial number of these dealers consider themselves to be "market- makers" in a more or less broad spectrum of products. This means that they are always willing to put a price on a potential transaction that is brought to their attention by a customer or another dealer, be it direct or through a broker. In roughly 80% of the transactions both participants are dealers/market-makers. Only 20% of the transactions have an "end-user" as party to the agreement. The role of the market maker is to provide liquidity to the market. The motivation for the market maker is to deal OTC derivatives in high volumes, both "buying" and "selling", making revenues out of the bid/offer spread.
The total outstanding volume of interest rate based OTC derivatives has passed the 60 TRILLION US$, the OTC market as a whole has a size of over 88 TRILLION US$ according to statistics as of December 1999 collected by the Bank for International Settlements. The motivation for entering into an OTC derivative transaction is the management of market risk positions, either for the purpose of hedging (immunising a position from changes in market prices) or for the purpose of deliberate position taking (trying to make a profit from an expected change in market prices).
The advantage of derivatives is that the market risk of substantial amounts can change hands without the need to really transfer the underlying (nominal) values.
Problems related to the state of the art
As explained above, the purpose of entering into an OTC transaction is the transfer of market risk, either to increase or decrease the total exposure to certain market parameters, depending on the preferences of the counterparties. As an unavoidable consequence however, the parties are getting a new exposure to 2 different risk categories, credit risk and operational risk.
As the market moves, the net present value of an transaction will move away from the roughly zero level at which the transaction was concluded. This means, one party gets a claim on the other party, the validity of the transfer of market risk is therefor always dependent on the ability of the parties to fulfil their obligations under the contract. It is of little use to have had the correct market view and get a claim of X on a counterparty that has gone bankrupt in the mean time. Therefore, an important factor in the development of the OTC markets has been the management of credit risk.
Originally, dealers calculated this credit risk as a percentage on each transaction (depending on the type of transaction and the maturity) and charged the so calculated amounts against the credit risk limit their institutions were willing to expose themselves to each counterparty. Bank supervisors force banks to hold capital against both market risk as well as credit risk. Therefore each transaction or a certain sub-set of transactions, for example all transactions with counterparty X. will be internally charged with the cost of capital tied up in the transaction. The total capital of the bank is a limited and scarce resource that effectively puts a limit on the volume of banking activities. It is therefor desirable for a bank to try to reduce the capital tied up by the unintentionally acquired credit risk in OTC derivatives.
As the market grew rapidly, it was soon discovered that the sum of all trans- actions between 2 counterparties was much less dependent on market changes then the sum of the individual transactions. Many active market participants are dealing back and forth with each other in both directions, sometimes even on a daily basis. This reduces the sensitivity of the portfolio: "all transactions with counterparty X" for
changes in market values. For this reason more sophisticated institutions are now calculating the credit exposure based on the sum of all transactions with each counterparty.
This development has been made possible by the fact that ISDA developed the so called "Master Agreement", whereby all individual trades are deemed to be one Agreement, and many countries around the world changed their bankruptcy laws to accommodate this development (the so called "netting" process). The European Banking Federation has recently as well developed a Master- Agreement. The bi-lateral netting will, for the banks that are using this method, reduce the credit exposure. Still, many banks are not using netting, thus calculating exposures on a gross basis, or with a combination, i.e. they use netting in combination with a limited "add-on" for each individual trade.
The credit exposure can for obvious reasons never be netted between counterparties. A claim on party X can not be offset by a debt to party Y. The asymmetric nature of credit exposure means that the total exposure of a bank normally grows with the number of counterparties it is dealing with. The problem is aggravated if the exposures are calculated on a gross basis, since then the exposure even grows with the number of transactions.
In another attempt to prevent the exposure from growing outside the available limits of a bank, especially in the longer time buckets, which are the most risky and therefor the most scarce, counterparties agree on so called "break-clauses". This allows one or both parties to terminate the transaction at specified points in time at the prevailing market price.
A party may also choose to assign his part of a deal to a new party. In such an assignment, a deal between two parties A and B, where A assigns his part of the deal to a third party C, C will step in as the new counterparty towards B. By an assignment, both A, B and C can get reduced or increased exposure towards the other parties. The assignment requires the consent of both party B and C.
Even then a substantial number of potential trades could not be done, because counterparties were no longer willing to accept each other. Hence, the concept of "credit-mitigation" has been invented. In this process, when the net present value of all transactions exceeds a certain, bi-lateral set limit, their parties deliver securities or cash to each other, as a guaranty for their obligations under the contract. This process is
called "collateralisation". It does in no way eliminate or reduce the underlying reason for the unwanted credit exposure, it only mitigates the effect of the exposure. To establish the value of cash or securities one party has to transfer to the other party, parties have to agree on the value of the derivative contracts. This can either be done in a multiple of bi-lateral relationships, or through a central service. Cedel Bank in Luxemburg has provided this service. (See the patent WO 97/03409, SAMPSON,
Method and system for providing credit support ). For several reasons, this service was not accepted by the marketplace and is for the moment discontinued. Which ever method is chosen, the collateralisation process incurs a lot of operational costs and operational risk, since the value of the credit exposure (the sum of the outstanding derivative contracts) and/or the value of the collateral may change in such a way that it no longer covers the credit exposure.
Another way of addressing the costs of OTC derivatives dealing is clearing. One example of this is a service called SwapClear provided by the London Clearing House. Counterparts that are members of a clearinghouse may choose to have the clearinghouse stepping in as the counterpart on both sides of the transaction. By doing this for several deals and counterparts, a clearing member can aggregate a lot of OTC derivative deals to the same counterpart and hence get an improved credit exposure netting effect. The clearinghouse is also calculating a settlement value on the cleared transactions in order to keep track of the net exposure towards their members. Any residual debt or claim is settled daily in cash on a margin account. Some clearinghouses have lately become exempt from capital charges according to the capital adequacy legislation in many countries. This means that the capital costs for cleared deals effectively becomes close to zero. Some clearinghouses also provide other administrative services where some of the operational costs and risks are reduced. The clearinghouses however do not clear all transactions under a master agreement. Hence, the netting arrangement under the master agreement is broken in two parts, one part that moves to the Clearing House, and another part that stays under the master agreement. This means that there will be a substantial, possibly even a higher, residual credit exposure under the master agreement.
Far from all OTC derivative market participants are members of a clearing house and it is only when both parties are members of the same clearinghouse that a deal can be cleared. When a clearinghouse steps in as the counterpart on both sides of the
transaction, the number of outstanding deals is doubled. Even for cleared deals, there are substantial operational costs and risks associated with the number of outstanding deals, both within the Clearing House as for its members. They still have to perform reconciliation of the tasks performed by the Clearing House as well as other administrative tasks linked to the number of outstanding transactions and risk management.
It is fair to say that both collateralisation and clearing are curing the symptoms of the problems and not the cause. The cause of the problems, be it unwanted credit risk or operational costs and risks is the high number of outstanding bi-lateral contracts in such a size that the market risk of the sum of these contracts in many bi-lateral relationships is close to zero. However, so far there has been no efficient mechanism to make parties agree on the early termination of these unnecessary transactions.
If the aggregated greeks of all transactions done between two counterparties are high, the credit exposure will vary a lot as the market is changing. This sensitivity to the market movements will cause problems in the sense that credit limits or collateral thresholds may involuntarily be exceeded and that collateral obligations may vary causing a lot of collateral transactions or large movements in the margin account at the Clearing House. For this reason is it desirable for the counterparts to reduce the aggregated greeks in their bi-lateral transactions. Operational risk can be defined as all risks an institution is exposed to, other than the intentionally accepted market and credit risk within the allowed limits of the bank. This includes, among others, unwanted market or credit risk as a consequence of staff errors (on purpose or by accident) or errors in the risk management systems, fraud by staff, customers or third parties, wrong valuation of transactions based on wrong parameter input, or mistakes in the maintenance of outstanding contracts during their lifetime (so called fixings, settlements, assignments, exercises or payments).
Although the majority of transactions has a lifetime of up to 5 years, transactions of 30, 40 or even 50 years do occur as well. A trader or risk manager, who for some reason is using the wrong valuation parameters, may build up substantial losses over a long period of time. Smaller banks and certainly many end-users are lacking the possibility to do an independent mark to market of their outstanding contracts.
Many high profile mishaps have been published over the years, (even to the extent that some commentators were seeing all derivatives as "evil" instruments) and this
should be considered only as the tip of the iceberg; only the low frequency/high impact accidents are getting publicity. The high frequency/low impact accidents, of which the total amount of damage may well exceed that of their published cousins are frequently overlooked. Forced by these developments, as well as higher legal standards, market participants have invested a lot of effort, staff and systems to prevent accidents from happening. Procedures have been sharpened, internal regulations improved and rigid control measures established. The overhead cost of dealing has for this reason increased dramatically, and still accidents do happen from time to time.
The operational risk is a concern for the supervisors of the financial markets. There are proposals for introducing capital charges to cover operational risk. So far, the supervisors have not been able to work out principles or a model for how these capital charges will be calculated. The issue is still on the agenda though, and it is likely that any such model will generate capital costs that grows with the number of outstanding deals. In addition to these new types of risks, the acquired market risk may well be according to the trader's intention. Still bank supervisors force the banks to hold capital against market risks as well. An individual transaction's contribution to the total market risk, for which capital must be held, can be quite complicated to calculate, especially when the banks are using so called proprietary models for these calculations. However, one method, accepted by the bank supervisors and used by many banks, is based on transforming all off-balance sheet items, e.g. OTC derivatives, into roughly equivalent on-balance sheet items to calculate the capital cost. In this case it is possible to assign a capital cost to each individual transaction.
Purpose of the invention
The object of the invention is to solve the above described problems related to the state of the art at the root, i.e. to decrease the number of outstanding financial derivative transactions between counterparties with the aim of lowering the gross and net credit exposure between counterparts, and/or to lower operational costs and/or risks and/or to enhance risk-control for each counterpart. The system has to create a tool that facilitates the early termination or assignment of transactions, both of which requires the consent of the counterparts.
The invention
The establishment of a computer based communications network and system, either through the Internet or any other suitable public or private network, with or without the help of centralised computing power, with or without accessing data in a centralised computing centre, collected for other purposes, which will perform the following tasks:
Allow communication between counterparties (dealers/market-makers as well as end-users) in derivatives, especially in OTC derivatives, to send to each other, or to a centralised computing centre, or allow access to a centralised computing centre where the whole or part of the data is collected, on a regular basis , possibly daily or in several daily cycles or even real-time, the following information:
1. The trade details necessary for an unambiguous identification of individual transactions of OTC derivative trades that the parties wants to take part in the following process. 2. The net present value ("mark-to-market" value) and the sensitivities to the valuation parameters (the greeks) calculated by the counterparties, using their own trading or risk-management systems, for all individual trades described under 1.
3. The aggregated value of all trades under the master agreement as well as any collateral that has been exchanged between the counterparts to cover their obligations under a master agreement as well as the threshold amounts under the collateral agreement.
The system will then:
Adjusting and converting input parameters to a common format
4. The net present values and the sensitivities will be converted to a common format, adjusting the units and scaling factors.
Trade matching and termination proposal routine 5. Identify the trades sent by the both parties, and selecting the trades that both parties wants to take part in the following process. 6. As the system has established that the net present value (mark-to-market value) and the sensitivities (greeks) are in a comparable format, the system will compare the
net present value and the greeks of the individual transactions. Assuming that the differences in the fore-mentioned values are within pre-agreed tolerance ranges, the system will start the process described from 9 below.
7. If the differences in the values in 6 are outside the pre-agreed tolerance levels, the parties involved will be notified of this mismatch. The system will store the mismatches for future analysis of the differences, whereby the system will be able to discriminate between random differences and systematic differences, whereafter the parties involved will be informed of the results of this analysis.
8. Participating counterparties will each enter into the system: - which trades should not be considered in the composition of termination proposals. which trades with which counterparts they want to be considered in the composition of assignment proposals if they will allow partial termination of existing trades.
9. The parties will enter tolerance-limits as they see fit on differences in the net present values, maturity dates and/or one or more of the greeks to identify intervals they consider to be risk-neutral. The system will now produce a list of individual transactions, fulfilling the criteria mentioned under 5 and 6, (the matching process) of which the net present values and the greeks match in total in such a way that they can considered to be risk-neutral. A risk neutral package of transactions is by definition insensitive to differences in valuation parameters. Hence, there is a high likelihood that the counterparts will agree on the value of such a risk neutral package. Also the value of such a package is not sensitive to market movements or when the mark-to-market values was established. Since the package is risk neutral, the market risk position is unchanged and the counterparts does not need to make any new transactions to retain their desired market position. The parties can agree to prematurely terminate these transactions (the termination proposal).
10. Participating counterparties can set the tolerance-limits on differences in the net present values, maturity dates and/or one or more of the greeks in such a way that existing exposures between parties can be reduced. The system will produce a termination proposal with the explicit aim of reducing credit exposures and collateral obligations, to prevent collateral obligations from occurring, where the
aggregated risk of the proposal is minimised. The system will also produce a termination proposal, with the aim of reducing the variability of the exposure. 1 1. The system may produce legally binding trade terminations if all pre-agreed parameters are met. 12. A termination should be interpreted as follows. It can either be a direct contractual agreement to tear up an existing contract, or to enter into a new inverted contract if there exists a way of legally cancelling two such mirror deals. A termination can also apply between two parties, A and B having the opposite position in a contract, in the case where a third party have taken the place as the counterpart to both these parties and the third party agrees to effectively cancel both A's and B's contracts on request of both A and B. This is also independent of the fact if A and B were the original counterparts in the transaction. 13. With the help of centralised computing power, the matching process can be expanded to tri-lateral or multilateral assignment and termination proposals. In this case, the chances are even higher to find packages of transactions that both reduces the net credit exposures and are risk neutral to all involved counterparts. This is due to the fact that when looking at three counterparts A, B and C, in a situation where they have circular claims, i.e., where A has a claim on B, B has a claim on C, and C has a claim on A, all parties can reduce their net exposures by the smallest amount of the claims in question, without having any net cash flow effect. In the bilateral case, reducing the credit exposure means that the debtor needs to make a cash payment to the creditor. Only the creditor gets a credit exposure reduction. Also, the two parties may have different funding costs and thereby attach different value to cash. This asymmetric benefit means that it is more difficult to get the parties to agree to perform credit exposure reducing terminations in the bilateral case. The multilateral termination proposals where there are circular claims make the proposal beneficial to all involved parties. Also when three or more parties are involved, two parties, A and B, that have opposite exposures to a third party C, can chose to assign a package of deals with C from A to B or vice verse, in order to reduce the exposure to C. The assignment package can also be combined with assignments to a fourth party D, where both A and B have the opposite exposures as they have to C in such a way that both A and B get a credit exposure reduction. The assignment package can possibly be combined with some terminations between
A and B in such a way that the whole package is still cash and risk neutral for both A and B. From a large population of counterparts that have dealt with each other, there will be many groups of three banks that have circular claims. The number of sub-groups of three out of a population of N parties, is proportional to the third power of N. In such a sub-group of three, the likelihood is 25 percent that the claims are circular, if the direction of all claims are evenly distributed in the population.
14. The system will also produce a list of all transactions where the individual or aggregated mark-to-market valuation of the transactions are such, that it would be favourable to both parties to terminate the transactions. This will happen when both parties valuations are within the tolerance range set by both parties.
15. If a participant enters the necessary information into the system, the system will quantify the savings in terms of capital for market risk, credit risk, operational risk en future operational costs that the participant can gain by following the termination proposal
16. The market conditions will most likely have changed from the time when the net present values and the risk parameters where calculated until the time when the termination proposals can be executed. Still it is not likely that the market conditions have changed so dramatically that the termination proposals are invalid. In most cases a new revaluation must be done in order to determine the actual termination fee of the package. The system will support this new revaluation by managing quote request for the termination packages using the prevailing markets rates
17. The system will support the internal enterprise-wide risk-management-tools of par- ticipants by giving them an external calibration tool.
18. The system will be capable, for each new deal that one user is about to make, to propose with which counterparts it would be favourable to execute the deal. This would be done by finding the counterparts where adding the deal will lower the net exposure on one or more of the greeks, thereby making the sum of the transactions less prone to shifts in the valuation parameters and possibly allow for an termination proposal in the near future, or where the deal would reduce current credit exposure.
19. The procedure described in 18. can also be used in an environment where
automated quote request are send using systems and networks, to automatically select the counterparts where the deal would be most favourable.