System and method for determination of incremental value at risk for securities trading

• US 5,819,237 A
• Filed: 02/13/1996
• Issued: 10/06/1998
• Est. Priority Date: 02/13/1996
• Status: Expired due to Term

• Current Assignee: Financial Engineering Associates, Inc. (ION Investment Group Ltd.)
• Original Assignee: Financial Engineering Associates, Inc. (ION Investment Group Ltd.)
• Inventors: Garman, Mark B.
• Primary Examiner(s): Weinhardt, Robert A.
• Assistant Examiner(s): GROUTT, PHILLIP J
• Application Number:
  US08/600,685
• Time in Patent Office: 966 Days
• Field of Search: 705/35, 705/36, 705/37, 705/38
• US Class Current: 705/36.00R
• CPC Class Codes: G06Q 40/04 Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
• CPC Class Codes: G06Q 40/06 Asset management; Financial planning or analysis

Abstract

A system, method, and product determines the incremental impact of any number of candidate trades on the value at risk (VaR) measure of a trading portfolio within a trading interval, without requiring that the VaR measure be redetermined individually with respect to each candidate trade. The method includes determining the VaR measure for the trading portfolio, and determining a derivative vector quantity for the VaR measure. For each candidate trade, the impact of the candidate trade on the VaR measure is determined as the vector product of the derivative vector and the mapped cashflows of the candidate trade. A negative sign indicates a desirable reduction in the VaR measure. This determination may be made for any number of candidate trades without having to re-determine the VaR measure. The software product employs this method in a financial analysis application in an optimized implementation. The system includes the software product along with databases storing the trading portfolio(s). Additionally, the method and product allow each candidate trade to be normalized with respect to selected criteria, so that a number of individual candidate trades may be ranked with respect to their incremental impact on the VaR measure to determine the candidate trade the best reduces the VaR measure.

Claims

1. A computer readable memory storing thereon a computer program for controlling the execution of a processor to determine whether any candidate trade in a selected trading interval reduces a value at risk measure of a trading portfolio having a selected financial instruments, the computer program controlling the processor to:

determine the value at risk measure for the trading portfolio once with respect to the trading interval;

determine a derivative vector of the value at risk measure;

determine a set of cashflows for the candidate trade;

determine an incremental value at risk measure for a candidate trade from the derivative vector of the value at risk measure and the set of cashflows for the candidate trade; and

,

determine from the sign of the incremental value at risk measure whether the value at risk measure of the trading portfolio is reduced by the candidate trade.

2. The computer readable memory of claim 1, wherein to determine the derivative vector of the value at risk measure for the trading portfolio, the computer program thereon controls the processor to:

convert the financial instruments in the trading portfolio into a first set of cashflows;

map the first set of cashflows unto a set of selected tenors;

determine a vector product W of transposed mapped cashflows and a scaled variance-covariance matrix of risk variances for the financial instruments in the trading portfolio at the selected tenors;

determine the value at risk measure from the square root of the inner product of W and the variance-covariance matrix; and

,

determine the derivative vector from the quotient of W and the value at risk measure.

3. The computer readable memory of claim 2, wherein the computer program thereon controls the processor to:

receive a plurality of candidate trades, each candidate trade including of at least one financial instrument;

for each candidate trade;

convert the candidate trade into a second set of cashflows;

map the second set of cashflows unto the set of selected tenors;

normalize the mapped second set of cashflows for the candidate trade by a normalization factor;

determine an incremental value at risk measure from the derivative vector and the mapped second set of cashflows for the candidate trade; and

,

rank each candidate trade by its incremental value at risk measure.

4. The computer readable memory of claim 3, wherein the normalization factor is:

a value at risk normalization factor determined according to the equation;

##EQU6## where λ

_i is the value at risk normalization factor, and a_i is the second set of cashflows.

5. The computer readable memory of claim 3, wherein the normalization factor is:

a return normalization factor equal to the value of the anticipated future returns of the candidate trade.

6. The computer readable memory of claim 3, wherein the normalization factor is:

a capital normalization factor equal to an amount of capital required sustain the candidate trade.

7. The computer readable memory of claim 3, wherein the normalization factor is:

a price normalization factor equal to a market price of the candidate trade.

8. The computer readable memory of claim 3, wherein the normalization factor is:

a notional normalization factor equal to a predetermined number of units for the candidate trade selected according to the type of the candidate trade.

9. The computer readable memory of claim 3, wherein the normalization factor is:

a cashflow normalization factor determined according to the equation;

##EQU7## where λ

_i is the cashflow normalization factor, and a_i is the second set of cashflows, and {c_j } is an arbitrary set of positive weights.

10. The computer readable memory of claim 3, wherein the normalization factor is:

a cashflow normalization factor determined according to the equation;

##EQU8## where λ

_i is the cashflow normalization factor, a_i is the second set of cashflows, and {c_j } is an arbitrary set of positive weights.

11. The computer readable memory of claim 3, wherein the normalization factor is:

a cashflow normalization factor determined according to the equation;

space="preserve" listing-type="equation">λ

.sub.i =||a.sub.i ||≡

max.sub.j {c.sub.j |a.sub.ij |}

where λ

_i is the cashflow normalization factor, a_i is the second set of cashflows, and {c_j } is an arbitrary set of positive weights.

12. The computer readable memory of claim 2 wherein the computer program thereon controls the processor to:

receive at least one candidate trade of at least one financial instrument;

convert the candidate trade into a second set of cashflows;

map the second set of cashflows unto the set of selected tenors; and

,

determine an incremental value at risk measure as the product of the derivative vector and the mapped second set of cashflows.

13. The computer readable memory of claim 2, wherein the computer program stored thereon controls the processor to:

determine the vector product W of transposed mapped cashflows p and the selected variance-covariance matrix Q according to the equation;

space="preserve" listing-type="equation">W=p'"'"'Q;

determine the value at risk measure, VaR, according to the equation;

space="preserve" listing-type="equation">VaR=Wp;

prior to determining the derivative vector, DELVAR, according to the equation;

space="preserve" listing-type="equation">DELVAR=W/Var.

14.

14. The computer readable memory of claim 1, wherein to determine the incremental value at risk for the selected set of candidate trades, the computer program thereon controls the processor to:

determine an incremental value at risk for each candidate trade in the selected set of candidate trades from the derivative vector and the set of cashflows for the candidate trade; and

accumulate the incremental value at risk for all of the candidate trades.

15. A computer readable memory storing thereon a computer program for controlling the execution of a processor to determine an incremental impact of a candidate trade on a value at risk measure of a trading portfolio P having selected financial instruments, the computer program controlling the processor to:

determine the value at risk measure VaR once for the trading portfolio for a selected trading interval;

determine the derivative vector of the value at risk measure VaR once for the trading portfolio for the selected trading interval;

determine a set of cashflows for the candidate trade;

determine the incremental impact of the candidate trade on the value at risk measure VaR from the product of the derivative vector and the set of cashflows for the candidate trade.

16. The computer readable memory of claim 15, coupled to an article of manufacture including:

a storage device storing at least one trading portfolio comprised of a plurality of financial instruments, and a selected scaled variance-covariance matrix Q of market risk variances for a plurality of financial instruments including the selected financial instruments, the market risk variances determined with respect to a set of selected tenors;

a processor coupled to the storage device, and controlled by the computer program stored in the memory to determine the value at risk measure VaR once for the trading portfolio for the selected trading interval by;

converting the financial instruments in the selected trading portfolio into a first set of cashflows;

mapping the first set of cashflows unto the set of selected tenors;

determining a vector product W of transposed mapped cashflows and the selected variance-covariance matrix; and

,

determining the value at risk measure VaR from as square root of the inner product of the mapped cashflows and the vector product W.

17. The computer readable memory of claim 16, wherein the computer program stored thereon controls the processor to determine the derivative vector of the value at risk measure VaR as the quotient of the vector product W divided by the value at risk measure VaR.

18. The computer readable memory of claim 15, the computer program stored thereon controlling the processor to:

receive a selected plurality of candidate trades;

for each candidate trade;

determine a set of cashflows for the candidate trade;

normalize the set of cashflows by a normalization factor;

determine for the candidate trade an incremental impact value on the value at risk measure VaR from the product of the derivative vector and the normalized set of cashflows of the candidate trade; and

,

rank the selected plurality of candidate trades by the incremental impact values.

19. A computer implemented method of determining whether any candidate trade in a selected trading interval reduces a value at risk measure of a trading portfolio, comprising:

determining the value at risk measure for the trading portfolio once with respect to the trading interval;

determining a derivative vector of the value at risk measure;

determining a set of cashflows for the candidate trade;

determining an incremental value at risk for a candidate trade from the derivative vector and the set of cashflows for the candidate trade; and

,

determining from the sign of the incremental value at risk whether the candidate trade reduces the value at risk measure of the trading portfolio.

20. The computer implemented method of claim 19, further comprising:

receiving a plurality of candidate trades, each candidate trade including of at least one financial instrument;

for each candidate trade;

normalizing the candidate trade by a normalization factor;

determining an incremental value at risk from the derivative vector and the normalized candidate trade; and

,

ranking each candidate trade by its incremental value at risk to determine at least one of the candidate trades that most reduce the value at risk measure of the trading portfolio.

21. The computer implemented method of claim 20, further comprising for each candidate trade:

converting the candidate trade into a second set of cashflows;

mapping the second set of cashflows unto the set of selected tenors;

normalizing the mapped second set of cashflows for the candidate trade by the normalization factor; and

,

determining an incremental value at risk from the derivative vector and the mapped second set of cashflows for the candidate trade.

22. The computer implemented method of claim 20, wherein the normalization factor is:

a value at risk normalization factor determined according to the equation;

##EQU9## where λ

_i is the value at risk normalization factor, and a_i is the second set of cashflows.

23. The computer implemented method of claim 20, wherein the normalization factor is:

a return normalization factor equal to the value of the anticipated future returns of the candidate trade.

24. The computer implemented method of claim 20, wherein the normalization factor is:

a capital normalization factor equal to an amount of capital required sustain the candidate trade.

25. The computer implemented method of claim 20, wherein the normalization factor is:

a price normalization factor equal to a market price of the candidate trade.

26. The computer implemented method of claim 20, wherein the normalization factor is:

a notional normalization factor equal to a predetermined number of units for the candidate trade selected according to the type of the candidate trade.

27. The computer implemented method of claim 20, wherein the normalization factor is:

a cashflow normalization factor determined according to the equation;

##EQU10## where λ

_l is the cashflow normalization factor, a_i is the second set of cashflows, and {c_j } is an arbitrary set of positive weights.

28. The computer implemented method of claim 20, wherein the normalization factor is:

a cashflow normalization factor determined according to the equation;

space="preserve" listing-type="equation">λ

.sub.i =||a.sub.i ||≡

Σ

c.sub.j |a.sub.ij |

where λ

_i is the cashflow normalization factor, a_i is the second set of cashflows, and {c_j } is an arbitrary set of positive weights.

29. The computer implemented method of claim 20, wherein the normalization factor is:

a cashflow normalization factor determined according to the equation;

space="preserve" listing-type="equation">λ

.sub.i =||a.sub.i ||≡

max.sub.j {c.sub.j |a.sub.ij |}

where λ

_i is the cashflow normalization factor, a_i is the second set of cashflows, and {c_j } is an arbitrary set of positive weights.

30. The computer implemented method of claim 19, wherein determining the value at risk measure comprises:

receiving a scaled variance-covariance matrix of market risk variances for a plurality of financial instruments including selected financial instruments in the trading portfolio, the market risk variances determined with respect to a set of selected tenors;

converting the financial instruments in the trading portfolio into a first set of cashflows;

mapping the first set of cashflows unto the set of selected tenors;

determining a vector product W of transposed mapped cashflows and the selected variance-covariance matrix; and

determining the value at risk measure from the mapped cashflows and the vector product W.

31. A computer implemented method of determining an incremental value at risk for a selected set of candidate trades in a trading portfolio during a selected trading interval, comprising:

determining a value at risk measure for the trading portfolio once with respect to the trading interval;

determining a derivative vector of the value at risk measure;

determining a set of cashflows for the selected set of candidate trades;

determining the incremental value at risk for the selected set of candidate trades from the derivative vector and the set of cashflows for the selected set of candidate trades.

32. The method of claim 31, wherein determining the incremental value at risk for the selected set of candidate trades further comprises:

determining an incremental value at risk for each candidate trade in the selected set of candidate trades from the derivative vector and the set of cashflows for the candidate trade; and

accumulating the incremental value at risk for all of the candidate trades.

33. A computer readable memory storing thereon a computer program for controlling the execution of a processor to determine an incremental value at risk for a selected set of candidate trades in a trading portfolio during a selected trading interval, the computer program controlling the processor to:

determine a value at risk measure for the trading portfolio once with respect to the trading interval;

determine a derivative vector of the value at risk measure;

determine a set of cashflows for the selected set of candidate trades;

determine the incremental value at risk for the selected set of candidate trades from the derivative vector and the set of cashflows for the selected set of candidate trades.