Hybrid multi-thread and multi-process computer simulation system and methods

• US 7,734,525 B2
• Filed: 09/26/2006
• Issued: 06/08/2010
• Est. Priority Date: 09/27/2005
• Status: Active Grant

• Current Assignee: Morgan Stanley Services Group Inc. (Morgan Stanley)
• Original Assignee: Morgan Stanley
• Inventors: Zborovskiy, Mikhail, Turchin, Dimitri
• Primary Examiner(s): Kazimi; Hani
• Assistant Examiner(s): Vyas; Abhishek
• Application Number:
  US11/527,324
• Publication Number:
  US 20070208649A1
• Time in Patent Office: 1,351 Days
• Field of Search: 705 35- 45
• US Class Current: 705/36.00R
• CPC Class Codes: G06F 9/505 considering the load
• CPC Class Codes: G06Q 40/00 Finance; Insurance; Tax strategies; Processing of corporate or income taxes
• 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 method for performing a calculation that includes determining solutions for a plurality of problem modules. The problem modules are of differing complexities, and their solutions are combined to determine a solution to the calculation. The method may include directing each of the problem modules to at least one master server, estimating a complexity for each of the problem modules, determining a threshold complexity level, sending problem modules having a complexity exceeding the threshold complexity level to at least one slave server and obtaining solutions for the problem modules therefrom, determining solutions for problem modules having a complexity not exceeding the threshold complexity level in the master server(s), and combining the solutions for the problem modules to determine the solution for the calculation.

Claims

1. A method for quantifying a credit exposure of an investment portfolio owner to a counterparty within said investment portfolio, said method comprising the following steps:

for each trade type of a plurality of trade types, empirically determining a base complexity based on a time required to price a trade representative of said trade type, said representative trade embodying one or more parameters of said trade type;

determining, by a computer system comprising a first server and at least one second server, relative complexities for a plurality of trades of said investment portfolio, said plurality of trades representing obligations of said counterparty to said portfolio owner, wherein determining said relative complexities comprises;

electronically receiving, by said first server, data identifying said counterparty and said plurality of trades, wherein said first server comprises a processor and a computer-readable medium;

for each of said plurality of trades of said investment portfolio;

defining a first complexity for said trade that is equal to said base complexity of a corresponding one of said plurality of trade types;

analytically determining a second complexity for said trade based on;

a first modification of said first complexity based on one or more parameters of said trade when a time required to price said trade is dependent on said one or more parameters of said trade; and

a second modification of said first complexity based on a number of time horizons over which said trade is to be priced;

for each of said plurality of trades having a second complexity not exceeding a threshold complexity level, determining a price for said trade in said first server;

for each of said plurality of trades having a second complexity exceeding said threshold complexity level, distributing a portion of said data received by said first server to said at least one second server to determine a price for said trade in said at least one second server; and

determining said credit exposure of said portfolio owner to said counterparty based on an aggregation of said prices determined in said first server and in said at least one second server.

2. The method of claim 1, wherein determining said credit exposure comprises adding notional values of said plurality of trades and subtracting therefrom said aggregation of said prices.

3. The method of claim 1, wherein, for each trade having a second complexity notexceeding said threshold complexity level, said price is determined by at least one thread of a multi-thread process.

4. The method of claim 3 wherein said multi-thread process is performed on a plurality of software-based server engines.

5. The method of claim 1, wherein said trades having a second complexity less than said threshold complexity level are sorted based on relative values of said second complexity, and wherein said step of determining a price for each of said trades is performed in order of decreasing second complexity.

6. The method of claim 1 wherein said trades having a second complexity greater than said threshold complexity level are sorted based on relative values of said second complexity, and wherein said step of determining a price for each of said trades is performed in order of decreasing second complexity.

7. The method of claim 1, wherein said first server comprises a plurality of first servers, and wherein each first server is in communication with a plurality of second servers.

8. The method of claim 1, wherein said first server comprises a plurality of first servers, and wherein each of said second servers is in communication with a single one of said first servers.

9. The method of claim 1, further comprising, for at least one trade, decomposing a calculation of said trade into a plurality of calculation components and distributing said calculation components among a plurality of second servers.

10. A method according to claim 1, wherein determining a price for said trades includes representing a plurality of scenarios for future values of parameters relevant to said prices of said trades, each of said scenarios comprising future values of said parameters at a plurality of time horizons.

11. The method of claim 1 wherein said method is directed by a root script executed on a control server.

12. A computer readable medium having stored thereon instructions which, when executed by at least one processor, cause the at least one processor to:

determine relative complexities for a plurality of trades of an investment portfolio, said plurality of trades representing obligations of a counterparty to an owner of said portfolio, wherein causing said at least one processor to determine said relative complexities causes;

electronic receipt of data identifying said counterparty and said plurality of trades by a first server;

for each of said plurality of trades of said investment portfolio;

definition of a first complexity for said trade that is equal to a base complexity of a corresponding one of a plurality of trade types;

analytical determination of a second complexity for said trade based on;

a first modification of said first complexity based on one or more parameters of said trade when a time required to price said trade is dependent on said one or more parameters of said trade; and

a second modification of said first complexity based on a number of time horizons over which said trade is to be priced;

wherein said base complexity for each of said plurality of trade types is empirically determined based on a time required to price a trade representative of said trade type, said representative trade embodying one or more parameters of said trade type;

for each of said plurality of trades having a second complexity not exceeding a threshold complexity level, determine a price for said trade in said first server;

for each of said plurality of trades having a second complexity exceeding said threshold complexity level, distribute of a portion of said data received by said first server to at least one second server to determine a price for said trade in said at least one second server; and

determine said credit exposure of said portfolio owner to said counterparty based on an aggregation of said prices determined in said first server and in said at least one second server.