IP Library › Granted Patent US 11,631,136

Granted Patent B2

US 11,631,136 · App. 17/682,120 · Granted Apr 18, 2023

Methods and systems for low latency generation and distribution of quote price direction estimates

Inventors: David Edward Taylor (St. Louis, MO); Andy Young Lee (Ballwin, MO); David Vincent Schuehler (St. Louis, MO)

Assignee: Exegy Incorporated

G06Q40/04 G06F17/18 G06N20/00 G06Q30/0201

View Patent ↗

Loading inventors, assignments & file history…

Monitor This Case

Get email alerts when status or documents change.

Order Certified Copies

Most orders are placed with the USPTO same day — all within 24 business hours.

Order via The Patent Place →

Pre-filled with this patent's details

Quick Facts

Patent No.: US 11,631,136
App. No.: 17/682,120
Granted: Apr 18, 2023
Kind: B2

External Links

USPTO Patent Center↗ Full Patent PDF↗

Abstract

Systems and methods are disclosed herein that compute trading signals with low latency and high throughput using highly parallelized compute resources such as integrated circuits, reconfigurable logic devices, graphics processor units (GPUs), multi-core general purpose processors, and/or chip multi-processors (CMPs). For example, an estimation that estimates a quote price direction for a quote on a financial instrument can be generated from streaming financial market data.

Claims (129)

1. A system for accelerated processing of streaming financial market data to derive trading signals at low latency, the system comprising:

at least one member of the group consisting of (1) a reconfigurable logic device, (2) a graphics processor unit (GPU), (3) a chip multi-processor (CMP), and (4) a multi-core general purpose processor (GPP);

the at least one member configured to (1) receive streaming financial market data pertaining to a plurality of financial instruments, (2) process the streaming financial market data through parallelized processing logic to compute a plurality of features of the streaming financial market data, and (3) compute a quote price direction estimation with respect to a price for a quote on a financial instrument as a weighted combination of the computed features, wherein each computed feature has a corresponding weight for the weighted combination;

wherein the computed features comprise features that are correlated to and predictive of the quote price direction estimation; and

wherein a selection of which features to use for the correlated and predictive features and what values to use for the corresponding weights are derived from a supervised machine learning model.

2. The system of claim 1 further comprising a processor configured to train the supervised machine learning model based on labeled training data representing known price direction changes for previous quotes to identify the correlated and predictive features and their corresponding weights.

3. The system of claim 1 wherein the parallelized processing logic is configured to compute a plurality of the correlated and predictive features in parallel.

4. The system of claim 1 wherein the computed quote price direction estimation comprises a probability that indicates a likelihood that a next change in price for the quote will be a higher price or a lower price relative to a current price for the quote.

5. The system of claim 4 wherein the quote comprises a current bid price for the financial instrument, and wherein the probability comprises a probability that a next change in bid price for the quote will be up or down relative to the current bid price.

6. The system of claim 4 wherein the quote comprises a current offer price for the financial instrument, and wherein the probability comprises a probability that a next change in offer price for the quote will be up or down relative to the curent offer price.

7. The system of claim 4 wherein the quote comprises a current bid price and a current offer price for the financial instrument, and wherein the probability comprises (1) a probability that a next change in bid price for the quote will be up or down relative to the current bid price and (2) a probability that a next change in offer price for the quote will be up or down relative to the current offer price.

8. The system of claim 4 wherein the quote comprises a current bid price and a current offer price for the financial instrument, and wherein the probability comprises (1) a first probability that a next change in bid price for the quote will be up or down relative to the current bid price and (2) a second probability that a next change in offer price for the quote will be up or down relative to the current offer price.

9. The system of claim 4 wherein the computed quote price direction estimation comprises (1) a first probability that indicates a likelihood that a next change in price for the quote will be a higher price relative to a current price for the quote and (2) a second probability that indicates a likelihood that the next change in price for the quote will be a lower price relative to the current price for the quote.

10. The system of claim 1 wherein the at least one member comprises the reconfigurable logic device.

11. The system of claim 5 wherein the features comprise (1) data that indicates a direction of a most recent bid price change for the financial instrument, (2) data that indicates a direction of a second most recent bid price change for the financial instrument, (3) data that indicates a direction of a most recent offer price change for the financial instrument, (4) data that indicates a direction of a second most recent offer price change for the financial instrument, and/or (5) a count of a number of bid quote updates since a most recent price change for the financial instrument.

12. The system of claim 6 wherein the features comprise (1) data that indicates a direction of a most recent offer price change for the financial instrument, (2) data that indicates a direction of a second most recent offer price change for the financial instrument, (3) data that indicates a direction of a most recent bid price change for the financial instrument, (4) data that indicates a direction of a second most recent bid price change for the financial instrument, and/or (5) a count of a number of offer quote updates since a most recent price change for the financial instrument.

13. The system of claim 7 wherein the at least one member updates the probabilities as new streaming financial market data is received and processed so that the probabilities change even when the current bid price and the current offer price for the quote do not change.

14. The system of claim 8 further comprising:

a trading application configured to (1) receive the first and second probabilities, (2) compare the first and second probabilities with a probability threshold, (3) make a directional prediction for the bid price in response to a determination that the first probability exceeds the probability threshold, and (4) make a directional prediction for the offer price in response to a determination that the second probability exceeds the probability threshold.

15. The system of claim 14 wherein the trading application is a first trading application, the system further comprising:

a second trading application configured to (1) receive the first and second probabilities, (2) compare the first and second probabilities with another probability threshold, wherein the another probability threshold is different than the probability threshold used by the first trading application, (3) make a directional prediction for the bid price in response to a determination that the first probability exceeds the another probability threshold, and (4) make a directional prediction for the offer price in response to a determination that the second probability exceeds the another probability threshold.

16. The system of claim 1 wherein the streaming financial market data comprises a plurality of messages that pertain to financial instruments, wherein each of a plurality of the messages comprises a plurality of fields of financial market data, and wherein the parallelized processing logic operates on a plurality of the fields to compute the features.

17. The system of claim 16 wherein the at least one member is further configured to evaluate the messages for the quote price direction estimation on a message-specific basis.

18. The system of claim 17 wherein the at least one member is further configured to append a subject message with the quote price direction estimation that has been computed for the subject message.

19. The system of claim 1 wherein the at least one member is further configured to present quote price direction estimations to one or more trading applications, wherein the quote price direction estimations are presented by the at least one member synchronously with the financial market data to which the quote price direction estimations pertain.

20. The system of claim 19 wherein the at least one member is further configured to pair the quote price direction estimations with corresponding quote price duration estimations for presentation of the paired quote price direction estimations and quote price duration estimations to the one or more trading applications.

21. The system of claim 1 wherein the at least one member is further configured to compute the quote price direction estimation for every quote for each of a plurality of financial instruments without impeding a flow of the streaming financial market data to consumers via the at least one member.

22. The system of claim 1 wherein the at least one member is further configured to generate or modify an order based on the quote price direction estimation.

23. The system of claim 1 further comprising a market making application that, for a market maker with (1) an open long position on the financial instrument at a first price and (2) a resting offer order for the financial instrument at a second price that is greater than the first price, is configured to modify the resting offer order to exhibit a higher price in response to the quote price direction estimation indicating an upward movement for the price of the quote.

24. The system of claim 1 further comprising a trading application that allocates capital for proprietary trading based on the quote price direction estimation.

25. The system of claim 1 further comprising a broker application that manages execution quality based on the quote price direction estimation.

26. The system of claim 1 further comprising a matching engine that supports order types that incorporate tactical trading logic based on quote price direction estimations.

27. The system of claim 10 wherein the reconfigurable logic device comprises a field programmable gate array (FPGA).

28. The system of claim 1 wherein the at least one member comprises the GPU.

29. The system of claim 1 wherein the at least one member comprises the CMP.

30. The system of claim 1 wherein the at least one member comprises the multi-core GPP.

31. A system for accelerated processing of streaming financial market data to derive trading signals at low latency, the system comprising:

wherein the quote comprises a current bid price for the financial instrument, and wherein the probability comprises a probability that a next change in bid price for the quote will be up or down relative to the current bid price; and

wherein the features comprise (1) data that indicates a direction of a most recent bid price change for the financial instrument, (2) data that indicates a direction of a second most recent bid price change for the financial instrument, (3) data that indicates a direction of a most recent offer price change for the financial instrument, (4) data that indicates a direction of a second most recent offer price change for the financial instrument, and/or (5) a count of a number of bid quote updates since a most recent price change for the financial instrument.

32. The system of claim 31 wherein the at least one member is further configured to compute the quote price direction estimation as a weighted combination of the computed features, wherein each computed feature has a corresponding weight for the weighted combination, the system further comprising:

a processor configured to train a machine learning model based on labeled training data representing known price direction changes for previous quotes to determine the corresponding weights for the features.

33. The system of claim 31 wherein the at least one member comprises the reconfigurable logic device.

34. A system for accelerated processing of streaming financial market data to derive trading signals at low latency, the system comprising:

wherein the quote comprises a current offer price for the financial instrument, and wherein the probability comprises a probability that a next change in offer price for the quote will be up or down relative to the current offer price; and

wherein the features comprise (1) data that indicates a direction of a most recent offer price change for the financial instrument, (2) data that indicates a direction of a second most recent offer price change for the financial instrument, (3) data that indicates a direction of a most recent bid price change for the financial instrument, (4) data that indicates a direction of a second most recent bid price change for the financial instrument, and/or (5) a count of a number of offer quote updates since a most recent price change for the financial instrument.

35. The system of claim 34 wherein the at least one member is further configured to compute the quote price direction estimation as a weighted combination of the computed features, wherein each computed feature has a corresponding weight for the weighted combination, the system further comprising:

36. The system of claim 34 wherein the at least one member comprises the reconfigurable logic device.

37. A system for accelerated processing of streaming financial market data to derive trading signals at low latency, the system comprising:

wherein the quote comprises a current bid price and a current offer price for the financial instrument, and wherein the probability comprises (1) a probability that a next change in bid price for the quote will be up or down relative to the current bid price and (2) a probability that a next change in offer price for the quote will be up or down relative to the current offer price; and

wherein at least one member updates the probabilities as new streaming financial market data is received and processed so that the probabilities change even when the current bid price and the current offer price for the quote do not change.

38. The system of claim 37 wherein the at least one member is further configured to compute the quote price direction estimation as a weighted combination of the computed features, wherein each computed feature has a corresponding weight for the weighted combination, the system further comprising:

39. A system for accelerated processing of streaming financial market data to derive trading signals at low latency, the system comprising:

a trading application; and

wherein the quote comprises a current bid price and a current offer price for the financial instrument, and wherein the probability comprises a (1) a first probability that a next change in bid price for the quote will be up or down relative to the current bid price and (2) a second probability that a next change in offer price for the quote will be up or down relative to the current offer price; and

wherein the trading application is configured to (1) receive the first and second probabilities, (2) compare the first and second probabilities with a probability threshold, (3) make a directional prediction for the bid price in response to a determination that the first probability exceeds the probability threshold, and (4) make a directional prediction for the offer price in response to a determination that the second probability exceeds the probability threshold.

40. The system of claim 39 wherein the trading application is a first trading application, the system further comprising:

41. A system for accelerated processing of streaming financial market data to derive trading signals at low latency, the system comprising:

wherein the streaming financial market data comprises a plurality of messages that pertain to financial instruments, wherein each of a plurality of the messages comprises a plurality of fields of financial market data, and wherein the parallelized processing logic operates on a plurality of the fields to compute the features.

42. The system of claim 41 wherein the at least one member is further configured to evaluate the messages for the quote price direction estimation on a message-specific basis.

43. The system of claim 42 the at least one member is further configured to append a subject message with the quote price direction estimation that has been computed for the subject message.

44. The system of claim 41 wherein the at least one member is further configured to compute the quote price direction estimation as a weighted combination of the computed features, wherein each computed feature has a corresponding weight for the weighted combination, the system further comprising:

45. The system of claim 41 wherein the at least one member comprises the reconfigurable logic device.

46. The system of claim 45 wherein the reconfigurable logic device comprises a field programmable gate array (FPGA).

47. A system for accelerated processing of streaming financial market data to derive trading signals at low latency, the system comprising:

the at least one member configured to (1) receive streaming financial market data pertaining to a plurality of financial instruments, (2) process the streaming financial market data through parallelized processing logic to compute a plurality of features of the streaming financial market data, (3) compute a plurality of quote price direction estimations with respect to a plurality of prices for a plurality of quotes on a plurality of financial instruments based on the computed features, and (4) present quote price direction estimations to one or more trading applications, wherein the quote price direction estimations are presented by the at least one member synchronously with the financial market data to which the quote price direction estimations pertain.

48. The system of claim 47 wherein the at least one member is further configured to pair the quote price direction estimations with corresponding quote price duration estimations for presentation of the paired quote price direction estimations and quote price duration estimations to the one or more trading applications.

49. The system of claim 47 wherein the at least one member is further configured to compute the quote price direction estimations as weighted combination of the computed features, wherein each computed feature has a corresponding weight for the weighted combinations, the system further comprising:

50. The system of claim 47 wherein the at least one member comprises the reconfigurable logic device.

51. The system of claim 50 wherein the reconfigurable logic device comprises a field programmable gate array (FPGA).

52. The system of claim 47 wherein the at least one member comprises the GPU.

53. The system of claim 47 wherein the at least one member comprises the CMP.

54. The system of claim 47 wherein the at least one member comprises the multi-core GPP.

55. A system for accelerated processing of streaming financial market data to derive trading signals at low latency, the system comprising:

wherein the at least one member is further configured to compute the quote price direction estimation for every quote for each of a plurality of financial instruments without impeding a flow of the streaming financial market data to consumers via the at least one member.

56. A system for accelerated processing of streaming financial market data to derive trading signals at low latency, the system comprising:

57. The system of claim 56 wherein the at least one member is further configured to compute the quote price direction estimation as a weighted combination of the computed features, wherein each computed feature has a corresponding weight for the weighted combination, the system further comprising:

58. The system of claim 56 wherein the at least one member comprises the reconfigurable logic device.

59. The system of claim 58 wherein the reconfigurable logic device comprises a field programmable gate array (FPGA).

60. The system of claim 56 wherein the at least one member comprises the GPU.

61. The system of claim 56 wherein the at least one member comprises the CMP.

62. The system of claim 56 wherein the at least one member comprises the multi-core GPP.

63. A system for accelerated processing of streaming financial market data to derive trading signals at low latency, the system comprising:

a market making application; and

wherein the market making application, for a market maker with (1) an open long position on the financial instrument at a first price and (2) a resting offer order for the financial instrument at a second price that is greater than the first price, is configured to modify the resting offer order to exhibit a higher price in response to the quote price direction estimation indicating an upward movement for the price of the quote.

64. A system for accelerated processing of streaming financial market data to derive trading signals at low latency, the system comprising:

a trading application; and

wherein the trading application allocates capital for proprietary trading based on the quote price direction estimation.

65. A system for accelerated processing of streaming financial market data to derive trading signals at low latency, the system comprising:

a broker application; and

wherein the broker application manages execution quality based on the quote price direction estimation.

66. A system for accelerated processing of streaming financial market data to derive trading signals at low latency, the system comprising:

a matching engine; and

wherein the matching engine supports order types that incorporate tactical trading logic based on the quote price direction estimation.

67. A method for accelerated processing of streaming financial market data to derive trading signals at low latency, the method comprising:

streaming financial market data through at least one member of the group consisting of (1) a reconfigurable logic device, (2) a graphics processor unit (GPU), (3) a chip multi-processor (CMP), and (4) a multi-core general purpose processor (GPP), wherein the streaming financial market data pertains to a plurality of financial instruments;

the at least one member processing the streaming financial market data through parallelized processing logic to compute a plurality of features of the streaming financial market data; and

the at least one member computing a quote price direction estimation with respect to a price for a quote on a financial instrument as a weighted combination of the computed features, wherein each computed feature has a corresponding weight for the weighted combination;

wherein the features comprise features that are correlated to and predictive of the quote price direction estimation; and

wherein a selection of which features to use for the correlated and predictive features and what values to use for the corresponding weights are derived from a supervised machine learning model.

68. The method of claim 67 wherein the at least one member comprises the reconfigurable logic device.

Assignments (2)

SECURITY INTEREST Recorded Dec 18, 2025

From: EXEGY INCORPORATED

To: FIRST-CITIZENS BANK & TRUST COMPANY

Reel/Frame 073257/0796 →

ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Feb 28, 2022

From: TAYLOR, DAVID EDWARD; SCHUEHLER, DAVID VINCENT; LEE, ANDY YOUNG

To: EXEGY INCORPORATED

Reel/Frame 059118/0692 →

Continuity (3)

Continuation 16874474 · May 14, 2020

Provisional Application 62847641 · May 14, 2019

Related Publication 20220277393A1 · Sep 1, 2022

Cited By (5)

US 12,340,414 US 12,354,160 US 12,412,213 US 12,620,028 US 12,718,295