Knowledge-driven Autonomous Commodity Trading Advisor

The myth that financial trading is an art has been mostly destroyed in the recent decade due to the proliferation of algorithmic trading. In equity markets, algorithmic trading has already bypass human traders in terms of traded volume. This trend seems to be irreversible, and other asset classes ar...

Full description

Saved in:
Bibliographic Details
Main Authors: LIM, Yee Pin, CHENG, Shih-Fen
Format: text
Language:English
Published: Institutional Knowledge at Singapore Management University 2012
Subjects:
Online Access:https://ink.library.smu.edu.sg/sis_research/1664
https://ink.library.smu.edu.sg/context/sis_research/article/2663/viewcontent/sky_commodity_iat12_final.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Singapore Management University
Language: English
id sg-smu-ink.sis_research-2663
record_format dspace
spelling sg-smu-ink.sis_research-26632019-12-10T02:53:29Z Knowledge-driven Autonomous Commodity Trading Advisor LIM, Yee Pin CHENG, Shih-Fen The myth that financial trading is an art has been mostly destroyed in the recent decade due to the proliferation of algorithmic trading. In equity markets, algorithmic trading has already bypass human traders in terms of traded volume. This trend seems to be irreversible, and other asset classes are also quickly becoming dominated by the machine traders. However, for asset that requires deeper understanding of physicality, like the trading of commodities, human traders still have significant edge over machines. The primary advantage of human traders in such market is the qualitative expert knowledge that requires traders to consider not just the financial information, but also a wide variety of physical constraints and information. However, due to rapid technology changes and the “invasion” of cashrich hedge funds, even this traditionally human-centric asset class is crying for help in handling increasingly complicated and volatile environment. In this paper, we propose an adaptive trading support framework that allows us to quantify expert’s knowledge to help human traders. Our method is based on a two-state switching Kalman filter, which updates its state estimation continuously with real-time information. We demonstrate the effectiveness of our approach in palm oil trading, which is becoming more and more complicated in recent years due to its new usage in producing biofuel.We show that the two-state switching Kalman filter tuned with expert domain knowledge can effectively reduce prediction errors when compared against traditional single-state econometric models. With a simple back test, we also demonstrate that even a slight decrease in the prediction errors can lead to significant improvement in the trading performance of a naive trading algorithm. 2012-12-01T08:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/1664 info:doi/10.1109/WI-IAT.2012.208 https://ink.library.smu.edu.sg/context/sis_research/article/2663/viewcontent/sky_commodity_iat12_final.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University autonomous trading commodity trading switching Kalman filter Artificial Intelligence and Robotics Finance and Financial Management Operations Research, Systems Engineering and Industrial Engineering
institution Singapore Management University
building SMU Libraries
continent Asia
country Singapore
Singapore
content_provider SMU Libraries
collection InK@SMU
language English
topic autonomous trading
commodity trading
switching Kalman filter
Artificial Intelligence and Robotics
Finance and Financial Management
Operations Research, Systems Engineering and Industrial Engineering
spellingShingle autonomous trading
commodity trading
switching Kalman filter
Artificial Intelligence and Robotics
Finance and Financial Management
Operations Research, Systems Engineering and Industrial Engineering
LIM, Yee Pin
CHENG, Shih-Fen
Knowledge-driven Autonomous Commodity Trading Advisor
description The myth that financial trading is an art has been mostly destroyed in the recent decade due to the proliferation of algorithmic trading. In equity markets, algorithmic trading has already bypass human traders in terms of traded volume. This trend seems to be irreversible, and other asset classes are also quickly becoming dominated by the machine traders. However, for asset that requires deeper understanding of physicality, like the trading of commodities, human traders still have significant edge over machines. The primary advantage of human traders in such market is the qualitative expert knowledge that requires traders to consider not just the financial information, but also a wide variety of physical constraints and information. However, due to rapid technology changes and the “invasion” of cashrich hedge funds, even this traditionally human-centric asset class is crying for help in handling increasingly complicated and volatile environment. In this paper, we propose an adaptive trading support framework that allows us to quantify expert’s knowledge to help human traders. Our method is based on a two-state switching Kalman filter, which updates its state estimation continuously with real-time information. We demonstrate the effectiveness of our approach in palm oil trading, which is becoming more and more complicated in recent years due to its new usage in producing biofuel.We show that the two-state switching Kalman filter tuned with expert domain knowledge can effectively reduce prediction errors when compared against traditional single-state econometric models. With a simple back test, we also demonstrate that even a slight decrease in the prediction errors can lead to significant improvement in the trading performance of a naive trading algorithm.
format text
author LIM, Yee Pin
CHENG, Shih-Fen
author_facet LIM, Yee Pin
CHENG, Shih-Fen
author_sort LIM, Yee Pin
title Knowledge-driven Autonomous Commodity Trading Advisor
title_short Knowledge-driven Autonomous Commodity Trading Advisor
title_full Knowledge-driven Autonomous Commodity Trading Advisor
title_fullStr Knowledge-driven Autonomous Commodity Trading Advisor
title_full_unstemmed Knowledge-driven Autonomous Commodity Trading Advisor
title_sort knowledge-driven autonomous commodity trading advisor
publisher Institutional Knowledge at Singapore Management University
publishDate 2012
url https://ink.library.smu.edu.sg/sis_research/1664
https://ink.library.smu.edu.sg/context/sis_research/article/2663/viewcontent/sky_commodity_iat12_final.pdf
_version_ 1770571444311293952