dorsal/arxiv
View SchemaAutomatic Trading Agent. RMT based Portfolio Theory and Portfolio Selection
| Authors | Malgorzata Snarska, Jakub Krzych |
|---|---|
| Categories | |
| ArXiv ID | physics/0608293 |
| URL | https://arxiv.org/abs/physics/0608293 |
| Journal | Acta Phys. Pol. B 37 (2006) 3145 |
Abstract
Portfolio theory is a very powerful tool in the modern investment theory. It is helpful in estimating risk of an investor's portfolio, which arises from our lack of information, uncertainty and incomplete knowledge of reality, which forbids a perfect prediction of future price changes. Despite of many advantages this tool is not known and is not widely used among investors on Warsaw Stock Exchange. The main reason for abandoning this method is a high level of complexity and immense calculations. The aim of this paper is to introduce an automatic decision - making system, which allows a single investor to use such complex methods of Modern Portfolio Theory (MPT). The key tool in MPT is an analysis of an empirical covariance matrix. This matrix, obtained from historical data is biased by such a high amount of statistical uncertainty, that it can be seen as random. By bringing into practice the ideas of Random Matrix Theory (RMT), the noise is removed or significantly reduced, so the future risk and return are better estimated and controlled. This concepts are applied to the Warsaw Stock Exchange Simulator http://gra.onet.pl. The result of the simulation is 18 % level of gains in comparison for respective 10 % loss of the Warsaw Stock Exchange main index WIG.
{
"annotation_id": "cf951b1d-3334-4927-af06-c3d99186f38b",
"date_created": "2026-03-02T18:01:10.527000Z",
"date_modified": "2026-03-02T18:01:10.527000Z",
"file_hash": "db6961e5f4f3fd74f2708ecaeaccd7fe0202772b6e32e17edb4027d5d9f0e023",
"private": false,
"record": {
"abstract": "Portfolio theory is a very powerful tool in the modern investment theory. It\nis helpful in estimating risk of an investor\u0027s portfolio, which arises from our\nlack of information, uncertainty and incomplete knowledge of reality, which\nforbids a perfect prediction of future price changes. Despite of many\nadvantages this tool is not known and is not widely used among investors on\nWarsaw Stock Exchange. The main reason for abandoning this method is a high\nlevel of complexity and immense calculations. The aim of this paper is to\nintroduce an automatic decision - making system, which allows a single investor\nto use such complex methods of Modern Portfolio Theory (MPT). The key tool in\nMPT is an analysis of an empirical covariance matrix. This matrix, obtained\nfrom historical data is biased by such a high amount of statistical\nuncertainty, that it can be seen as random. By bringing into practice the ideas\nof Random Matrix Theory (RMT), the noise is removed or significantly reduced,\nso the future risk and return are better estimated and controlled. This\nconcepts are applied to the Warsaw Stock Exchange Simulator http://gra.onet.pl.\nThe result of the simulation is 18 % level of gains in comparison for\nrespective 10 % loss of the Warsaw Stock Exchange main index WIG.",
"arxiv_id": "physics/0608293",
"authors": [
"Malgorzata Snarska",
"Jakub Krzych"
],
"categories": [
"physics.soc-ph",
"cs.CE",
"q-fin.PM",
"stat.AP"
],
"journal_ref": "Acta Phys. Pol. B 37 (2006) 3145",
"title": "Automatic Trading Agent. RMT based Portfolio Theory and Portfolio Selection",
"url": "https://arxiv.org/abs/physics/0608293"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "d9f86697-5755-4695-a8da-415991cb31f4",
"id": "arXiv Dataset IDs",
"type": "Model",
"variant": "snapshot-2026-03-01",
"version": "0.1.0"
},
"user_id": 1000002
}