1. 제 목 : Optimizing sparse mean reverting portfolios for algorithmic trading
 

2. 발 표 자 : 야노스 교수
 

3. 일 시 : 2012년 5월 16일(수) 10:00 ~ 11:00
 

4. 장 소 : IT-1호관 318호 

      5. 초청교수 : 이동익 교수 

      6. 강사약력 :

1986, 1989, 부다페스트기술경제대 전기공학석사, 공학박사  

2004 DSc (Doctor of Sciences, 헝가리과학아카데미)

7. 내용요약 :

In this lecture, novel heuristic algorithms for finding sparse, mean reverting portfolios in multivariate time series are presented. The new results can be applied to developing profitable convergence trading strategies by identifying mean reverting portfolios which can be traded advantageously when their prices differ from their identified long-term mean. After mapping the optimal portfolio selection into a generalized eigenvalue problem, novel approaches are introduced to identify the optimal portfolio vector in a subspace which satisfies the cardinality constraint. These heuristic algorithms provide good approximate solutions to the identification of sparse portfolios (which is proven to be NP-hard) and have polynomial run-time. As a result, convergence trading becomes possible not only on daily- but on intraday price series, as well, and some of the methods could even be implemented for real-time, high frequency algorithmic trading. In addition to these results, we will interpret trading as a decision theoretic problem and present detailed algorithms for implementing convergence trading strategies. The effectiveness of the new methods is tested by extensive simulations on generated and historical real market data (SP500, SWAP rates, Forex data).

※ 주최 : BK21 정보기술연구인력양성사업단, 전자전기컴퓨터학부

◀ 문의처 : BK21정보기술연구인력양성사업단 ☎ 950-6613 ▶