habib.zouaoui@gmail.com Nadjat_meriem@outlook.fr
Measuring of Optimal Investment Portfolio Using Genetic Algorithm
Case of Algeria Stock ExchangeHabib ZOUAOUI 1& Meriem-Nadjet NAAS 2 1. University of TAHAR Moulay SAIDA –Algeria
2. University Center of Relizane –Algeria
Received:04 Aug 2014 Accepted: 26 Dec 2014 Published: 30 June 2015
: ) GA ( 2010/01/31 -2010/12/31 . . jel G11 Abstract:
This study aims to address the problems of risk management portfolio of bank loans based on the diversification strategy sector, and seeks primarily to the application of genetic algorithms (GA) to improve the Markowitz model (return - risk), The problem of portfolio optimization is a multi-objective problem that aims at simultaneously maximizing the expected return of the portfolio and minimizing portfolio risk. Present study is a heuristic approach to portfolio optimization problem using genetic algorithms technique.
The present study data on a sample of stocks price between 31/01/2010 -31/12/2010 in the Algerian stock Exchange. Further more in an attempt to evaluate the effectiveness of genetic algorithms to improve the level of risk optimization.
Keywords : Optimal portfolio, Diversification, Artificial intelligence, Genetic Algorithm, Return & Risk . (JEL) Classification : G11
Modern Portfolio Theory
1990 1952 1 (Métaheuristic) 2 1 2
3 Efficient portfolio 4 2010/01/31 -2010/12/31 1990 Diversification H. Markowitz,1952 . . Static Model Dynamic Model .
1 1.1 1952 3 Covariance) 4 2.1 Static Model 5 (2.1)…. ....
1 : / : e w R R w c s Vw w Min P 1 : / ; : ' ' ' e w k Vw w c s R w Max 1 (2.2) (2.3) 3.1 Dynamic Model Continuous-time
Regime Switching Models
6 ,1969 P.A.Samuelson N.H.Hakansson ,1971 ,1997 S.R.Pliska Optimal Dynamic Strategies H.Richardson, 1991 & D.Duffie M.Schweizer ,1996 2000 , Zhou & Li
Stochastic Linear Quadratic
7 1999
J.Yong & X.Y.Chou
Stochastic Differential Equations * Markov Chain Brownian Motion 8 ,1995
R. Korn & S. Trautmann
9
ij
ij n n n n V e R R R R R R w w w 1 1 1 1 ) 1 ,...., 1 ( ) ,...., ( ) ,...., ( ) ,...., (
RP
Engle & Ferstenberg , 2007
Garleanu & Pedersen, 2009
Transaction Costs 10 (2.4) (2.5) (2.6) (2.7) (2.8) : t : t Xt : t ft : t : G : pt : t . Artificial Intelligence
– 1970 Michigan University of Johon Holland
Optimization Problem 11 1993 Franklin Risto Karijalainen 1998 Herbert Dawid Michael Kope Sylvie Geisendorf 2000 Alfons Balmann Katrin Happe 2003 Pmar Keskinocak Feryal Erhun 2012 12 146 M. Garkaz 2011
S.Sefiane & M. Benbouziane
2012
1.2 13 Start Fitness Function . New Population Selection Parents Chromosomes (2.9) fi i n r ] 0 ، 1 [ r<p1 : r<pi < pi-1 Crossover Offspring Mutation Replacement
Test Stopping Creteria : R+
wi Monthly closing prices
31 / 01 / 2010 31 / 12 / 2010 w1 : SONELGAZ/14
w2 : Air Algérie w3 : Alg Telecom w4 : EGH EL AURASSI w5 : SAIDAL 1
RP
2 Fitness function (2.10) 3 Population The 50 50 4 . 5 Stopping Creteria 100 Current portfolio
Inputs R+ Arithmetic Crossover 13 4 Chromosome
Objective function value
Variance of Portfolio Mean Return of Portfolio Computing time 1 0,90612% 2,984% 0,3610% 3,5840 2 0,146695% 3,238% 0,4750% 3,2580 3 0,153187% 3,734% 0,5720% 3,1950 5 0,157452% 4,522% 0,7120% 2,7560 11 0,176732% 4,934% 0,872 % 2,5480 R+ 5 W5 W4 W3 W2 W1 0,1699% 0,1182% 0,1788% 0,4102% 0,1229% R+ GA fitness function Maximization 0 10 4 5 W1 = 12,29% : SONELGAZ/14 W2 = 41,02% : Air-Algérie W3 = 17,88% : Alg-Telecom W4 = 11,82% : EGH EL AURASSI W5 = 16,99% : SAIDAL 4,934% 0,872 %
Markowitz
1 01 / 2010 -12 / 2010 http://www.sgbv.dz 2 Portfolio weights Excel R+ 01 2010 SAIDAL EGH EL AURASSI Alg Telecom Air Algérie SONELGAZ/14 0,00067 0,0224 0,00016 0,00025 -0,00197 Mean 2,706 0,00068 0,000315 5,2 1,1454 Variance http://www.sgbv.dz 02 Var-Cov Matrix SAIDAL EGH EL AURASSI Alg Telecom Air Algérie SONELGAZ/14 1,146 1,036 -2,145 8,983 1,1454 SONELGAZ/14 1,146 1,036 -3,938 5,2 8,982 Air Alegrie -1,0489 0,00016 0,000315 -3,94 -2,145 Alg Telecom -1,852 0,00068 0,00016 1,097 0,00068 EGH EL AURASSI 2,7062 -1,852 -1,0489 -2,51 1,1464 SAIDAL
http://www.sgbv.dz 03 http://www.sgbv.dz 1 . » « 2008
R. J. Kuo, and C. W. Hong" , Integration of Genetic Algorithm and Particle Swarm Optimization for Investment Portfolio Optimization", Applied
Mathematics & Information Sciences An International Journal ; Sci. 7, No. 6, PP :2397-2408 .
2 . 2 2014 117 -136 . 3 . 1998 4 . » « 2 1989
5. Jean-Luc Prigent, « Portfolio Optimization and Performance Analysis », Financial -Mathematics Series, Chapman & Hall/CRC is an imprint of Taylor & Francis Group, U.S, 2007, PP: 70-78.
6. James D.Hamilton, « Regime-Switching Models», Palgrave dictionary of Economics, USA, 2005.
7. Min Dai , Zuo Quan Xu, Xun Yu Zhou, “Continuous-Time Markowitz’s Model with Transaction Costs”, 2009 .
www.math.nus.edu.sg/~matdm/mv-transaction6.pdf (16/10/2012). 8 . 2011 108 -120 .
9. Xun Yu Zhou and G.Yin ,”Markowitz’s Mean-Variance Portfolio Selection with Regime Switching : A Continuous-Time Model “ ,March 2006,
PP:5-11.
10. Esben Hedegaard, “Robust Dynamic Asset Allocation With Imperfect Predictors”,November ( 2011) , PP:4-6 . web Site: www.people.stern.nyu.edu/ehedegaa/PDFs/RobustDynamicAssetAllocation.pdf (16/10/2012).
11. A.K.Misra, « Portfolio Optimization of Commercial Banks - An Application of Genetic Algorithm », European Journal of Business and
Management , 2013, Vol.5, No.6, P 120.
12 . 21 2012 304 -315 . . (Stochastic Process) .
13. Slimane Sefiane,Mohamed Benbouziane, “Portfolio Selection Using Genetic Algorithm”,Journal of Applied Finance & Banking, 2012, vol.2,
no.4.
14. Luca Scrucca, "GA: A Package for Genetic Algorithms in R", Journal of Statistical Software, 2013 ,Volume 53, Issue 4. http://www.jstatsoft.org/ . Correlation.Matrix SAIDAL EGH EL AURASSI Alg Telecom Air Algérie SONELGAZ/14 0,20591 0,11683 -0,35691 0,36805 1 SONELGAZ/14 -0,21122 0,58071 -0,00308 1 0,36805 Air Alegrie -0,00359 0,34610 1 -0,00308 -0,35691 Alg Telecom -0,04297 1 0,34610 0,58071 0,11683 EGH EL AURASSI 1 -0,04297 -0,00359 -0,21122 0,20591 SAIDAL