Markowitz

January, 2013

Naïve Strategy vs Markowitz Model

EDHEC students:

Laurie, Xianghui, Kathleen, Yuchen, Monalisa

Etude de Satisfaction – CCI du Var - Tout droit de reproduction réservé à EDHEC Junior Etudes. Cette proposition est valable 3 mois.

1

1

Summary

Overview & Objectives Naive Strategy Methodology Markowitz Strategy Methodology Results & Conclusion

Etude de Satisfaction – CCI du Var - Tout droit de reproduction réservé à EDHEC Junior Etudes. Cette proposition est valable 3 mois.

2

2

Objectives

Naïve strategy Model

Markowitz Model

Conclusion

• The objective of the assignment on Naïve Strategy vs Markowitz model is to find that of selected sample-based strategies on given datasets which model is superior statically. • Basically we want to find out if the Markowitz model can beat a naïve strategy • In the article “Optimal Versus Naive Diversification: How Inefficient is the 1/N Portfolio Strategy” by DeMiguel et al. (2007), they address the out-ofsample performance of selected sample-based strategies on U.S. datasets. Their striking results : They tested 14 models based on the Markowitz model and found that no one can beat statistically a naïve portfolio strategy on different markets !!!! • Our goal : to check their results with two models : the naïve strategy and the sample based Markowitz model (MV)

3

Objectives

Naïve strategy

Markowitz Model

Conclusion

For the naïve strategy approach, we calculate the mean of each asset’s monthly return from July, 1963 to July, 2011.

Naïve strategy Model

Markowitz Model

Conclusion

The weight of each of the ten assets is equally 10% of portfolio. Summing up the mean of each asset and then divided by to calculate the rate of return of the portfolio, we get the rate of return of the portfolio

E(R)= 0.94%

4

Objectives

Naïve strategy

Markowitz Model

Conclusion

For Markowitz Model, we divide the dataset into sets with 60 samples. We run...