Dynamic Portfolio Using Association Rules

We provide dynamic portfolio strategy capitalizing on the time-varying association rules. We set the moving observation window and employed changes in the number of negative rules as a signal for portfolio weights. The back-testing result shows that increasing number of negative rules reflects newly formed negative linkages between Korea stock market and related variables. The result suggests that as a proxy for systemic risk measure, signal from association rules can be an effective tool for portfolio risk management.

1. Introduction

2. Data and methodology

3. Test Results

4. Conclusions

References

(0)

(0)

(0)

(0)