Özyeğin University, Çekmeköy Campus Nişantepe District, Orman Street, 34794 Çekmeköy - İSTANBUL

Phone : +90 (216) 564 90 00

Fax : +90 (216) 564 99 99

E-mail: info@ozyegin.edu.tr

Aug 06, 2024 - Aug 13, 2024

Thesis Defence – Gülşah Bıçkı (MSFE)

Gülşah Bıçkı - M.Sc. Financial Engineering 

Asst. Prof. Özcan Ceylan – Advisor

Date: 13.08.2023

Time: 15.00

Location: Özyeğin University, Çekmeköy Campus - AB2.345

 

"Enhanced Dynamic Multi-Period Portfolio Optimization: Integrating Advanced Transaction Cost Functions"

 

Asst. Prof. Özcan Ceylan, Özyeğin University

Prof. Taylan Akdoğan, Özyeğin University

Asst. Prof. Yasin Kütük, Gebze Technical University

Abstract:

This study addresses the often-overlooked impact of transaction costs in multi-period portfolio optimization. Traditional models underestimate transaction costs such as bid-ask spreads, market impact costs, and commission fees by assuming these costs are fixed, while in reality, they are time-varying. Using 13 years of data from the most liquid ETFs from January of 2012 to May of 2024, we identify the variables  that best predict bid-ask spreads such as the VIX and trading volume, and integrate these variables  into advanced forecasting methods like GARCH-X, ARIMA-X, and SARIMA-X. We then compare the forecasting performance of these models using MSE, RMSE, and MAE metrics to select the best-performing model. The best forecasting results are integrated into our multi-period portfolio optimization framework, based on Boyd et al., using the Model Predictive Control (MPC) technique.

Our findings show that accurately modeling transaction costs is crucial for effective portfolio optimization. By incorporating time-varying factors, we significantly enhance traditional optimization techniques. The effectiveness of these techniques is rigorously evaluated through backtesting and advanced performance metrics.

In conclusion, this study highlights the importance of detailed transaction cost modeling in developing effective investment strategies and demonstrates how advanced methods can improve portfolio management and adapt to alternating market conditions.

Keywords:

Portfolio optimization, transaction costs, bid-ask spread, market impact, financial markets, Model Predictive Control (MPC), ETFs, GARCH-X, ARIMA-X, SARIMA-X, moving averages, investment strategies, market volatility.

Bio:

Gülşah Bıçkı received her MSc degree from Boğaziçi University Faculty of Science Institute,Istanbul in 2013. She has 10 years of experience as a sales director in the private sector and has been working as a finance expert in financial engineering at Özyeğin University for the last 3 years. This thesis focuses on the optimization of transaction costs in financial markets.