بهینه سازی سبد سرمایه گذاری با استفاده ازمدلهای dcc ،cccو الگوریتم مارکوئیتز :شواهدی از بورس اوراق بهادار
محورهای موضوعی : فصلنامه اقتصاد محاسباتی
زهرا قربانی
1
,
علیرضا دقیقی اصلی
2
,
مرجان دامن کشیده
3
,
رویا سیفی پور
4
1 - گروه آموزشی اقتصاد، دانشکده اقتصاد و حسابداری، دانشکاه آزاد اسلامی واحد تهران مرکز
2 - استاد یار دانشگاه آزاد واحد تهران مرکز
3 - مدیر گروه تحصیلات تکمیلی، دانشکده اقتصاد و حسابداری، دانشگاه آزاد اسلامی واحد تهران مرکزی
4 - استادیار، عضو هیئت علمی، گروه اقتصاد، واحد تهران مرکز، دانشگاه آزاد اسلامی ، تهران ، ایران
کلید واژه: بهینه‌سازی سبد سهام, الگوریتم مارکویتز, گارچ چند متغیره, بازار سرمایه, ریسک,
چکیده مقاله :
مقاله حاضر به بررسی عملکرد و مقایسه مدل های گارچ چند متغیره و الگوریتم مارکویتز در بهینه سازی سبد سرمایه گذاری برای سهام‌های برتردر چهار صنعت منتخب، شامل صنایع منتخب ماشین آلات برقی، استخراج کانه های فلزی، خودرو و ساخت قطعات و فرآورده های نفتی،که دارای بازدهی و ریسک متغیر هستند، برای سال‌های 1395-1399 می‌پردازد. براساس نتایج بهینه سازی پویا و میانگین گیری از متوسط اوزان بهینه این چهار صنعت در هر سه مدل، وزن بالاتر به سهام صنایعی اختصاص یافته است که نوسانات کمتری در بازدهی شان وجود دارد. در واقع، اوزان کمتر در بین چهار صنعت به صنایع با نوسانات شدیدتر در بازدهی یعنی صنایع خودرو و ساخت قطعات و فرآورده های نفتی اختصاص دارد. برعکس بیشترین سهم متوسط بهینه از سبد تشکیل یافته در بین چهار صنعت به صنعت کانی های فلزی با کمترین نوسانات در بازدهی تعلق دارد. لذا با توجه به نتایج حاصل شده، هر سه مدل نتیجه یکسانی را برای هر چهار سبد نشان می دهند. لذا در راستای تنوع بخشی به سبد سرمایه گذاری و کنترل ریسک سرمایه گذاری، به سرمایه گذاران توصیه می گردد همبستگی بین روند بازدهی سهام و نوسانات بازدهی سهام دارایی های مختلف قابل نگهداری را مدنظر قرار دهند.
Extended Abstract This study investigates the impact of the capital market using multivariate GARCH models and the Markowitz algorithm to optimize the stock portfolio. The statistical population of this research includes stock exchange companies that were admitted to the stock exchange before 1395 and were active until the end of 1399 and had the following characteristics: The financial year of the companies should have ended on March 20th and the companies' shares should have been traded on the stock exchange during each year of the research period and the end-of-period price was available. In addition, the financial information of the companies must also be available. Considering the above characteristics, 4 top industries, including the automotive and parts manufacturing industry, the selected electrical machinery industry, the metal mining and oil products industry, were selected as the screening population in our portfolio based on a combination of stock liquidity, stock trading volume in the trading hall, stock trading frequency in the trading hall, and the company's impact on the market. The sample size is 800 and is daily during the period from 1395 to 1399. Purpose The results of this study show that the optimal weights are more allocated to stocks with less volatility in the stock return trend of that industry. In fact, lower weights are allocated to industries with more volatile returns among the four industries, namely the automotive and parts manufacturing and oil products industries. Conversely, the largest optimal average share of the portfolio among the four industries is for the non-metallic minerals industry with the least return volatility. Methodology The results of this study also show that industry stock return shocks have reciprocal effects on each other. For example, a positive shock to the stock return of the non-metallic minerals industry leads to a negative shock to the stock return of the automotive and parts manufacturing industry. In addition, the results of this study show that the CCC and DCC models have different results in estimating the optimal weights of the industries and risk-free assets that make up the investment portfolio. So that, the DCC model, compared to the CCC model, allocates less weight to the stocks of the automotive and parts manufacturing and oil products industries and, conversely, allocates more weight to the stocks of the non-metallic minerals industry. Finally, the results of this study show that the portfolio formed using the Markowitz optimization algorithms can track the risk-averse individual's utility to maximize profit. And Based on the results of this study, it is suggested that investors pay attention to the volatility of the stock return of that industry when selecting stocks for investment and allocate a greater share to stocks of industries with less return volatility. Finding It is also suggested that DCC models be used alongside CCC models to estimate the optimal weights of the investment portfolio. In addition, it is suggested that Markowitz optimization algorithms be used to form an investment portfolio that matches the risk-averse individual's utility. Now, let’s address the limitations of this study, that one of the limitations of this study is the use of daily stock return data. It is suggested that in future research, data with higher frequency such as hourly or minute data be used. Another limitation of this study is the non-consideration of other factors affecting stock returns, such as macroeconomic factors. It is suggested that in future research, these factors should also be considered. Conclusion The results of this study have important implications for investors and portfolio managers. The use of multivariate GARCH models and the Markowitz algorithm can help to optimize stock portfolios and improve risk-adjusted returns. Investors should consider the volatility of stock returns and the correlation between industries when making investment decisions. DCC models can be used to estimate optimal portfolio weights, and Markowitz optimization algorithms can be used to form portfolios that match the risk-averse individual's utility. Future research should focus on using higher frequency data and considering other factors affecting stock returns.
