利用馬可夫鏈模型挑選行為股票之投資組合最佳化

Behavioral Stock Portfolio Optimization With Complementary Markov Chain Process

劉建志(Chien-Chih Liu)

指導教授 ： 張國華

中原大學/電機資訊學院/工業與系統工程研究所/碩士(2018年)

https://doi.org/10.6840/cycu201800725

摘要

隨著低利率時代以來臨，投資人為了追求較高的報酬率，並不會將資金存入銀行放著，而是開始把資金投入股票市場中，然而股票市場行情變換迅速，真正獲利的投資人並不多，只有在適當時機點進場交易的投資者才能從中獲利。本研究利用馬可夫鏈探討行為股票，投資人在投資時，有許多不確定因素造成投資人在做決策時不與一般常理來的一樣，產生不理性的行為，所產生的不理性行為有可能造成股價的變動，而這些受到不理性行為影響的股票我們稱之為行為股票；這些不理性的行為在其他研究利用操作型定義來觀察，依據大量觀察的數據，我們可以找出何時哪一支股票何時開始受到不理性的行為的起因，在經過多長的時間影響該股票的報酬率轉為正報酬率與它相對的會發生這影響的機率。
本研究觀察行為股票所發生的起因、發生的效應相對機率及時間，再使用馬可夫鏈計算明天股票上漲機率，使投資者能夠更準確的選擇股票並且得到相對高的報酬，最後將挑選出的行為股票放入Safety-first投資組合模型中，利用Python混合型整數規劃執行最佳化，找出投資組合最佳化，使報酬可以贏過市場大盤。經本研究證實，利用馬可夫鏈的機率挑選明天報酬率上漲的行為股票，最後獲得到的累積報酬率勝過於大盤累積報酬率3倍。

關鍵字

行為股票 ； 投資組合最佳化 ； 馬可夫鏈

並列摘要

Due to low-interest-rate, investors choose to put their money into the stock market instead of depositing it in banks. However, market information changes rapidly so only few investors can analyze them on time so only few investors also earn profit. This study focuses on behavior stocks (B-stocks), which are stocks subjected to irrational behaviors of investors caused by uncertainty such as emotional decision that can simultaneously affect the price movement of the stocks. We can use the B-stocks from established operational definitions of irrational behaviors by other researchers by studying the occurrence of a cause-and-effect relation, the time between them and the corresponding probability.
Considering the established B-stocks, we can use them as the investment pools. Before doing so, to support the probabilities of such B-stocks to have the desired effect, we also study the Markov chain processes within these B-stocks that results to positive returns. Those identified B-stocks that also have a desired transition probability of having positive returns are considered as the new investment pool. Using this investment pool, we select the portfolio considering the Safety-first portfolio model by optimization through Python mixed-integer programming. From the results, the cumulative return rate of the portfolio considering B-stocks over the test period outperforms its market portfolio counterpart.

並列關鍵字

Behavioral stock ； Portfolio optimization ； Markov chain

參考文獻

[1] Amini, S., B. Gebka, R. Hudson and K. Keasey (2013). "A review of the international literature on the short term predictability of stock prices conditional on large prior price changes: Microstructure, behavioral and risk related explanations." International Review of Financial Analysis 26: 1-17.

[2] Bauerle, N. and U. Rieder (2004). "Portfolio optimization with Markov-modulated stock prices and interest rates." IEEE Transactions on Automatic Control 49(3): 442-447.

[3] Bondt Werner F, M. and R. Thaler (1985). "Does the Stock Market Overreact?" The Journal of Finance 40(3): 793-805.

[4] Boussaidi, R. (2013). Representativeness Heuristic, Investor Sentiment and Overreaction to Accounting Earnings: The Case of the Tunisian Stock Market.

[5] Brinson, G. P., L. R. Hood and G. L. Beebower (1995). "Determinants of portfolio performance." Financial Analysts Journal 51(1): 133-138.