應(yīng)偉德國際1946bv官網(wǎng)的邀請，香港大學(xué)Waiki Ching教授于2024年10月19日上午9點在中關(guān)村校區(qū)主樓317會議室做了題為《On Adaptive Online Mean-Variance Portfolio Selection Problems》的學(xué)術(shù)報告。報告會由郭思尼老師主持，學(xué)院眾多師生參加了本次報告會。
程教授圍繞在線投資組合選擇問題展開深入闡述，聚焦市場快速變化情境下如何借助先進技術(shù)和模型提升投資者決策的準(zhǔn)確性。他指出，傳統(tǒng)投資組合理論如Markowitz的均值-方差模型雖構(gòu)建了理想的風(fēng)險與收益平衡框架，但在復(fù)雜多變的市場環(huán)境中顯得力不從心。為此，程教授引入適應(yīng)性技術(shù)，提出兩種創(chuàng)新模型，為在線投資組合選擇提供了強有力的支持。
程教授首先介紹了在線投資組合選擇問題的背景，強調(diào)投資者需根據(jù)市場實時信息動態(tài)調(diào)整資產(chǎn)配置，而短期內(nèi)精確預(yù)測未來資產(chǎn)回報并合理規(guī)避風(fēng)險是當(dāng)前研究的難點。針對這一問題，他提出了適應(yīng)性在線移動平均方法（AOLPI），通過結(jié)合歷史數(shù)據(jù)和同行資產(chǎn)影響，動態(tài)調(diào)整資產(chǎn)預(yù)測中的衰減因子，提高預(yù)測準(zhǔn)確性。實驗證明，AOLPI在多種市場環(huán)境下顯著優(yōu)于傳統(tǒng)簡單移動平均（SMA）和指數(shù)移動平均（EMA）方法。隨后，程教授介紹了適應(yīng)性均值-方差模型（AMV），該模型綜合考慮風(fēng)險與收益的平衡，通過動態(tài)更新協(xié)方差矩陣捕捉資產(chǎn)間的風(fēng)險關(guān)聯(lián)，并通過風(fēng)險偏好參數(shù)靈活調(diào)整投資策略。AMV使投資者能夠根據(jù)最新市場數(shù)據(jù)動態(tài)管理風(fēng)險，特別在高風(fēng)險市場中展現(xiàn)顯著優(yōu)勢。
程教授通過多組實際市場數(shù)據(jù)集驗證了AOLPI和AMV模型的有效性，并展示了兩者結(jié)合形成的AOLPIMV算法在提升投資組合優(yōu)化效果方面的卓越表現(xiàn)。該算法在交易成本考慮下仍能保持較高回報，優(yōu)于傳統(tǒng)在線投資組合選擇方法。程教授的研究為在線投資組合選擇提供了新的技術(shù)工具，幫助投資者在復(fù)雜的市場環(huán)境中做出更科學(xué)的決策，同時推動了金融科技領(lǐng)域的進一步發(fā)展。
報告結(jié)束后，程教授和與會師生展開了積極的討論交流。報告反響熱烈，得到了師生們的一致好評。
匯報人簡介：
Professor Waiki Ching is a distinguished academic at the University of Hong Kong, well-known for his contributions to stochastic modeling, financial mathematics, and computational biology. He has extensive expertise in areas such as matrix computations, operations research, and quantitative finance. His research focuses on applying mathematical techniques to solve real-world problems in finance and biology, including portfolio optimization, risk management, and biological network analysis. Professor Ching has published a wide array of peer-reviewed papers in top journals and has been awarded numerous research grants from prestigious funding bodies such as the Hong Kong Research Grant Council.
He made significant advancements in both finance and biology through the development of algorithms and models. His work in financial mathematics includes dynamic portfolio selection, risk management, and online investment strategies. One of his key contributions is in adaptive online portfolio optimization, where he has developed innovative models such as the adaptive mean-variance model and adaptive online moving average methods, which have been successfully applied to real-world investment scenarios. These methods improve investment decision-making by balancing risk and return in dynamic, high-frequency trading environments. In computational biology, Professor Ching has focused on analyzing biological data and modeling complex systems like gene regulatory networks. He applies matrix computation techniques to study large-scale biological networks, helping to understand how biological systems function and evolve.
