題目：Leveraged Portfolio Selection under Liquidity Risk: Model, Theory,

and Computation

主講人：Chanaka Edirisinghe, 美國倫斯勒理工學院 講席教授, Lally管理學院學術副院長

主持人：陳靖楠 副教授

時間：

2018年7月9日19:00-21:00

地點： 北航新主樓A座618

摘要：

When a financial portfolio is rebalanced under market conditions to

satisfy leverage and other restrictions, asset illiquidity adversely-impacts

trading prices, and hence, the portfolio's performance. Using a continuous-time

trading model, we study the Pareto-efficiency between risk-adjusted return,

leverage, and target return. We show analytically that the Sharpe-maximizing

unlevered portfolio is no longer a tangency portfolio, and

proportionate-leveraging is not an optimal strategy under liquidity risk. As

target return increases, the required minimum portfolio-leverage increases at

an increasing-rate, while the Sharpe-Leverage frontiers are

progressively-dominated. These results contrast with the classical portfolio

theory that assumes no liquidity risk, and our empirical analysis using ETF

asset-data verifies that ignoring liquidity impact may lead to severe portfolio

under-performance.

If time permits, I will also consider a specific situation involving only

de-leveraging, where the model is simplified to maximize portfolio’s expected

value under leverage and margin limits. This leads to a separable model, but it

is extremely difficult to solve due to non-convexity. I will present a new and

general dual cutting plane technique that solves the Lagrangian dual

more-efficiently. The sensitivities of the optimal deleveraging strategy to

leverage and margin limits will be discussed in the context of the above data

set.

當調整投資組合以滿足杠桿和其他限制時，資產流動性對交易價格會產生不利影響，進而影響投資組合的表現。運用連續時間的交易模型，我們研究風險調整收益率、杠桿率和目標收益率之間的帕累托效率。我們的解析結果表明，夏普率最大化的無杠桿投資組合不再是切向投資組合，在流動性風險下按比例加杠桿也不再是最優策略。隨著目標收益的增加，所需的最低投資組合杠桿率以遞增的速度增長，而夏普杠桿前沿逐漸被主導。這些結果與忽略流動性風險的經典的投資組合理論不同，并且我們基于ETF資產數據的實證分析證實忽略資產流動性會嚴重影響投資組合表現。

如果時間允許，我還會考慮一個特定的情況，只涉及去杠桿化，其中模型被簡化以最大化投資組合在杠桿率和保證金限制下的預期價值。這導致了一個可分離的模型，但是由于其非凸性，求解非常困難。我將提出一個新的一般的雙切割平面技術，更有效地解決拉格朗日對偶問題并討論最佳去杠桿化策略對杠桿率和保證金限制的敏感性。

主講人簡介：

Dr. Chanaka Edirisinghe holds a BS (Mechanical Engineering), an M.Eng

(Industrial Engineering and Management), and a Ph.D. (Management Science) from

University of British Columbia, Canada. He has published extensively in

operations research and finance, focusing on quantitative finance topics, as

well as stochastic and quadratic optimization. His research appears in

Management Science, Operations Research, Mathematical Programming, Mathematics

of Operations Research, as well as in Journal of Financial and Quantitative

Analysis, Journal of Banking and Finance, and Quantitative Finance, among

others. He received the Citation of Excellence Award by Emerald Management Reviews

in 2009 for publishing one of the top 50 management research articles in the

world. He was a former Vice Chair of Financial Services Section, as well as

Optimization Society of INFORMS, and he was the General Chair of the INFORMS

2016 annual conference.

Chanaka Edirisinghe教授于加拿大英屬哥倫比亞大學獲得管理科學博士學位、工程學碩士學位及工業機械工程學士學位，并在運籌學和金融學領域發表了大量文章，他專注于量化金融、隨機和二次優化。他在Management Science,

等雜志上發表過文章，并于2009年獲得Emerald Management Reviews頒發的Citation of Excellence Award。他曾擔任美國運籌與管理科學協會（INFORMS）金融服務分會和優化分會副主席，并且擔任INFORMS 2016年會的大會主席。

編輯：宋超