Ying Becker | Brandeis University (original) (raw)

Papers by Ying Becker

Financial Analysts Journal, Nov 1, 2012

Financial Analysts Journal, May 1, 2012

Pacific-Basin Finance Journal

Financial Analysts Journal, 2010

... Didier Rosenfeld, CFA Managing Director Head of the EAFE & Global Active Equity Strat... more ... Didier Rosenfeld, CFA Managing Director Head of the EAFE & Global Active Equity Strategies State Street Global Advisors ... F. Bryson, Otgontsetseg Erhemjamts, Atul Gupta, Henry Marigliano, Kartik Raman, Anya Suvorov, Anand Venkateswaran, Roy Wiggins, Mike Yang, and ...

Applied Catalysis A: General, 1993

Genetic Programming Theory and Practice V

Active quantitative investment strategies have been gaining more traction among asset managers an... more Active quantitative investment strategies have been gaining more traction among asset managers and institutional investors. At the core of quantitative investment strategy is a quantitative model that could effectively rank the assets based on the likelihood of their excess returns against a relative benchmark, or sometimes simply forecast assets' future returns. To achieve this, the model extracts and utilizes the key characteristics of the complex reality to accurately reflect the interested behavior. Thus, the biggest challenge of developing a stock selection model is how to find the most efficacious factors and their interactions from large amounts of financial data and market information. Our previous study has demonstrated that genetic programming is an effective technique for selecting factors and constructing multi-factor models for ranking stocks. By constructing various "simple objective" fitness functions based upon investment philosophy specific with respect to risk, we are able to use the GP algorithm to develop more powerful stock selection models than with the traditional methods. However, the empirical results have shown that such developed GP model does not guarantee a well-balanced performance that factors multiple investment criteria a portfolio manager would consider. The research on GP theory and practice has shown that the issue might be addressed with multiple objective functions. In addition, the multiple objective functions might result in a more robust model to deal with the data overfitting issue. In this study, we implement and evaluate three simple multi-objective algorithms to simultaneously optimize the information ratio, information coefficient, and intra-fractile hit rate of a portfolio. These algorithms, the constrained fitness function, sequential algorithm, and parallel algorithm take widely different approaches to combining these different portfolio metrics. The results show that the multi-objective algorithms do produce well-balanced portfolio performance. The constrained fitness algorithm performs much better than the sequential and parallel multi-objective algorithms. In particular, the constrained fitness function algorithm produces formulae that have better test statistics than that of any of the single fitness function algorithms. Moreover, this algorithm generalizes to the held-out test data set much better than any of the single fitness algorithms. The model performance diagnosis also shows that the GP models resulting from the constrained fitness function and the one from the sequential algorithms are the best. These two models complement each other in different financial market environments.

SSRN Electronic Journal, 2006

The ecologically enhanced versions of the quantitative strategy were superior to those of the bas... more The ecologically enhanced versions of the quantitative strategy were superior to those of the baseline portfolio, although all the active strategies exceeded the benchmark for the time frame tested, even after accounting for transaction costs and incorporating style constraints. It was found that the information ratio on the portfolio is most significantly improved when the extreme ratings are incorporated into the process. The results of this analysis support the assertion that sensitivity to environmental issues, particularly for the extreme performers,may enhance returns of an active strategy over time.

Econometric Modeling: Capital Markets - Portfolio Theory eJournal, 2018

Quantitative equity management is concerned with rigorous, disciplined approaches to help investo... more Quantitative equity management is concerned with rigorous, disciplined approaches to help investors structure optimal portfolios to achieve the outcomes they seek. At the root of disciplined, modern investment processes are two things: risk and return. The notion of total return is obvious—price appreciation plus any dividend payments. Risk may not be so straightforward. In most quantitative approaches, risk is viewed as more akin to a roulette wheel; that is, the possible outcomes are well specified and the likelihood of each outcome is known, but in advance, an investor does not know which outcome will be realized. In this piece, we curate the history of quantitative equity investing, which traces its origins to the development of portfolio theory and the capital asset pricing model (CAPM). In equities, some of the first quantitative approaches were aimed at confirming the theoretical predictions of the CAPM. In particular, the expected return of a risky asset depends only on the ...

Advances in Pacific Basin Business, Economics and Finance

Key Engineering Materials

ABSTRACT

Genetic Programming, 2000

This is a narrative describing the implementation of a genetic programming technique for stock pi... more This is a narrative describing the implementation of a genetic programming technique for stock picking in a quantitatively driven, risk-controlled, US equity portfolio. It describes, in general, the problems that the authors faced in their portfolio context when using genetic programming techniques and in gaining acceptance of the technique by a skeptical audience. We discuss in some detail the construction of the fitness function, the genetic programming system's parameterization (including data selection and internal function choice), and the interpretation and modification of the generated programs for eventual implementation.

Gecco, 2009

We provide an overview of using genetic programming (GP) to model stock returns. Our models emplo... more We provide an overview of using genetic programming (GP) to model stock returns. Our models employ GP terminals (model decision variables) that are financial factors identified by experts. We describe the multi-stage training, testing and validation process that we have integrated with GP selection to be appropriate for financial panel data and how the GP solutions are situated within a

SSRN Electronic Journal, 2000

There have been numerous studies to date on the performance effects of incorporating socially res... more There have been numerous studies to date on the performance effects of incorporating socially responsible criteria into investment portfolios. While many studies point to out- performance of socially responsible investments (SRI) over time, it has been a challenge to prove conclusively that incorporating ethical criteria enhances performance. Some researchers have uncovered flaws in the earlier studies. This research explores whether

Proceedings of the first ACM/SIGEVO Summit on Genetic and Evolutionary Computation - GEC '09, 2009

Genetic and Evolutionary Computation, 2009

Genetic programming has been shown to be an effective technique for devel- oping quantitative mod... more Genetic programming has been shown to be an effective technique for devel- oping quantitative models to rank stocks with respect to their future returns. Stock selection models derived via genetic programming can often outperform models based on more traditional linear factor combination methods. However, genetic programming also poses an inherent challenge when applied to financial markets: its powerful optimization capability

Journal of Asset Management, 2004