EXPERIENCE AND EXPERTISE:

• Research department lead on statistically based bond pricing, idea generation and implementation for predictive models, design of cross - departmental IT systems to perform automated predictive scoring.
• Analyzing mortgage prepayment models - splined, multivariate survival models of mortgage loan events based on borrower, loan, and macroeconomic variables. Data and calibration of model are disparate and complex.
• Quick, ad-hoc, practical analysis of datasets and generation of summary write-ups daily, weekly, monthly.
• Estimation of variance components from empirical data using quantitative genetic techniques, for example least squares and maximum likelihood methodology.
• Estimation of genotype by environment interactions, including 'special environmental effects' (maternal effects, other indirect second-order effects), performing mixed (fixed & random) effects analyses, QTL analysis.
• Modelling to examine the influence of social interactions between organisms. Published work indicating profound implications for crop biology and genetics. Specialty in non-additive variance (interaction effects in complex systems) and deducing implications for predictive models.
• Developing proprietary algorithm for effects estimation and analysis using SAS/Base, STAT, SQL, and Macro. Incorporated permutation test to generate distributions for hypothesis testing.
• Using multivariate analysis in the identification of adaptive traits including various regression techniques (SAS-based, Mathematica-based solution designed around system of differential equations), structural equation modelling (AMOS). Theoretical and experimental work.
• Use of R, S-plus and SAS to perform analysis of ordinal and nominal data.
• Data handling skills - statistical and coding expertise in SAS, C++, C, XML, SQL in an industry environment. Conducting simulation studies (C++, Markov-Chain Monte-Carlo) examining evolution of genetic effects over environmental changes, presenting work at international conference.
• Managing several concurrent experimental protocols with international collaboration, and interacting with investigators and support staff to manage workloads on a regular basis.
• Writing and reviewing of scholarly papers and book chapters in quantitative genetics, providing methodological assistance for statistical analysis in neuroscience. Grant writing independently and collaboratively.

EMPLOYMENT:

Lead Quantitative Researcher

Responsibilities:

• Electronic trading system. Development of a Java based asynchronous electronic pricing system for mortgage bonds using market-driven model results and trader inputs.
• Driving collaboration with trading, market risk, and IT. Arranging the components and groups needed to make projects succeed (data sources, collaborative structures, human resource). Delivering efficiently by focusing strategy and fostering cooperation.
• Mentoring juniors. Hiring, mentoring, developing domain and quantitative knowledge as well as bringing interns and junior team member to understand enterprise-level work. Help junior quantitative researchers navigate the transition from academia to finance.
• Active communicator. D elivering results and analyses to management, investors, vendors, and the trading desk.

Quantitative Researcher

Responsibilities:

• Statistical attribution analysis. Use of 1010data, R, Tableau, as well as writing C++ algorithms. Elucidating causal effects of refinancing and default of mortgage loans. Various regression techniques and proprietary pseudo-orthogonalization techniques. Engineering bonds using predictive data mapping techniques.
• Model tuning. Optimization in Python to adjust model parameters. Algorithm also used for ‘ reverse model ’ that deduces traders ’ implied perception of the market ’ s direction. Data link across systems.
• Economic and market scenario analysis. Scenario analysis for portfolios of residential mortgage loans. What if analyses of home price appreciation and unemployment scenarios. Segmentation modelling using loan level, economic, and market data. Reporting for clients, trading desk, and management via Python scripted process.
• Model validation. Validation of third party models, monitoring of parameterization and calibration. Adjusting model parameters after deficiency analysis.
• Liaison with internal and external groups. Attending industry conferences to gain understanding of technical, economic, and regulatory factors. Contract negotiation. Direct work with other quantitative modelling staff at houses to enhance areas of concern in models that BNP uses.

Analytical Developer

Responsibilities:

• Development of predictive algorithms and analytic libraries. C++ libraries linked to Excel, digesting underlying SQL and raw text data sources. is.
• Developer on C++ system. Pricing and risk generation on a distributed processing farm. Data capture and management systems and static data storage; use of python, SQL, C, C#, XML, R, SAS, JMP.
• Collaborate with research group for statistical arbitrage. Statistical models based on market and economic data.
• Estimating variance components using least squares and maximum likelihood; performing mixed (fixed and random) effects analyses using SAS.
• Parameter estimation of correlated independent variables (canonical correlation) with permutation tests. Intensive SAS usage, some C++, and SQL for data handling and screening. Use of SAS/Stat, /SQL and /IML.
• Data screening, manipulation of data sets across systems/formats for multiple analyses, transformations including normalization and first order linear transformations (based on the specific data set moments).

Selected Conferences and Seminars

• Fundamentals of Deep Learning, with Applications - Confidential, Chief Data Scientist, Confidential. NYC.
• Polyglot Data Scientist & Learning Algorithms, Neural Turing Machines - Confidential, Confidential. NYC.
• Survey of Bayesian Methods Considering Dynamic Linear Models - NYC.
• Augmenting Simple Parametric Models with Machine Learning - Confidential. NYC.
• CoreLogic Risk Summit: Three days of Modelling, Analysis, and Policy. Dana Point, CA.
• DIY "Medium Data" Analysis - Confidential. NYC.
• Summer Institute in Statistical Genetics -Confidential, NC