We would like to search for a Statistician / Econometrician / R programmer with good statistical/econometrician knowledge, modeling/modeling diagnostic experience, and statistical research experience, to develop the programs and consult on model and experiment design for our academic statistical research on hedge funds.
We would be analyzing the influencing factors that affect the performance, risk, and life cycle(hazard rate) of hedge funds using regression analysis (including multivariate regression, time series regression, Cox regression), difference-in-difference regression, heckman 2 stage model, Regression Discontinuity Designs. Doing causal inference and using identification strategies to setup appropriate experiment/tests that establish causal claims, and avoiding issues such as reverse causality, simultaneity, omitted variable, and endogeneity in general. The position’s responsibilities and qualifications are described below.
Work Period: From Immediately to 01/15
Pay: Total project pay is determined by your availability and hourly rate. We would do flat rate only for the entire project (the budget amount would not be a good reference because it depends on the freelancer).
Responsibilities
- Model Consultation and Model diagnostic:
-Consult on procedures for specifically detecting the presence of and treating the endogeneity issues of reverse causality, simultaneity, omitted variable, and endogeneity in general
-Consult on finding suitable identification strategies for establishing claim for causal relationship for what we are looking to test
-Consult on modeling design based on what we are looking to test and our data. –
-Consult on methodology or model diagnostic procedure that we should use to make sure our models/analysis/experiments are robust, and the model parameters are specified to our analysis, and properly adjusted for factors that may affect our model/results (such as to avoid multicollinearity, endogeneity, Heteroscedasticity, biase, distribution)
- Programming/model implementation/data processing:
-programming in R to code our models and variables, data processing, and various other programming, data, statistics related tasks.
- Models implementation: Replicating or applying models and procedures from reference papers to our research
- Description of procedures: Provide detailed and technical description of
the methodology used models and implementation
Qualifications
- Master’s or PhD’s degree in Statistics/Econometric that used R significantly
- Minimum 4 years of experience implementing programming using R (must),
preferably statistical models
- Have experience in Causal inference and using identification strategies to setup appropriate experiment/tests that establish causal claims
- Have practical experience in implementing tests for detecting the presence of and methods for treating the endogeneity issues of reverse causality, simultaneity, omitted variable, and endogeneity in general
- Have experience in applying statistical methodologies procedures to ensure robustness (such as to avoid multicollinearity, endogeneity, Heteroscedasticity, biases) (having experience in model diagnostic procedures)
- Knowledgeable and years of practical experiences in basic and advanced statistical modeling, including many of the below statistical analysis methods, models, procedures
- Regression analysis
o cross-sectional regression (t-statistic, chi square-statistic, F-statistic)
o logistics regression (regular and conditional logit model)
o Time series regression (important)
o Multivariate regression
o Cox proportional hazard model (z-statistic) used for survival analysis
- Quasi-experiments
- Experience in academic research would be preferred
- Very good English communications skills in oral and writing
- Weekly availability to be at least 25 hours from now to 01/15, prefer 40-60 hours/weekly availability