Skip to content

Latest commit

 

History

History
78 lines (62 loc) · 3.57 KB

Datacamp Systematic Investment Strategies.md

File metadata and controls

78 lines (62 loc) · 3.57 KB

Datacamp course Systematic Investment Strategies

DataCamp github repo
How to create a DataCamp course

some comments regarding Zivot "Intro to Computational Finance with R"
Zivot Intro to Computational Finance with R
https://www.datacamp.com/courses/computational-finance-and-financial-econometrics-with-r

  • contains some time series
  • no quantmod
  • no factor models
  • no active portfolio management strategies
  • contains some CAPM portfolio analysis, but no optimizaton
  • no Machine Learning (backtesting and shrinkage)

reproducible perform an operation from start to finish
leverage my experience as PM
practitioner approach

model regularization shrinkage

  • Loading and scrubbing time series data: packages xts and quantmod,
  • Estimating risk and performance measures: volatility, skew, CVaR, risk-return ratios (Sharpe, Sortino, Calmar), package PerformanceAnalytics,
  • CAPM model: market portfolio, regressions of asset returns, alpha, beta, CML, SML, package PerformanceAnalytics,
  • Factor models: CAPM, Fama-French, Barra, statistical,
  • Asset pricing anomalies: size, value, momentum, volatility,
  • Investor risk preferences and utility functions: investor prudence and temperance,
  • Kelly and CAPM,
  • Performing rolling calculations using vectorized functions: package caTools,
  • Performing factor model regularization shrinkage
  • Constrained portfolio optimization: Akaike and Bayesian information criteria, coefficient shrinkage,
  • Out-of-sample performance of optimized portfolios,
  • Portfolio management strategies: risk parity, minimum correlation, minimum variance, maximum Sharpe, maximum CVaR,
  • Estimating model parameters,
  • Forecasting returns and volatility,
  • Active portfolio management strategies: tactical asset allocation, universal portfolios,
  • Strategy backtesting and metaparameter tuning: data snooping, cross-validation, model overfitting, parameter regularization,
  • High Frequency trading strategies: volatility pumping and harvesting,

I envision each vignette would contain reproducible R code samples, relying on fast, vectorized code. The R code samples would use actual market data, and would be self-contained and include data loading, formatting and preparation, analysis, model building, and visualization.

  • Machine Learning for Systematic Investing
  • Investment Portfolio Optimization with R
comments:

teach to use packages xts, PerformanceAnalytics, PortfolioAnalytics, backtesting framework backtesting

Both are very good, and the course I envision would combine the concepts from these two and move beyond them, as a logical continuation. Each lecture would consist of several vignettes, each illustrating a particular technique or model. Here are some topics to start:

I will be travelling over the next few weeks, but I will have time to refocus on this project starting in the second week of December.

explore and adapt:

machine learning courses

Python