Notes on: Alberg, J., & Lipton, Z. C. (2017): Improving Factor-Based Quantitative Investing By Forecasting Company Fundamentals

Table of Contents


  • Certain key factors which can be computed from the reported data are known to out-perform the market average, known as fundamentals
  • Using DNNs they accurately predict these key factors and thus beat the market average when running backtesting


  • alberg17_improv_factor_based_quant_inves_473fd65ee4501a70c2e1a5987c9279035ff1b4a8.png
  • alberg17_improv_factor_based_quant_inves_780c54181b328a77e2016d2184a07f2634680259.png is the expected portfolio return (i.e. the average return)
  • alberg17_improv_factor_based_quant_inves_1f7926eae1f02986a2c4f146d5b8ec00af99dad1.png is the risk-free return , i.e. a benchmark investment which for which there is no risk, e.g. interest rates
  • alberg17_improv_factor_based_quant_inves_d900d8b33e269687684d44c4fb477f2d9f058d7d.png is the std. of the portfolio
  • LFM = Lookahead Factor Models
  • book value is the value of an asset according to its balance sheet account balance


  • Book-to-market: book value normalized by market capitalization
  • EBIT/EV: operating income normalized by the enterprise value