GET IN TOUCH

Development of analyst ranking model

3
Quant analysts.
1
Uncorrelated model.
100+
Analysts covered.

Client

A US-based hedge fund.

Opportunity

Our client wanted to develop an analyst ranking model to identify how analyst crowding could be used systematically to assess stock performance.

Implementation

Considering the specifics of our client's business model and investing strategy, we deployed a multi-step process:
  1. Downloaded all historic revision of the analyst ranking from Bloomberg and Reuters
  2. Created a scenario of analyst ranking dates R and stocks reaction on R+1, R+2 and R+n days
  3. Identified correlations and number of holding days for analysts, who were top-ranked for respective stocks
  4. Developed a highly proprietary technology and models of crowd behavior
  5. The model was maintained and used by the hedge fund to deploy capital, whenever there was a ranking change

DATA SOURCES USED:

Bloomberg
Reuters

Value Delivered

►
1
SG Analytics delivered a model that was completely non-correlated with the hedge funds existing model.
►
2
The model enabled the client to track their existing portfolio.
►
3
The hedge fund could take overweight positions on certain stocks and underweight position on certain other stocks based on the analyst ranking and SG Analytics backtests.
►
4
The proprietary model was used by the hedge fund to take large size positions in new stocks.
 

Related Case Studies

Monte Carlo VaR modeling on options portfolio
Factor modeling backtest for New Yorker hedge fund
ETL mechanism for risk-analysis feeder files