Skip to content

Latest commit

 

History

History
42 lines (32 loc) · 1.61 KB

5. Screens & Data Cleansing.md

File metadata and controls

42 lines (32 loc) · 1.61 KB

Screens

Screen Description
Major stock exchanges only Only stocks traded on NYSE, NASDAQ
Accrual Screen Remove potential fraudsters and manipulators
No companies with less than 5 years of operation Necessary for calculating 5 Year Average ROIC metric from ERP5
Last month's momentum While calculating generic momentum, we will skip most recent month's returns due to short-term reversal effect

Data Is Always Dirty

  • Once the screens were applied:

    • There were 138,696 records

    • On average, 3.8% of data was missing across 16 variables

    • Removed all companies with any missing metrics

    • Final dataset consists of 89,588 records (i.e. stock years)

From 138,696 records that were available after applying the screens, distribution of missing values was as follows:

Variable Name Number of Missing Values Missing %
Earnings Yield 2,255 1.6%
Return on Invested Capital (ROIC) 4,013 2.9%
Price-to-Book 4,736 3.4%
5 Year Average ROIC 1,128 0.8%
Generic Momentum Measure 0 0%
Frog-in-the-Pan (FIP) 53 0.04%
Net Profit Margin 2,557 1.8%
Long-term Debt/Invested Capital 2,472 1.8%
Total Debt/Invested Capital 2,699 1.9%
Price/Sales 2,554 1.8%
Total Debt/Total Assets 2,127 1.5%
Pre-tax Return on Total Earning Assets 22,317 16.1%
Long-term Debt/Total Liabilities 1,907 1.4%
Enterprise Value Multiple 2,364 1.7%
Interest/Average Total Debt 29,863 21.5%
Asset Turnover 2,985 2.2%