I was recently inspired by Charlie Munger and decided to use his philosophy of inversion to build a back test. The question I asked myself is what makes a bad investment? To that I answered:
- They take on a lot of debt which adds risk to the company if things go wrong.
- They give away the company to employees or in acquisitions diluting existing owners.
- They don’t grow
- Earnings don’t arrive in cash
- They have a business model that can easily be recreated
- They sell at a high price
To make a good company I reasoned I want to invert the bad company metrics. To model these factors I back tested through Portfolio 123 by ranking the following data:
-
They take on a lot of debt which adds risk to the company if things go wrong.
- Low Debt to Assets: Reviewed Quarterly
- High Interest Coverage: implies debt load can be covered by earnings
- They give away the company to employees or in acquisitions diluting existing owners.
- Low Equity Dilution: Kinda obvious
- They don’t grow
- High Sales Per Share Growth
- Earnings don’t arrive in cash
- High Free Cash Flow per Share Growth
- High Operating Cash Flow Growth per Share: Companies showing an upward trend in cash flow growth are preferred, as this signals ongoing operational success and growth, avoiding those that have become static.
- They have a business model that can easily be recreated
- High Average Gross Margins: High gross margins suggest a company can add a lot of value to the raw materials or by the info they have. Economic theory would suggest if others could do the same, the price charged would not be that much higher than the inputs.
- They sell at a high price
- High Yield of Gross Profit to Market Cap: Going down an income statement or cash flow has messier and messier financials. Focusing on the top level should give a cleaner Yield and I believe would allow for better valuation of earlier stage companies.
Here were the results:
https://docs.google.com/spreadsheets/d/1M1Kxw5m6sFBr-a-b-Hks5qAF7jsewPFqkU2BpszlymM/edit?usp=sharing
- The back test tab selects the higher ranked securities and market cap weights them subject to a 6% max weight and owns 100 stocks at a time from the Russell 1000 index.
- The Ranking Score Tab shows an equal weighted average annualized return.
Portfolio123 adjusted for look ahead bias and delisted securities and all that.
The big question of course is, does anyone think this will work in the real world? To me it’s a very intuitive strategy and has promise. I would like to hear your thoughts and in particular connect why my assumptions of what a bad company and good company is could be wrong.
Other Fun Facts:
- Top performer in backtest: NVDA, been held for over 3,000 days
- Top 10 Holdings Currently: NVDA, AAPL, STLD, LRCX, NUE, FAST, V, SNPS, TXN, KLAC
Disclosure: I own NVDA and V (unfortunately not for the past 3000 days.)
Leave a Reply