2

I’m looking to build a portfolio using the following Peter Lynch Style strategy. What I’m wondering is, do you think this is a good approach? Is there anything you’d change about it?

Here’s the strategy:

  1. Screen stocks using Peter Lynch’s approach for Fast Growers:

    Market Cap is >= US$ 600 million (basis year 2000) adjusted yearly, Average 5y EPS growth % is greater than or equal 20%, 0 <= PEG ratio <=1, For non-financial and utility companies: 1y Inventory to sales ratio growth % is < 5%, For financial companies: Equity to asset ratio is >= 13.5 AND latest filing RoA% is >= 1%, For non-financial companies: Debt to Equity ratio is < 30, If TTM Sales is > $1 billion, then latest filing P/E ratio is <= 40

  2. Investigate any stocks meeting this criteria in detail and buy stocks in 20 companies than I’m happy with (with a good mix across industries). Initially my budget is $10,000 so I will buy 5% ($500) of each stock.

  3. Continue to send $1000 a month buying more stock ie buying $50 extra of each stock every month.
  4. At the end of the year, run the screen again- Any companies no longer meeting the screen criteria will be sold and replaced by companies that do meet the screen.
  5. Stock in these companies will be bought in such away that the amount owned in each company is rebalanced to 5% of the total portfolio.
  • 1
    It would be telling if you have access to past data and you could back test your idea on previous company performance (10 years?). That would offer insight on the viability of your strategy. – Bob Baerker Jun 5 at 13:22
  • Thanks Bob. In order to do this back test I would need to screen the stock market based on the what the financial ratios (PEG, Debt/Equity values etc) were for all the companies 10 years ago so I could decide which companies to include in the portfolio. Do you know where I could get the info to perform a historical screen like this? Thanks – Kieran Daly Jun 5 at 14:10
  • I'm not a FUNdamental kinda guy so I have no clue if such a database is available. let alone for free. And even if available, this would be a bit of a brute force project, year by year: Select companies, determine price performance, dividends received (reinvested performance?), M&A changes, portfolio substitutions, etc. – Bob Baerker Jun 5 at 14:23
  • 1
    How much are you paying your broker in commissions for the trades? Can you buy partial shares? – NL - Apologize to Monica Jun 5 at 14:58
  • @KieranDaly your brokerage might have such information. – RonJohn Jun 5 at 15:33
1

Looking at the question and some of the comments, I think you would need to look into backtesting. Because what you describe is a set algorithm (a program can run this portfolio), you can easily backtest everything if you have the right datasets. Assuming you can do basic programming/can learn quickly, I would recommend Quantopian which is generally used to create trading algorithms which are then entered in a competition. You could just use this for their data sets and to backtest your idea. There are also many other places to get the data from.

A couple suggestions

One: Seeing as you have limited capital, I would suggest investing in probably around 10 companies or even less because you have a limited amount of time. In the time it will take you to research 20 companies, the market will have changed too much.

Two You should probably run the screener more often than a year. Companies put out earnings every quarter so at a bare minimum screen every quarter.

Three Since you are asking this here, I assume that you don't have that much trading knowledge or experience. I think your time could be better spent learning more about the market and fundamentals. For all you know, he could be telling his students to invest in these companies to drive the price up so he can make profits (this probably isn't happening but it is possible).

Anyway, good luck and have fun!

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.