I did a simulation, that turned out a bit controversial, to sanity check whether a guy's trading strategy had any chance of delivering as advertised:
If you buy the top 5 largest food producers by market cap (currently Nestle, Mondelez, Hershey, General Mills, Kraft Heinz) right after ex dividend and sell before Quarterly Earnings. Rinse and repeat every quarter. They statistically yield 29% annually.
My original sim just used those 5 named stocks for the last 4 and 10 years, on the basis that if it can't perform even with ultimate market winners, it has to be a dud. The result was 6% annual returns.
But some disagreed, arguing that choosing today's winners would make things worse, not better, as it loses the benefit of exploiting short term movers. So I took on the challenge of improving the simulation as a fun project.
And here are the results and some extra info showing how the top 5 shuffle over time. The new approach gives an annualised return of 1%.
The algorithm: wake up each day and check if there's any earnings dates and if so sell. Then if you have available capital, check the top 5 stocks, and if there's ex div dates, buy.
I didn't bother with restaurant chains, drinks co's etc as that didn't seem to be the original intention.
Leave a Reply