Skip to main content
No. of Recommendations: 6
what one does with the freed up money matters a lot as to whether you should have sold a position. With Pro that is a known. But for other services? What kinds of assumptions had to be made for that study?

I vaguely recall the study in question, and I think the answer was that in the backtest, they just applied the normal rules that would be followed by Stock Advisor. It was a simple "what if" test of what the SA overall return would show if those stocks hadn't been "sold."

I don't think the SA return is a good representation of real world investing - in particular, I don't believe the ROI is a time-weighted internal rate of return, but rather just an average of all the individual returns. If you're just averaging individual returns, it biases you to keep the big winners around, since they can underperform individually but still outperform at the service level due to their size.

For example - make it really simple. Say you have a recommendation service with just one stock, which is showing a 1,000% return. It reports returns as the average return of all stock picks, both active and closed (where closed returns are frozen at the time of sell). This service would show an overall return of 1,000% since there's just the one stock and no sales have been made.

Now suppose you're running this service and you're faced with these choices:

1. Sell this stock and buy something better
2. Keep this stock and buy something better

For argument's sake, let's say that you know for a fact that the current big winner is going to gain another 5% in the coming year, and the new stock you're considering will double in a year. In that case, here are the returns you can expect to see in the coming year for each of those choices:

1. Sell the stock and buy the new one - in this case you freeze that 1,000% gain, and average in the new gain of 100%: (1,000% + 100%) / 2 = 550%

2. Keep the stock and buy the new one too - in this case your big position gains 5% to grow to 1,050% total, and you also get the 100% gain on the newcomer: (1,050% + 100%) / 2 = 575%

So even though you know for a fact that your current stock will be a dog compared to the new one, you're best off keeping it around anyway.

Interestingly, the best idea of all, given this computation method, would be to not buy anything new at all, since buying something new increases the denominator, but is very unlikely to increase the numerator sufficiently to keep the same average. But in a service that is based on making periodic picks, given those two choices, the best idea is to keep the big winner, as long as you believe it will gain anything at all. Only if you really think it will reverse course and lose money should you sell it.

In a real money port, if you had the above two choices, the answer would be easy. Sell your big winner, and reinvest the proceeds into the new idea. That would get you from a 1,000% return to a 2,000% return in a year. In a real port, this is what you would do every time.

So I think the computation method that Stock Advisor uses is inherently not a good way to model real world investing. A better approach would be to use some sort of TWIRR system where you allocate a fixed amount to every idea. That initial investment amount would need to be adjusted over time for inflation and to keep up with the portfolio - maybe add 10% per year to the virtual amount of money being injected.

Print the post  


What was Your Dumbest Investment?
Share it with us -- and learn from others' stories of flubs.
When Life Gives You Lemons
We all have had hardships and made poor decisions. The important thing is how we respond and grow. Read the story of a Fool who started from nothing, and looks to gain everything.
Contact Us
Contact Customer Service and other Fool departments here.
Work for Fools?
Winner of the Washingtonian great places to work, and Glassdoor #1 Company to Work For 2015! Have access to all of TMF's online and email products for FREE, and be paid for your contributions to TMF! Click the link and start your Fool career.