The Motley Fool Discussion Boards

Previous Page



Subject:  Re: CAPS Geniuses Date:  2/15/2013  2:05 AM
Author:  XMFCrocoStimpy Number:  394360 of 424

This has been a great thread, and I'm glad to see that there is still an active group within CAPS still looking at the quantitative uses of the system. For those of you who don't know me, I was previously heavily involved in work to extract investable signals in a long/short design of higher ranked stocks minus lower ranked stocks, and in building the simulators of the CAPS system to allow us to test the impact of changing the scoring metrics and stock eligibility. Unfortunately my role at The Fool has taken me elsewhere for the last year and a half, though I have continued to monitor the performance of the overall CAPS system.

One of the competing sets of interests that have always been present when I read discussions about the scoring metrics/ranking system has been the tug of war between wanting this to be just about picking stocks (which is its fundamental design), and wanting to impose some real world constraints to make the picks and the use of the picks to be more like investing. Both have their place, but they do not always play nice with one another, and retro-actively superimposing an investing metric across the decision making process that was just based on stock picking rules could lead to some pretty unrealistic measures (not to mention really tick some people off since they wold have used a ompletely different strategy). I'd like to remedy this in the way that most any legacy system gets a major change: I want to build an add-on.

This is the basic concept I'm working with, and hope to get some feedback from players like yourselves that have a vested interest in building additional functionality. The system would be SimFool (with apologies to Will Wright), and at its heart it would be both a portfolio investment simulator and a CAPS player generator. Simplicity is the goal to prototype it quickly, with the ability to expand functionality based on active participants interest.  Each player in SimFool would start out with say $1M virtual cash, the goal is to make money. Stocks can be purchased with different sized investments (unlike CAPS, where they are all implicitly equally weighted). However, a CAPS doppelgänger player will also be generated with the buys being listed as "outperform" and having equal sized positions. In SimFool we will monitor annualized returns, plus characteristics of the portfolio such as volatility, exposure to different market benchmarks and factors, and ultimately attempt to bracket the expected uncertainty around a player's "skill" that might simply be luck. 

This initial approach wouldn't allow for tackling the down side of the market - no shorting, to translation to "underperform" calls for the CAPS doppelgänger. Those features could come later, but why weigh down the initial attempt?

I'll be very straight with people - this is exploratory on my part, nothing else. I see a tremendous amount of value that we could unlock from this system, all the while building a large scale example of the difference between stock picking by itself, and stock picking wrapped in investment strategy.  This would also help to revitalize the activity within CAPS itself, though what will be seen is an influx of long only players (which I don't see as a bad thing). 

 Feedback is what I need. Interest in such an add-on system? Does it have the right potential functionality, or is it missing something crucial? Anyone concerned it would have an adverse impact on CAPS?

Thanks in advance for any help,

Xander (TMFCrocoStimpy) 

Copyright 1996-2020 trademark and the "Fool" logo is a trademark of The Motley Fool, Inc. Contact Us