Use access key #2 to skip to page content.

anticitrade (99.80)

BLACK BOX ROYAL RUMBLE! (New game proposal)

Recs

21

November 04, 2009 – Comments (20)

It seems you can hardly throw a virtual rock on the internet without hitting some joker pimping his own automated stock picking system that is guaranteed to make you both wealthy and attractive to women (or men I suppose).  Even Motley Fool is rife with these pundits.  But is it really possible that an automated system can provide abnormal returns?  If so, how do you (an idiot in all likelihood) separate the good from the bad?  (sorry about calling you an idiot… I just seemed so funny to me)

 

Well I have the answer!  We start a BLACK BOX ROYAL RUMBLE here on CAPS involving as many automate systems as possible.

 

Here are the rules I propose:

 1) No one can start before 12/1/09 (Ends on 5/1/10)2) All participants must create a new profile with the name "BBRR" (BLACK BOX ROYAL RUMBLE) followed by their standard name (this will make it easy to search member names and find the participants).3) Accuracy will be irrelevant, only points will count (so users will move freely in and out of stocks).4) You can have up to 200 picks.

5) Users are encouraged to specify the basics of their methodology at the initiation of the event.

6) Ideally users will abstain from any soft/personal bias in their stock selection but stick to the constraints of their automated system (impossible to police, but makes it a lot more interesting).

7) Anytime anyone writes BLACK BOX ROYAL RUMBLE the must do so in all caps, and if they speak it they must physically yell it.  Just to show how serious this is.

 

Any other rules we should add?  

Are you interested in this? 

Will you participate?

 

(I think if we can get enough participation the black box charm may be a reality.)

20 Comments – Post Your Own

#1) On November 04, 2009 at 12:28 PM, jddubya (77.83) wrote:

This sounds like an awesome idea for a contest!  Although I personally don't have any automated system it would be great to see one or more of them in action.

Report this comment
#2) On November 04, 2009 at 12:28 PM, anticitrade (99.80) wrote:

Weird how the first 4 rules got rolled up into one big mess....  It should look more like:

1) No one can start before 12/1/09 (Ends on 5/1/10)

2) All participants must create a new profile with the name "BBRR" (BLACK BOX ROYAL RUMBLE) followed by their standard name (this will make it easy to search member names and find the participants).

3) Accuracy will be irrelevant, only points will count (so users will move freely in and out of stocks).

4) You can have up to 200 picks.

Report this comment
#3) On November 04, 2009 at 12:32 PM, portefeuille (99.96) wrote:

exclude the usual culprits like leveraged ETFs and low volume stocks. otherwise the winner will almost certainly have used them extensively.

Report this comment
#4) On November 04, 2009 at 12:48 PM, davejh23 (< 20) wrote:

Only green thumbs.

Report this comment
#5) On November 04, 2009 at 12:48 PM, JakilaTheHun (99.95) wrote:

I probably won't participate one way or another (due to time constraints), but isn't a 200 stock rule a bit excessive?  You are basically forcing everyone entering the contest to make 200 picks (since it is scored on points alone). 

Just seems like it should be more limited --- maybe 50 picks or 35 picks.  It's just a thought --- feel free to completely reject my idea for any reason or every reason :)

Report this comment
#6) On November 04, 2009 at 12:51 PM, JakilaTheHun (99.95) wrote:

Just seems like it should be more limited --- maybe 50 picks or 35 picks.  It's just a thought --- feel free to completely reject my idea for any reason or every reason :)

Of course, on the other hand, I guess that's why it's an "automated stock picking" contest.  But I would still imagine that some people's  automated systems might be better at chucking out 35-50 good picks rather than 200 picks --- and that wouldn't necessarily mean their system was not superior. 

Report this comment
#7) On November 04, 2009 at 1:25 PM, vriguy (44.58) wrote:

Real life investing also involves weighting - your system may tell you to put x$ in stock A and y$ in B, but counting points assumes equal weights.  What you describe sounds like the usual CAPS scoring system.

Report this comment
#8) On November 04, 2009 at 2:16 PM, anticitrade (99.80) wrote:

port,

I agree with elimination of the leveraged ETFs but not sure what we should use to define low volume though.  I kind of thiink it should be anything that you can't enter or leave with 1000$ without some trouble.

davejh23,

I want to agree with you (I think my system works better for green thumbs vs red thumbs), but I don't know the justification for eliminating them.  It seems that a system that can identify both red and green thumbs would be better than a system that can just identify green.

Jakila,

One of the reasons I wanted to enable up to 200 stocks is that it favors the automated system approach.  Plus with a greater number of picks you get better statistical support......  Although, I agree that some systems may only produce 30 or so great stocks, putting hte cap that low would facilitate non-automated stock picking as well.  Additionally, It could be a bit of a headache to verify that everyone only has x amount of picks.... Whereas if you allow people to go to 200, you don't have to do any policing since it is a CAPS established ceiling.

vriguy,

I agree that weighting is an important aspect of investing.  However, within the confines of the established CAPS system there is not an opportunity for it.  However, this competition will be different from the standard caps game since only your points count (No accuracy factor) and it occurs during a strict 6 Month time frame. 

I am not looking to create a perfect competition here.  Just one that is 80% perfect, with only 20% of the work.  Ideally, the rules shouldn't require individual policing of accounts.

Thanks for the response.

Report this comment
#9) On November 05, 2009 at 9:10 AM, anticitrade (99.80) wrote:

It doesn't appear that there is adequate interest in this.  I really thought there were more Fools using automated systems to pick stocks!

Report this comment
#10) On November 05, 2009 at 1:24 PM, TMFJake (94.75) wrote:

Yes, if there was more interest, I'd be willing to create for the contest with a page similar to what we do with the DA Davidson Student Investment Challenge:

http://caps.fool.com/Partnerships/DaDavidson/index.aspx

 

 

Report this comment
#11) On November 05, 2009 at 1:25 PM, Tastylunch (99.46) wrote:

Anticitrade, my comment got eaten by CAPS last night

 anyway at the fool there is an awesome board called Mechanical Investing. You should ask those guys if they are interested.

http://boards.fool.com/Messages.asp?mid=28077165&bid=100093

they all use automated means I believe and on average are msarter/more accomplished than most CAPS players I believe.

Most of them are oldtimers by CAPS standards (Some have been here for ten years)

Report this comment
#12) On November 05, 2009 at 2:05 PM, anticitrade (99.80) wrote:

Good call tasty, I will drop a message on their board and see if I can get a few of them interested.

Report this comment
#13) On November 05, 2009 at 5:22 PM, Popnfresh100 (93.32) wrote:

This sounds like an awesome idea,  here are a couple additional rules that might make it more realistic:

 1. Users MUST start BLACK BOX ROYAL RUMBLE picking system on the start date, to decrease the chances the winner will be someone who just happened to get in at the right time. (Or someone who entered at multiple points in time with the same system)

2.  Have a minimum and maximum number of active picks at all times, in order to prevent sector betting.  (In CAPS you can pick 100 banks or REITs to underperform or outperform at the exact same time, in real life you have to allocate) 

I don't use any automated systems, but I know of a couple I wouldn't mind testing.  I might even make up one or two.

Report this comment
#14) On November 05, 2009 at 5:45 PM, anticitrade (99.80) wrote:

Popnfresh100,

I totally agree with your first suggestion.

I like your 2nd rule as well, but I worry about restricting the types of systems that may be included or create a policing nightmare (checking every users current picks to verify that they are within a certain band).  Part of me would actually like to require everyone to maintain 200 picks at all times.  Because the heart of an automated system should be that it outputs large amounts of quality picks.  It's possible that someone will put together an automated system that just picks sectors, then they just grab the top 200 picks from that sector (although that could lead to someone winning more by accidently hitting the right sector than haveing a working model). 

In general, I would hope we can run 10-20 different systems against eachother and then after 6 months we can say:

"This system did this well, it used this strategy, and it might have just been lucky because of x.  But at least we have a better idea of its true value against other methods."

I really think there are many simple ways that users could test different systems..  Some ideas:

1)  One that does all 5 star rated caps picks.

2)  One that just uses a basic motley fool screen to pick stocks that have a small P/E ratio.

3)  One that uses google search engine software to identify most searched for ticker symbols.

Ect.

Report this comment
#15) On November 05, 2009 at 5:52 PM, anticitrade (99.80) wrote:

Could a member of the mechanical investing board invite me to be a member? 

Report this comment
#16) On November 05, 2009 at 6:16 PM, Tastylunch (99.46) wrote:

I think all you have is ask the Fool for general asccess to the boards. They usually approve in you 24 hours or so. That should give you access to any board.

 there might be an automated form for this I can't remember, I've been a Fool board member for several years now.

If you can't get in for some reason just post here what you want to say there and I'd be happy to copy and paste it in the baord for you.

Report this comment
#17) On November 05, 2009 at 6:48 PM, anticitrade (99.80) wrote:

Thanks for the help Tasty, I will check back tomorrow and see if it will let me post.

Report this comment
#18) On November 05, 2009 at 10:21 PM, UltraContrarian (99.82) wrote:

I'm in.  I'll run the following screens on the CAPS screener each month and pick 100 outperforms and 100 underperforms.

Outperform criteria: P/E < 15, LT debt/equity < .5, current ratio > 1, 3 yr earningsgrowth > -3, 3 yr revenue growth > -3, ROE > 7. Must have up to date financial statements.

(40) Dividends - yield over 1, sort by lowest book value
(15) High Growth - P/B < 3, sort by highest 3 year revenue growth
(10) High Insider Ownership - P/B < 2, sort by highest insider ownership
(5) Short Squeeze Candidates - P/B < 2, sort by highest short interest
(30) General - yield less than 1 (or 0), sort by lowest book value

Underperform criteria: P/S > .01, Short Interest Ratio > 1, CR < 2, Earnings Growth < 20, Revenue Growth < 20, ROE < 10

(40) P/B < 0, P/E < 0, sort by lowest current ratio (+all financials)
(5) P/B < 0, P/E > 20, sort by lowest current ratio (+all financials)
(30) P/B > 2, P/E < 0, sort by lowest current ratio (+all financials)
(15) P/B > 2, P/E > 20, sort by lowest current ratio
(10) Financials - P/B > 2, PE > 20, sort by highest P/B

FYI, I plan to disregard any rules you guys make that I don't like.

Report this comment
#19) On November 05, 2009 at 10:24 PM, UltraContrarian (99.82) wrote:

My name will be BBRRstreetflame (although I would prefer RUMBLINstreetflame - it's catchier).  It's time to retire my UltraContrarian account anyways.  Keeping track of 200 picks takes enough time.

Report this comment
#20) On November 06, 2009 at 3:59 PM, Popnfresh100 (93.32) wrote:

Ant,

I like the everyone-picks-200 rule.

I think the best way to make a CAPS portfolio perform like a real life portfolio is to maintain a set number of picks to simulate allocation, and to close the picks on a regular basis to prevent the majority of the outperformance coming from caps point compounding on one or two high-flying picks (re-open with the same pick if you want a buy-and-hold system). 

But I also think you're right about wanting to keep this simple. Just having a set # of picks would do a lot to help with simulating allocation, and keeping the number of picks high, and the time frame relatively short, will probably help to prevent caps point compounding on high performing stocks from swaying the data too much.

The portfolios wouldn't simulate a real-world portfolio, as I don't know anyone who holds 200 stocks at a time and isn't an ETF, but I think they would do a good job testing the systems.

Report this comment

Featured Broker Partners