Wanted to share this and ask the audience for insight/comments: I screened on Yahoo Finance for the top % gainers for the last 12 months, and various stocks returned such as NTES, DCEL, SOHU, MICC, etc. I "profiled" these and found that (out of the top 100 gainers) - A 50/125 SMA cross would have signaled entry into these as they climbed - Almost all were $3 to $10 at/near time of 50/125 cross - Most all were less than 100 million shares outstanding - Most all had 100K to 500K ave. volume at/near 50/125 cross - Most all currently (and likely at the time of cross) have Investor Business Daily Group Strength of "A" or better - Most all currently (and likely at time of cross) have Investors Business Daily Composite Rating of "80" or better * looking for comments, criticisms, ideas, reference my observations, and maybe I am missing some things. This study came after observing NTES and ASKJ and saying "dammit, how can I catch those" so here I am today. thanks BILL
The problem with this analysis is that you're not accounting for the many that DID cross the mov avg and yet failed to proceed and would have caused you losses. I believe there's an appropriate description of this oversight but it escapes me at the moment. It's kind of like saying...you know, people who win at jackpot machines have several things in common, one of which is that they all pull the lever a certain way. What about the rest who lose?
this could not be better stated. if you take some logic courses, you run into this all the time. one is necessary and sufficient for another to happen, but the inverse is not always true. sure, all the big breakouts are gonna have their ma's crossing (ma's are just a picture of what happened in the past).
yes I agree with the slot machine example, no argument however, if we come across stocks meeting the above criteria, and say place Buy Stop - Market orders higher than say their recent highs, as an added "filter" to confirm trend, is this not somewhat of a sound methodology?
Do you pay to advertise here? I have to admit, you are a machine. You take every opportunity to pimp your service.
If you notice I only point things out when it is relevant to what we do and pertains to something that people are asking about on the board. In this example the link I gave is to a section that is completely free. I just like to participate in Elite discussions which are relevant, this is not where we get most people to look at our site, not even close. To get people on our site we use our brains ;-) Just look up "50 day moving average" or "volatility stock screener" on google.
Start following stocks that have broken out. How many went up in a straight line and how many pullback? there's always a chance to get in after the breakout.
I think what we are seeing here is the intersection of good core system research versus survivorship bias and data mining.
Snooptrader - Yes, I believe they have paid to pimp (or at least they had at one time) - although it would be more truth in advertising if his/her handle was identified as being a rep for that company in the same way that Bright, def, and the people from other advertisers are (instead of tagging him as a "Senior Member") - his AOL IM address is even SupprtTradeIdeas. They've definitely "used their brains" - though his suggestion to google those phrases is an example of how - if you take a quick look at the HTML of their web pages, they've done a nice job specifically engineering them for web crawlers - incorporating most of the usual trading phrases into the text or meta names (everything from "stop loss" to "ECN" to "moving average") so the search engines will find and multiply catalog the site. Now whether their overall notion has any merit or usefulness, that's a whole different subject. As far as your moving average crossover goes - all you're seeing is a common factor in all trend moves. Look at it over a longer period, especially during range trading periods - in the long run, systems based solely on any moving average crossover don't pass the test of time. Got to combine it with other factors to try to reduce noise and false positives.