A new indicator

[doctorj]

I've often pondered the potential of using the number of posts generated about a particular share as an indicator of it's short term potential. Obviously the indicator would be skewed towards small caps and wouldn't distinguish between stocks that are experiencing price action at either end of the spectrum but it might be a way to create a watchlist for eyeballing.

Someone more proficient in coding than I should be able to code something to troll through ASF, SScene, Sharetrader and HC and stat the number of posts about particular shares. Daily highs, 1-5MAs and some other simple stats could be used by individuals to create their shortlists based on personal preference.

This does sound random, but I've noticed some unqualified observations that there SEEMS to be a correlation between the number of posts on stocks across the various forums and short term price volitility. I'm interested in anything that may help shortlist dogs for speculation. Thoughts?

 

[tech/a]

Really the challenge is to all who look for runaway stocks to get on them.
Unless you know about them thats not going to happen.

Doc this is a complex issue and a very important one in my veiw as Id say that most here trade that way.

(IE,Small caps,Small investment funds,and short timeframes,with expectations of high returns in short timeframes.(+30%)

See if your trading $2000 you want another $2000 quickly.
If your trading $200,000 then $2000 or 1% is of very little impact.

So is it more about the style of trading that your/and lots of others are attempting to tap into?

Hows the Breakout alerts going for you?

I have a lot to add to this topic but not the time at the moment perhaps a few others would like to comment first.

 

[doctorj]

Tech, I have a few theories regarding your breakout alerts but its very difficult to test with the infrequent access I have to them. I think the first issue with the breakout alerts is at what time of the day are they most affective? Ideally, I'd like to see several from one day over the course of many days to play around with some things.

Trading these stocks is a challenge. You need to be able to jump aboard early enough to get enough of a ride to warrent the risk, but not too early that you're jumping at false signals.

The main reason I started this thread is not really to discuss the specifics of an indicator related to forum posts but as you've said, a collection of theories, similar to my breakout alerts theories in progress, to assist in better trading dogs.

 

[Mofra]

I've often pondered the potential of using the number of posts generated about a particular share as an indicator of it's short term potential.

Its worth pointing out at this point that the MUL thread on this site is currently up to its 69th page of replies. Interesting thought but in practice I'm not sure how effective it would be in comparison to more traditional methods of analysis.

Good luck

 

[doctorj]

A valid point. A quick comparison shows that MUL has averaged about 85 posts per month on this forum and ZFX about 70. In the same time, they've changed -63% and +31% respectively - both of which seem to reflect the volitility I was referring to in the parent post.

Ofcourse, without research (which I lack the programming ability to do efficiently), my evidence is only empirical. The thing that attracts me to something like this is the basis that many people with many different specialties can cover the 1500 odd companies much better than I can and many shortlist their universe of stocks in vastly different ways than I do. A stat that can capture the experience of others is appealing, even if it only provides a basis for me to quickly create an additional shortlist for eyeballing and cross referencing with the list I create through my more traditional approaches.

Or maybe I'm just crazy.

 

[RichKid]

There are always issues with the quality of stats and methodology when using postings/threads for this purpose.

One issue is how many forums do you include? Does some of the bickering on hotcopper count? If you look at the two threads compared, MUL and ZFX, do we look at the number of posters (for eg Chicken has posted almost all the posts on ZFX; MUL had Stefan posting prominently but now he's away for a bit, so do we then consider the posts of certain posters with a greater weight? Does the absence of a prominent poster affect things? ie do we look at the profile of posters?).

Also the issue of entry, some people may have been on a stock for awhile and then post on it to push it a bit higher before it slows, how do we pick rampers? Also don't forget that we don't really know about the volume of posts on a stock till AFTER the posts build up so we may have missed the bulk of the move. With cases like MUL the sheer volume of posts may not be conclusive as you pointed out. There are some threads which have looked promising but have died with the stock after an initial flurry of posts- it that sense would we get more info off the price chart or anncts than the threads?

If we are looking at posts for an indication of crowd interest or behaviour wouldn't existing indicators be sufficient?

Rather than trying to throw cold water on the idea I'm trying to point out how difficult it may be to come up with a system that can be applied consistently, with defined parameters- perhaps that is something we can continue to discuss. My primary concern is that another indicator is just going to be a distraction, I have trouble enough with the current ones!

 

[doctorj]

Obviously, certain statistics drawn from the raw data would be more useful than others. The theory is that in some cases it may function as a leading indicator, particularly in the instance where people that want to get other opinions before dishing out their hard earned.

In hindsight, this thread is largely irrelevant without testing to either support or refute my empirical observation. Unfortunately I lack the tools to program something capable of testing the observation and it appears I've not managed to convince anyone else it is worthwhile

 

[Mofra]

This thread reminds me of the ANZ job index figures released on a monthly basis, which many have argued is irrelevent due to the number of internet jobs advertised nowadays.

Perhaps comparing the volume & volitility of a stock before it was widely mention to after would at least give a basis for how effective these chatrooms are at raising awareness of a company, so that some basic measure could be applied, ie x number of new messages/posts about a company = y increase in average volitility

 

[tech/a]

Doc.

Inside trader have a monthly reporting on 3000 analysts who give their choices as best value stocks.All purely fundamental and all collated and rated by number of consultants who choose the stock.

I went through the list and chose those with n my veiw the best "Value" in terms of cheap stocks with what thy thought was fundamental value.
There were about 8 which i have on a watchlist at the office.

I think 3 are below purchase price and the rest above.
While by no means an exhaustive test Im sure you could collate and play with the old results knowing now stock prices a few months down the track.


Im sure I have a print out here.
Ill email you that printout if you think it would be benificial to you.
I for one would be interested in a more indepth study and the results.
Wouldnt 3000 "expert" selections be better than those from forums?----hmmm maybe not!


While finding a prospect is important how you trade it is in my veiw MORE important.

What do you have in mind doc---you methodology?

 

[doctorj]

I wouldn't call it a methodology, more of a social experiment I guess

At the moment, in parallel to testing some theories to develop a similar-but-different system to techtrader, I'm working on some other outside the box things for either discretionary trading or as mentioned here the construction and maintenance of shortlists, the premise being that we don't trade the same way in bull and bear markets, why should we trade the same way in small/mid/large caps?

Any information you could give me would be wonderful. I was hoping for a market scan for about 45mins after open today...

 

[tech/a]

Doc Ill get your scan tommorow sorry!!!!

Just noticed that Inside trader do what you have suggested with their clients.