MBA Project

A

Albert

Hello all.
I am in the beginning weeks of a statistics class online through Baker College and in need of a research project. Since I love trading I thought I'd start here.
I am interested in starting a survey of traders and evaluating the impact of longevity on success. My research paper's stated hypothesis is that there is a positive and direct correlation between a trader's success(?) and the length of time involved in the trading business.
Now I can see a few problems with definitions already such as how to define success- earned income versus return on the dollar versus the intangible benefits of making enough to not go into a job: Is a guy that spends 8 hours a day at it over a year the same as somebody who looks at charts for 30 minutes each night for a year?
Looking for feedback before I submit the survey.
Albert
 
Albert,
You are going to have to get very general with your definitions or your going to have to make them very precise and have lots of variations. I'm not sure how much time you have, but there are all kinds of traders out there. All have different ways of going about what they do.
 
What if you never find a successful trader?

Trust me.. With the time, work, effort, mental agony etc involved, all the traders, even those rare successful ones are net negative.
 
Longevity definitely has a strong positive correlation to success. When I compare myself now to where I was when I started I can see a marked difference. It may not necessarily be related to the number of hours spent in the market. What I did in 8 hours as a beginner I can do in 30 minutes now, and with much more accuracy. In my opinion, success is related to experience, and experience is definitely tied to longevity. Just my 2 cents, hope it helps. :-)

-- Paccc
 
Give some thought as to where survivor bias might come into this. Also with fixed position size (min 1 future or 100 shares) you might also want to give some thought as to starting account size being a factor in success.
 
I really think you'll be better off with doing something related to market.

As to explain, why the month of January is special for economic growth? why major drops in equity markets happen in October? and why you should never sell a stock on a Monday etc etc..

Make it easier on yourself.
 
Tend to agree, you may well find that you come up against all kinds of ways someone can pick holes in your methodology.
 
you will experience what true statisticans call, the sampling bias. period.




Quote from Albert:

Hello all.
I am in the beginning weeks of a statistics class online through Baker College and in need of a research project. Since I love trading I thought I'd start here.
I am interested in starting a survey of traders and evaluating the impact of longevity on success. My research paper's stated hypothesis is that there is a positive and direct correlation between a trader's success(?) and the length of time involved in the trading business.
Now I can see a few problems with definitions already such as how to define success- earned income versus return on the dollar versus the intangible benefits of making enough to not go into a job: Is a guy that spends 8 hours a day at it over a year the same as somebody who looks at charts for 30 minutes each night for a year?
Looking for feedback before I submit the survey.
Albert
More...
 
I want to thank all of you for me being generous with your suggestions and professional with the tone of all the comments.
To reply:
Survivor bias is the issue that most concerns me, but maybe I'll get lucky and some beginners that frequent the site will participate.
Sampling errors are always an issue. I can avoid it by increasing the poll size, but it surely will not go away, perhaps just mitigate it.
As far as being"net" negative over the years I may not be getting to that issue with the set of questions that I envision using. But the point would be how successful the trader is at this date, not "in total to date." This may bring up more questions than it solves (or addresses) but isn't that the point of delving into something, to get a core kernel of truth that you can build on.
To change the project is always an option, but the die has been cast at this point and I'm feelin' lucky. C'mon baby, gimme 25 players (minimum sample size, per assignment.)
Thanks again, those suggestions and criticisms are exactly what I'm looking for.
Albert
 
There are a couple of other issues that I need to hammer out. I need to find a way to keep the information private. I don't mind sharing general numbers with everybody, such as average age or profitability or style of participant, but I surely will not specify that Joe X does this or has that.
Survey construction is an art form especially if you have an inherent bias (as implied by your hypothesis.) I hope the class will relieve me of my bias but I wouldn't mind having an actual trader preview it to see if it leads to specific answers.
Perhaps an administrator can chime in here.
Albert
 
You must log in or register to reply here.
Back
Top