What do you think of MrM's system

[Lyehopper]

Well there are quite a few companies out there that either sell you the information that there AI has produced, or sell you some software to train a neural network.

But basically, training neural networks to find patterns, there’s also other approaches such as data mining and genetic algorithms. Neural networks are kind of a new tool and becoming more reliable, they are often used as other pattern recognition tools such as:

Radar systems, face identification, object recognition, voice recognition, handwritten text recognition, medical diagnosis, but if used correctly can identify trades that have good potential for positive feed back/rapid growth.

Usually these systems are used in combination with each other and before the information is applied to the ANN, filters and other trading techniques that are already know to work quite well are used.

Many in the past have found ANNs to be unreliable at predicting the stock market. However they did not reduce the noise of the samples (you wouldn’t pick a Rick Shaw stock would you) and they used no other predicting systems already know, these academics did not look deep enough into the trading side of there system.

So I’m aiming for a method that uses a combination of filtering systems, and a combination of modelling systems (like Mr M's) before being applied to the ANN

All that said, I wouldn’t be able to give you any reliable stock picks right now, unless you wanted it based just on the filtering method. I’m not really interested in that, I just want to prove that ANNs can be applied to the stock market.

My current filtering methods alone give a 180% annual increase (I think I calculate Mr Ms system at 205% so I’m quite interested in testing this)

However I have concentrated on a system that finds all the data with rapid increases, it also imports a lot of small losses

After Appling this type of data to the ANN I expect most of the small losses to be filtered out and to have an annual profit between 200 and 300%. Still it’s a working process so I’m afraid I’m not much use to you guys right now.

There are however, many ANN companies out there doing this right now. Have a search for neural network and stock market.

Sounds alot like Spikefader's Cream-of-the-Crop filter system. AND guess WHAT?.... Spike dosen't charge us anything for his services (but he does accept donations....jejeje)

Hey jamiew.... Do you ever sell short?

[jamiew]

Yes. I haven’t concentrated on marketing yet, I need a bit of time to get it all finished first. I’ll probably let the info go free too, for at least a couple of years. So hopefully i may join in on your picks section in the (not that near) future.

[lemonjello]

Hey jamiew,

What do you do with a PhD in physical biological chemistry? I think I might have some friends that would be interested. What are the primary texts? Which university are you attending?

Best of luck.

Ah, doesn’t quite have the same meaning to me. PhD in physical biological chemistry.

as well as studying AI ANN systems (no insemination unfortunately) and will continue a post doc in this area.

[IIC]

I have a PhD in lookin' good...Doug(IIC)

[Karel]

ah ha, so somebody has back tested it, i would be quite keen on these results too.

Hi Jamie, welcome on board.

The first I heard of $$$Mr.Market$$$ was on The Motley Fool, were he was bragging about his winning streak. People, including me, just summarily dismissed him, until I dug up his picks from his Fool board (now defunct). His method did not do bad at all in a bad time for the market. It was also a method that made sense to me, so I became really interested. Since then I have been forward testing his method, with real money, on paper (see my Following MM thread), and I have developed an approximation. MM has changed his method since I created my approximation, but I am still happy with it.

It is difficult to really backtest MM's method. If you can find a way to do it, or to backtest an approximation, I would be really interested.

And MM's method is the best excuse for playing golf an amateur investor can have. When you have your limit orders in place of course . It is very low maintenance indeed.

Regards,

Karel

[jamiew]

Ah good to see you Karl, I agree it is actually quite hard to back test it since it relies on filtering calculations that are done in real time. What sort of average % annual profit is your real time test making? I think most people will disregarded this system on first look, how ever the profits that have been quoted are impressive, so I’m looking in to it.

If i get time, i may design a little program that automatically goes through Mr M's system every day to give the top 5 stop picks, see how well it stands against the test of time. It shouldn't take too long to design, its just I have a few other projects i need to do first. I’ll go check your thread now thanks.

I`ve now quoted Mr Ms system on a site (hope he doesnt mind, i did ask him before but never got a reply), i`havent put anything bad or good, just put down what i found out and I`ll let people make there own minds up. Seems it could be a good system though.

Hi Jamie, welcome on board.

The first I heard of $$$Mr.Market$$$ was on The Motley Fool, were he was bragging about his winning streak. People, including me, just summarily dismissed him, until I dug up his picks from his Fool board (now defunct). His method did not do bad at all in a bad time for the market. It was also a method that made sense to me, so I became really interested. Since then I have been forward testing his method, with real money, on paper (see my Following MM thread), and I have developed an approximation. MM has changed his method since I created my approximation, but I am still happy with it.

It is difficult to really backtest MM's method. If you can find a way to do it, or to backtest an approximation, I would be really interested.

And MM's method is the best excuse for playing golf an amateur investor can have. When you have your limit orders in place of course . It is very low maintenance indeed.

Regards,

Karel

[jamiew]

Hey jamiew,

What do you do with a PhD in physical biological chemistry? I think I might have some friends that would be interested. What are the primary texts? Which university are you attending?

Best of luck.

Hmmm, a good question. Originally work within the biotech sector/pharmaceutical industry, however I’m a more interested in AI/programming/trading so i would rather go into any of these sectors.

What are the primary texts? not really sure what you asking, the area is "type I modification-restriction enzymes and a protein that over comes classical restriction by mimicking DNA"

Which university are you attending? Edinburgh University (UK)

[jamiew]

Karl

"41 winners (including open positions), average 14.0%
9 losers (including open positions), average -37.1%
50 total positions, average 4.8%"

What is the average life span of these trades (the average holding time)?
What stop loss are you using?
What is your exit strategy? as I cant seem to find one on MR M's home page.

[diogenes]

Hmmm, a good question. Originally work within the biotech sector/pharmaceutical industry, however I’m a more interested in AI/programming/trading so i would rather go into any of these sectors.

What are the primary texts? not really sure what you asking...

I assume he means something like what are the first year texts that nearly every program has, e.g., in math Rudin’s Mathematical Analysis and in economics Mas-Colell et al’s Microeconomic Theory.

[jamiew]

hmm well maybe its a different system, chemistry PhDs in UK require a lot of journal reading in the 1st 2nd and 3rd year, something i have never been that fond of. There are no standard text books, since if it was in a text book it would be explained and therefore the research would not be necessary. Are you sure you are not talking about undergrad courses?

Even a masters student will get most of the background reading from journal reviews.

A good place for research reading is pubmed, that's where most of the research journals I read are from. You can look up just about any scientific journal there.

I feel my academic back ground is being disputed? trust me its nothing to be proud of, I would trade it in for industrial experience any day. Its not so easy to get a job with no industrial experience!

[jamiew]

ah .. I think you do mean undergrad, Atkins has a good chemical physic book if that's what interested you? depends in what area you are interested in, oxford primers are the easiest way of learning from scratch, they don't go in to heavy and tell you what you need to know for each course

If any one wants to buy these books I'm quite willing to sell them, as I'm moving house now I need to get rid of all my stuff (hundreds of Sci-books)!!!! although I don't imagine any one here would want them.. and the cost of shipping probably woulnd't make it worth it?

[skiracer]

Jamiew,
I think the single most important facet that you must learn is the how and why of both the technical and fundamental sides of trading. How can anyone who doesn't understand the basics of either area as it can be applied to short, intermediate, or long term trading analysis comment on the subject or someone's elses discipline or strategy without a good understanding of the basics. I think that you have alot of studying to do on both sides of the analysis to get a better grip on what it takes to develope a successful strategy. I'm feeling that you are leaning heavily towards the A1 concept as a strategy that is superior in respect to whatever else you may get from others but are looking for other strategys that you will eventually prove to be not as good as this neural system that you are talking about. It would be good to develope some credentials to pyramid off of first. Then other more experienced, although academically less advanced, traders might feel more comfortable with what you are trying to convey or accompolish. Nothing personal but I want to see the beef, the money, and the credentials. I'm going out on a limb here, but I'm taking for granted that you're intelligent enough to absorb enough of what I am talking about to make us appreciate what you are trying to do and talk about here. It takes credentials an experience to make experienced traders take notice and listen. Without those two factors in your reportoire it makes it hard to listen or try to appreciate what you're putting forth.
Have you been trading or have any experience putting your hard-earned money on the line based on strategys that you have developed or learned from your research?

[diogenes]

hmm well maybe its a different system, chemistry PhDs in UK require a lot of journal reading in the 1st 2nd and 3rd year, something i have never been that fond of. There are no standard text books, since if it was in a test book it would be explained and therefore the research would not be necessary. Are you sure you are not talking about undergrad courses?

In the US system the first year of grad school is spent in seminars based on material from both books and articles. Then, you get to take comprehensive exams based on your first year classes.
If you fail to pass your exams after a given amount of attempts, then you are asked to leave.

Edit:

Those across "the pond" have a more interesting system for uni, I think.

[diogenes]

Jamiew,
I think the single most important facet that you must learn is the how and why of both the technical and fundamental sides of trading. How can anyone who doesn't understand the basics of either area as it can be applied to short, intermediate, or long term trading analysis comment on the subject or someone's elses discipline or strategy without a good understanding of the basics. I think that you have alot of studying to do on both sides of the analysis to get a better grip on what it takes to develope a successful strategy. I'm feeling that you are leaning heavily towards the A1 concept as a strategy that is superior in respect to whatever else you may get from others but are looking for other strategys that you will eventually prove to be not as good as this neural system that you are talking about. It would be good to develope some credentials to pyramid off of first. Then other more experienced, although academically less advanced, traders might feel more comfortable with what you are trying to convey or accompolish. Nothing personal but I want to see the beef, the money, and the credentials. I'm going out on a limb here, but I'm taking for granted that you're intelligent enough to absorb enough of what I am talking about to make us appreciate what you are trying to do and talk about here. It takes credentials an experience to make experienced traders take notice and listen. Without those two factors in your reportoire it makes it hard to listen or try to appreciate what you're putting forth.
Have you been trading or have any experience putting your hard-earned money on the line based on strategys that you have developed or learned from your research?

Actually, I would think that a quant role would be more in line with AI etc, than a trader.

My understanding, all second hand (kinda), is that the quants at banks etc do not mess with tech or fundamental analysis as we know it.

[jamiew]

Jamiew,
I think the single most important facet that you must learn is the how and why of both the technical and fundamental sides of trading. How can anyone who doesn't understand the basics of either area as it can be applied to short, intermediate, or long term trading analysis

gosh I feel like I'm going in circles. I'm not putting any thing on a plate, or suggesting any possible tips, trying to sell anything, or even giving strategies . I'm sure I would be wrong more often than right. I'm just here to find out as much as I can about Mr Ms system. I'm sure you are all vastly more experienced than me, of that I have no doubt, I am merely interested in it from an academic point of view as a direct comparison for a project I'm developing.

I think most academics fail in this area as they don't do there home work from the traders point of view, I have been studying this area for a few years, I'm interested to learn, so here I am, doing a bit of side homework.

comment on the subject or someone's else's discipline or strategy without a good understanding of the basics.

what do you mean by this, once again, I have made no comment on anyone's strategy, only asked for comments on it?

I'm feeling that you are leaning heavily towards the A1 concept as a strategy that is superior in respect to whatever else you may get from others but are looking for other strategies that you will eventually prove to be not as good as this neural system that you are talking about.

Its A.I. not a1, and no, im a scientist, a null or a proof is good for me. results are what I need, and if Mr Ms is better, I will develop in that direction.

I'm not criticising anyone's techniques, telling them how they should or shouldn't trade, or trying to prove that one system is better than the other, merely enquiring about Mr Ms techniques and as scientist I do not have any hard earned cash, but this is the life I chose.

please go through this thread again if any of this is unclear to you. However i feel I am once again unwelcome here and since I am offering nothing in return to you, i can understand.

However thank you for those people who have given me a bit of info, it has been most helpful, I hope to return the favour