Trading and Statistics
- Thread starter ChkitOut
- Start date
If you want to get a bit more sophisticated, there is always the field of "distribution-free" statistics. Here's a reasonable reference:
http://www.amazon.com/Distribution-...=sr_1_2?ie=UTF8&s=books&qid=1214608167&sr=8-2
http://www.amazon.com/Distribution-...=sr_1_2?ie=UTF8&s=books&qid=1214608167&sr=8-2
From what I wrote before, I want to underscore the value of keeping your own timed prices data. I always have. Setting it up in tabulated formats and maintaining it is time consuming while also you have to work out an effective array of parameters for compiling and managing such extensive datasets. All of this is to assist accurate daytrading; you may not realize it but it gives you the model to track the market live and plot what will happen next. You therefore are playing in the market with the benefit of certainty. However this may well be beyond the considerations of any amateur trader whether part time or full time. At least you should know that being a professional and professionally exploiting markets puts the professional player way ahead of almost all independent traders and their outlooks.Quote from Cheese:
What is it that you see? What is that never gets discussed on ET? Price and? Price and time. The vertical axis is price and the lateral axis is time. Price by time. When? Just when are upmoves and downmoves occuring in the daily menu? What is their order and magnitude on the time axis? Certainly it is crucial in day trading to know the metrics of the market you are using, if you want to extract most of what the market daily offers. The reason all of this doesn't matter is because you do not keep time data and cannot otherwise access precise time data from back data. And you wouldn't have the time to set it up and tabulate it anyhow.
When I develop a trading system, I find a reason for why the market should do something, and why it should continue doing it. Then I develop a system to exploit that reasoning in a statistically sound manner.
The common notion about fundamental data is that is too abstract to quantify. My theory is that anything - when analyzed enough - can be broken down into a system of some sort. To see some ways you can utilize fundamental research in a systematic fashion, look into "econometrics". It is perfectly possible to utilize fundamental reasoning in your system development, and still take a statistical approach.
It is possible to develop a trading system that stays profitable over time without intentionally factoring in a fundamental premise. If you have accomplished this, I am very happy for you. However, take a look at that system deeply, and you will find that the reason why it stays consistent is due to an underlying fundamental factor staying consistent.
The chances of a system designed strictly from a technical standpoint aligning profitably with a fundamentally sound premise is substantially increased when the fundamental premise is thought of first. Failing to consciously factor in fundamental reasoning in the development process of a system is to rely on chance coincidence for consistency and system viability.
When one of my systems is on a profitable streak, or a losing streak, I do not just regard it as a statistical anomaly. I know why it is happening. I am sure that a lot of "technicals-only" quants made money last year, but I guarantee that a greater than normal amount lost money too during the high volatility conditions over the past year. This is because they failed to consciously factor in a fundamental premise, so when the market conditions shifted in a manner therefore voiding the fundamental premises of their systems which once made them profitable (if even unintentionally), they were not able to realize it.
In Summary:
Being able to build a profitable system is not enough. One must understand why it is profitable to keep it that why. Consciously setting a fundamental foundation to your system increases your chances of being able to do this.
Sure, I use correlation solely in my trading:
I get up in the morning and go to the toilet to piss. I have a stop watch to measure the seconds I pissed, I have proprietary glass to measure the color of my urine (on a scale from 1 to 7), and I have proprietary device to measure the intensity of the smell of my urine. With the three measurements, I decide on the number of contracts I will trade for the day, the long or short choices I will use, and the number of trades with the holding time for each of them. Pure correlation, nothing else.
Believe it or not, the result is good.
i trade a mean reversion strategy based on 30 day moving averages and standard deviations. Very simple in concept. Buy and short at 2 std. deviations. I always close or take profits on half the posion at 1 std. deviation and the rest at the moving average mean.
I never take losses but just keep adding layers in a systematic way until a reversal occurs and then begin to peel back the layers
I only trade ETFs in order to eliminate company risk. Some traders like baskets, but i prefer the ETF approach. I track over 100 etfs and generate buy/short signals based on end of day prices. At any given time i may have 30 positions open.
Love to buy in after hours to take advantage of mass liquidations and to take advantage of overnight gaps in the markets at the open.
just brief description, Hope it helps.
Good luck to all.
I never take losses but just keep adding layers in a systematic way until a reversal occurs and then begin to peel back the layers
I only trade ETFs in order to eliminate company risk. Some traders like baskets, but i prefer the ETF approach. I track over 100 etfs and generate buy/short signals based on end of day prices. At any given time i may have 30 positions open.
Love to buy in after hours to take advantage of mass liquidations and to take advantage of overnight gaps in the markets at the open.
just brief description, Hope it helps.
Good luck to all.
Which instruments you trade with this approach ?
eminis mainly.
You may not believe me, but my method has its validity in removing human errors from trading. By eliminating human intervention, my correlation method drastically increases odds of winning. It's not that my urine smell, color and pissing length have any inherent relation with the market. They don't. As a matter of fact, many indicators used by traders have no relation with where the market is going to go in the next minute, hour, day or year.