You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Quantifying randomness: variance ratio
- Thread starter stephencrowley
- Start date
Yeah, very true, there are things related to profitible ideas that won't spill the beans. Of course no one really wants to talk about their core strategy.
Variance ratios, parksin numbers, exponents, etc are all just measures that can be used to screen strategies, do "post-moterm", etc.
I don't think anyone is suggesting these things will make money by themselves, they are used to give you a general idea of what types of strategies have the possibility of working on which data sets.
Indeed, the money is there in the data.. just depends on how you look at it. Just so happens that statistics is a really useful tool to do just this.
Variance ratios, parksin numbers, exponents, etc are all just measures that can be used to screen strategies, do "post-moterm", etc.
I don't think anyone is suggesting these things will make money by themselves, they are used to give you a general idea of what types of strategies have the possibility of working on which data sets.
Indeed, the money is there in the data.. just depends on how you look at it. Just so happens that statistics is a really useful tool to do just this.
Quote from nononsense:
Hi All,
Some make it seem as if the market has to be tackled somewhat like a problem in Statistical Mechanics.
In my experience, mathematics can indeed be useful. However, the main thing is to fret out some rather subtile properties well hidden from view of the masses, irrespective of how learned some may seem to be in this mass.
I'm sure that if you happen to stumble on something like this, you will not start with dispensing mathematical formulae related to your find.
The problem with whatever statistics you use is that you are always... late.
To succeed in trading, you mainly need specific criterions to heop you take a decision in the minute. Statistics are of great help when it comes to valuing what/which criterion is useful or not.
To succeed in trading, you mainly need specific criterions to heop you take a decision in the minute. Statistics are of great help when it comes to valuing what/which criterion is useful or not.
Right. The use of statistics inherently implies empirical analysis. Properly collected, derived, transformed, evaluated and implemented analysis of empirical data will provide an extremely valuable tool though: Anticipation with an adequately applied confidence measure.Quote from science_trader:
The problem with whatever statistics you use is that you are always... late.
To succeed in trading, you mainly need specific criterions to heop you take a decision in the minute. Statistics are of great help when it comes to valuing what/which criterion is useful or not.
Exactly. Of course gathering statistics is always after-the-fact. But, when you gather stats over a sufficient period of time and can apply robust statistical tests to see if the stats are stable, due to chance, whatever, can greatly help you apply confidence intervals for the forecasting into the future.. greatly controls and reduces risk when done properly.
Quote from ktmexc20:
Right. The use of statistics inherently implies empirical analysis. Properly collected, derived, transformed, evaluated and implemented analysis of empirical data will provide an extremely valuable tool though: Anticipation with an adequately applied confidence measure.
Quote from stephencrowley:
Exactly. Of course gathering statistics is always after-the-fact. But, when you gather stats over a sufficient period of time and can apply robust statistical tests to see if the stats are stable, due to chance, whatever, can greatly help you apply confidence intervals for the forecasting into the future.. greatly controls and reduces risk when done properly.
The problem in finance and trading is that the number of data is quite limited if you work with daily data. Once you have collected enough data to obtain a good statistical indicator, you don't have enough left to prove significance. In that case you have no other choice than working with high frequency data. In that framework you might have a chance to find something interesting.
Quote from stephencrowley:
Found a very good paper here..
Extending the variance ratio test to visualize structure in data: an application to the S&P 100 Index.pdf
This is quite a good paper.
The modifed varience ratio works very well.
Makes a change to to get something simple and practically useful from the academics.
It comes down to knowing where to look, how to impliment. Knowledge, skills, experience. It's an art along side the science in that creativity plays a big part, imo.Quote from science_trader:
The problem in finance and trading is that the number of data is quite limited if you work with daily data. Once you have collected enough data to obtain a good statistical indicator, you don't have enough left to prove significance. In that case you have no other choice than working with high frequency data. In that framework you might have a chance to find something interesting.
Yes of course, I laugh when I still see people trying to find "patterns" in daily closing prices, open, high, low, etc etc. It's kind of sad really.
Quote from science_trader:
The problem in finance and trading is that the number of data is quite limited if you work with daily data. Once you have collected enough data to obtain a good statistical indicator, you don't have enough left to prove significance. In that case you have no other choice than working with high frequency data. In that framework you might have a chance to find something interesting.
Indeed, I've found the modified VR test as well to be really useful.. it can also be used to find the optimal length of rolling windows or moving avg lengths.. although I haven't read any papers using it for that.
Many traders seem to deride any academic stuff as "useless" but there is a lot of great stuff out there.. but just as with random sequences you need to filter the noise out of academia as well.
Many traders seem to deride any academic stuff as "useless" but there is a lot of great stuff out there.. but just as with random sequences you need to filter the noise out of academia as well.
Quote from squeeze:
This is quite a good paper.
The modified variance ratio works very well.
Makes a change to to get something simple and practically useful from the academics.