Search results

Hi Hugin, Most state based models that I've worked with or find in literature/internet fall into the 1st description; i.e. conditional gating or regime modelling, whereby some systems have a better chance of thriving under a certain regime. That is not to say the other approach might not be...

The same problem persists when discussing things like markov chains. Plenty of people have a keen interest in hidden markov models, without understanding that something like a visible markov model might be far better suited to solving their needs (usually because they want to feed garbage into a...

There are several well-known methods available to avoid overfit; regularization and introducing complexity penalties into the fitness function come to mind. Regarding market states, nobody seems to discuss much how to even define the states rather than learn them. A much better discussion...

Wal-Mart slashing ipad/ipod prices... http://money.cnn.com/2012/12/14/technology/mobile/walmart-apple-iphone-ipad-sale/index.html?iid=HP_LN&hpt=hp_t3

A slightly different angle to ponder if you will... One of the things I distinctly recall about the crash of 08, is that it was preceded by regulation changes more favorable to shorts e.g. removal of certain collars, short uptick rule, etc... A cynic might say the floodgates were opened for...

Just to clarify... ARIMA can model non stationary series. Hence, the I=Integrated in AR-I-MA.

One of the best.:) http://www.amazon.com/Financial-Modeling-Simon-Benninga/dp/0262026287/ref=sr_1_1?s=books&ie=UTF8&qid=1351824005&sr=1-1

The idea is that you are generating sequences of random binary signals over the duration of the asset you want to compare your own signals to. The 1s and -1s represent long and short signals which you can use as random long/short signals. Each time you substitute one of the sequences as your buy...

I'd love to see a cogent thread... unfortunately, there are few who want to contribute any actual work... :( Dozens of books. I often re-read PDQ statistics (Streiner) as a short coffee table refresher. If you can grasp the concepts within, it's a great fast practical read. Till...

Excel. =IF(NORMSINV(RAND())>=0,1,-1) Run many times will give a sampling distribution of signals to compare. It makes more sense to run against same period as system validation set.

My personal experience has been that nothing is obvious except for hindsight and maybe inside information. But that's just me.

nice job.:) Appears very clean with good font choice, IMO. Green color highlight fonts kind of clash though; maybe darker?

R About only advantage to matlab is some signal processing support-- but, R has huge support and development behind it as well as financial libraries available.

You can look into stepwise regression.

I was hoping for a more statistical assessment. Other than giving a few examples, it would be nice to see a collection of statistics related to your observations. You might have interest to track down toby crabel's book for similar ideas (that have been explored long ago).

I think it's a good endeavor. One thing I've observed, however, is HL order tends to be efficient for reasonably long samples. Have your observations or methods shown otherwise? ..also, don't know if this belongs under journals.

Do you know if they provide any kind of historical database of the information? TIA

http://www.r-project.org/ Do you have all of the back adjusted contract data for his examples (40 yrs or max)? What source are you using to obtain them?

It can be done in R. If you want them to line up as in his illustrations, you have to be careful how you want to handle rescaling the real time data, because you can't scale 0 to 100 in advance. As the series unfolds, they may not align nicely.

Hi Craig, From your example, I think you are asking for relative adjacent returns? try : library(quantmod) x<-c(1,2,3,6) 100*Delt(x) returns: Delt.1.arithmetic [1,] NA [2,] 100 [3,] 50 [4,] 100