FXEZ's Quantitative Research In R | Page 6 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Quantitative Approach to Backtesting 24 replies

applications using quantitative data on the NYSE 0 replies

FXEZ's E/J Gravy Train 26 replies

The Shark's Trading Book - Quantitative Trading 2 replies

Quantitative Qualitative Estimation for MT4? 0 replies

Attachments: FXEZ's Quantitative Research In R
Exit Attachments

FXEZ's Quantitative Research In R

  • #101
  • Apr 20, 2017 2:21pm Apr 20, 2017 2:21pm
Quoting alphaomega
What you don't realize is that your system never worked in the first place, and you are just one lucky poor soul for not knowing the difference!
I got directed here to your post because someone was pointing to it on another thread. Thanks for the post, it's interesting to read. I'll try to keep this short and I have to ask, how can you be sure that the system never worked in the first place?

What I mean is you're speaking as if you somehow know of unforeseen future events. Maybe what you assert is a losing strategy now, is really just a losing period in a slightly longer term winning strategy, and so on.

Of course it would be interesting if you could formally prove that all strategies are necessarily losing strategies by definition.

I think I realize you're essentially just presenting an intuitive or common sense explanation for things. It's just that I personally get triggered when smart people categorically assert they know of something with certainty in this forex business when in fact, I don't think such a thing is possible in this business because we are in the here and now and the future, with all its Black Swans, White Swans, etc., hasn't occurred yet. I mean, you could endlessly settle on one idea only to find out it was incorrect and then switch ideas, only to find out your original idea made more sense, etc.

Anyway, I wished it were possible to reveal a deeper truth but all these statistical methods and computations are just human creations to help us feel better and shouldn't be taken literally to the nth degree.

My thought is that the essence of what is important is making money. The positive results, no matter how you get there, is essentially what matters for our finite life in order to make our remaining life a little bit easier.
  • #102
  • May 13, 2018 8:39am May 13, 2018 8:39am
Some nice ideas in here ,lets keep them alive
  • #103
  • May 28, 2019 8:01am May 28, 2019 8:01am
Quoting FXEZ
I keep writing these test systems in case I stumble upon one that actually does make consistent green on random data. I still believe in Santa Claus too.
Hi FXEZ, I know this is an older thread and I know that you haven't been active @FF for quite some time - but here you go: It is possible to be successful on random data. However, the data must be stationary and symmetrically distributed around the E(X). The idea is from the link above, the implementation is my own ...

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 19 KB


Inserted Code
par(mfrow=c(2,1))
average <- 1.2
#x <- rnorm(1000, average, 0.01)
x <- rlogis(1000, average, 0.01)
plot(x, type="l", xlab="close price = next open price", ylab="fictive price value")
abline(h=average, col="blue")
profit <- NULL
for (i in 2:999) {
  if (x[i]<average) {
    profit[i] <- x[i+1]-x[i]
  }
  else if (x[i]>average) {
    profit[i] <- x[i]-x[i+1]
  }
  else {
    profit[i] <- 0
  }
}
profit <- profit[-c(1,1000)]
plot(cumsum(profit), type="l", xlab="trade number", ylab="equity")
  • #104
  • Last Post: Jul 3, 2020 4:58am Jul 3, 2020 4:58am
Quoting Dr.Phoenix
{quote} Correct, PipMeUP Guys, that is why it is time here to remind of markets are neither uniformly nor normally distributed on the whole. However, of course the RW assumption could be used as the first step to discover the truth (that is out there). One day I found some researchings showing that the most probable distribution to a market behaviour was the Levy distribution. Below you can find a link to a page where all known distributions are formalized, described and coded (in Russian). These may give you a new breath for further discoverings....
ok this wot paying attention
0 traders viewing now