- Search Crypto Craft
-
FXEZ replied May 15, 2020I spent a long time chasing cointegration. It's a long term relationship and it can fall apart, I think most of us are more interested in short term relationships and in things that don't fall apart. But I keep coming back to the triangular ...
Real or Algo
-
FXEZ replied May 14, 2020There is a triangular relationship that runs at the core of all FX because we don't trade currencies, but rather currency pairs. So E/G = E/U * U/G (note the U cancels out). So if you don't have EG quotes but do have quotes for EU and GU, you can ...
Real or Algo
-
FXEZ replied May 14, 2020Your calcs are off in your example. If you: buy 100k EU @ 1.08593 then you own 100k EUR and are short 108,593 USD. So for GBPUSD you need to sell 108,593 worth of GBP or 108593/1.24682 approx 87096 GBPUSD. Then this just leaves EURGBP - where you ...
Real or Algo
-
FXEZ replied May 8, 2020The snippet above is correct for all cases. If you introduce a random number generator to make trades then the trades are independent. Any calculation where odds change implies dependency, like the odds changing in blackjack when the deck is running ...
Random Trading EA with success
-
FXEZ replied Apr 27, 2020Just to reiterate - the statistics appear to just be descriptive rather than predictive. It seems if you put enough conditions together with and/or you can end up with really high non-predictive stats because you're basically covering all the ...
The similarity system
-
FXEZ replied Apr 27, 2020If I were doing it over I'd focus more on Python due to the huge speed advantage it has over R and the fact that so many of the new tools are being developed there vs. R. Now that visual studio 2019 has native python support, I really don't have any ...
The similarity system
-
FXEZ replied Apr 27, 2020Anticre: I replicated your 94% and 76% on close prices but ultimately there is no predictive power in these probabilities - meaning there is no way to use them to get that probability out of the next bar. So this quote above is how I think this ...
The similarity system
-
FXEZ replied Apr 27, 2020Thanks for letting me know - it's been 6 years since my original post and I only by chance saw your reply - having given up subscription to this thread long ago. So now I'm playing again in R. Thanks a lot!

The similarity system
-
FXEZ replied Apr 14, 2020Most losing strategies lose, not because they have a large negative edge, but because they have no edge at all. Transaction costs eat the system or its reverse to shreds. Both the system and its opposite typically lose over the long run due to ...
A or B
-
FXEZ replied Apr 1, 2020I don't think any of the original principals of the company are still running things, and it shows. The current crew are a bunch of half-witted rent seekers. Why care for your reputation when there is literally no competition? Every policy change by ...
OANDA Discussion
-
FXEZ replied Mar 31, 2020I've been away from FF for a long time. This topic was dealt with in excruciating detail way back then, and apparently still is of interest to someone today. I guess some things never change. Some in this thread have posted examples of ...
Forex is random
-
FXEZ replied Dec 14, 2017Thanks for posting this Sis.yphus. I see this as a thought experiment. If I have time I may use this as a starting point for performing some research, which if interesting I'll try to post some of the highlights.
Do you manage losers and let your winners run?
-
FXEZ replied Sep 27, 2017Yes you're correct that your analysis tests the reference point concept. And I think what my previous post shows relative to the reference point concept is that resetting the reference point more often is generally superior to using the same ...
Higher Edge within a Single Candlestick
-
FXEZ replied Sep 27, 2017Nice job AntiCre. It is rare to see code posted to test hypotheses, and it was nice to see this code in a language that I'm conversant, R. I changed your code slightly to allow for testing different values for games and rounds (but probably not ...
Higher Edge within a Single Candlestick
-
FXEZ replied Mar 31, 2017CRB creation logic: After you select a CRB range amount (like 20 pips), the logic for constant range bars (CRB) is quite simple and goes something like: 1) start a new bar at the next/current price 2) keep track of bar's range (high and low prices) ...
Machine Learning with algoTraderJo
-
FXEZ replied Mar 31, 2017PipMeUp is correct that constant range charts converge from different starting points. From my tests you can start multiple constant range charts at different points and in approximately 3-5 bars they will synchronize. The number of bars to ...
Machine Learning with algoTraderJo
-
FXEZ replied Feb 22, 2017I've come to the conclusion that with Oanda you can't trade using market orders due to the frequent news and non-news spread "spikes" you cite. But with (marketable) limit orders based on spread midpoint it is doable, though you miss some fills.
OANDA Discussion
-
FXEZ replied Feb 6, 2017Hey GEfx, thanks for your post. Can you comment on how forex.com is in terms of fills / slippage and spreads vs say Oanda (with which I am familiar) from your experience? Thanks.
FXCM Discussion
-
FXEZ replied Jan 17, 2017Calculus, on the lower right corner of the TWS there is a green button that says "Data" right next to the clock. Click it. It shows your data connections. Next to the one called Primary (under the middle "Purpose" column) note the Farm name. That is ...
Interactive Brokers - experiences and opinions