A place for redditors to discuss quantitative trading, statistical methods, econometrics, programming, implementation, automated strategies, and bounce ideas off each other for constructive criticism. Feel free to submit papers/links of things you find interesting.
Hello and welcome to the MLFX (Machine Learning Forex) sub! MLFX is the culmination of 10 years of academic research into using machine learning to create predictive algorithms for the Forex markets. Recently we've put more focus on creating a consumer product and we're looking for people to help us with Beta testing the system. There's still a decent amount of work for us to do before we'd consider MLFX v1.0 to be "feature complete" but we're already getting some pretty amazing results. Now comes the test of consistency! This is what our signals look like (27/08/2020) https://preview.redd.it/jvwqvw0oahj51.jpg?width=1128&format=pjpg&auto=webp&s=1c80e3390d7e63bddbda2987ba613cb9adb12d62 We're currently working on a way to check every trade after the fact and determine if it was a winner or loser, we will also plot each individual trade on a chart (like below) as we already have that functionality. Red = SL | Blue = TP | Solid = Result Our algorithm uses multi-agent systems and evolutionary algorithms to create agents (traders) that have been specifically optimized to perform in that current market. I'm not going to go too in-depth in this post if you want to find out more you can here: https://www.mlforex.com/algorithm/agents If you're interested in assisting with beta testing please join our Telegram group and express your interest. https://t.me/joinchat/IFyZTEw_uhEHe0Oo5uSAVw
what is the learning pathway to make these machine learning forex strategies ?
My knowledge of MQL4 didn't help me with the patterns that you use your eye to recognize it like ( head and shoulders, divergence, fibonacci retracement ... etc) What is the learning path for me to use machine learning to feed the robot with historical data and point out the range of the pattern example " from Hour 14 at day 14/5/2019 to hour 21 at day 17/5/2019" that is a head and shoulder (I pick these ranges through manual backtest) and so on through the whole data set so the algorithm can learn to recognize the pattern through these examples and perform trading on future data? I already started to learn python but I don't want to get distracted with learning stuff that I won't need to make these machine learning robots so I better ask the experts first.
Hi, I'm making a machine learning trading bot. Using the NEAT algorithm, the simulation starts with a bunch of traders trying random things and the best surviving traders move on to the next generation. The trader instances will be awarded from monetary gain and die when they are broke. It's likely not entirely possible to create a successful trading AI, but that's not the goal. The goal is to make an AI that is better than me. I've made a "trading platform" from scratch in javascript and implemented the NEAT algorithm, after days of work. I now have to decide what data the AI should look at to determine its trading actions. So far I have a trend line, wether a candle is red or green, and the current unrealized gains. It's enough information for the AI to actually decide trades, but it's not making very good decisions right now. So here's a question for you guys, what sort of trading data would you present for the "eyes" of an AI? Maybe I should try to add trading volume?
e-Forex Magazine | Machine learning stirs up competition in FX Algo Trading
fintech #algotrading #hedgefunds #quants #hft
e-Forex Magazine | Machine learning stirs up competition in FX Algo TradingIvy Schmerken Though capital markets firms have been adopting artificial intelligence and machine learning to train algorithms for equity trading, recently this trend has expanded to foreign exchange. Ivy Schmerken, Editorial Director at Flextrade Systems, has written widely about this topic and we asked her to revisit it for e-Forex. By crunching vast quantities of data by computer, machine learning algorithms can identify hidden patterns in past data and learn to forecast stock market returns or FX currency pairs. Large banks have been investing millions into advanced technologies such as AI and machine learning to capture a bigger share of the algo trading market. JP Morgan developed a new algorithm dubbed DNA – or Deep Neural Network for Algo Execution to merge what a multitude of algos do into a single strategy, allowing the framework to decide how a client’s order should be executed, reported Reuters ..... Continue reading at: https://www.e-forex.net/articles/aug-2020-machine-learning-stirs-up-competition-in-fx-algo-trading.html
Anyone experienced with Statistical/Machine/Deep Learning in Forex?
After fiddling around with EAs, I got interested in ML. Maybe it was just because of the hype, or because way back in college I did a project on neural networks, I was interested to see if ML/Deep Learning has any application in trading. According to the guy who wrote the book on Deep Learning in Python, there was a disclaimer early on that DL is not applicable to markets, because of their inherently unpredictable nature. I took his advice and stopped knocking on that door. But now, I am finding books on Statistical Learning, and these authors don't shy away from claiming that Stock Markets and Finance in general benefit from Machine Learning concepts or thereabouts. They are hinting at it, at least (I just got started). So, my question to anyone in the community who has experience with this: is there any scope for a retail trader to gain some insight through ML/DL/SL? Even if it is about getting better at identifying patterns, or finding probabilities of possible price direction, or better optimisation of regression analysis. Anything that makes it worth it?
Anyone experienced with Statistical/Machine/Deep Learning in Forex?
After fiddling around with EAs, I got interested in ML. Maybe it was just because of the hype, or because way back in college I did a project on neural networks, I was interested to see if ML/DL has any application in trading. According to the guy who wrote the book on Deep Learning in Python (along with creating the Keras librarby in Python), there was a disclaimer early on that DL is not applicable to markets, because of their inherently unpredictable nature. I took his advice and stopped knocking on that door. But now, I am finding books on Statistical Learning (also another advanced one by these authors), and these authors don't shy away from claiming that Stock Markets and Finance in general benefit from Machine Learning concepts or thereabouts. They are hinting at it, at least (I just got started). So, my question to anyone in the community who has experience with this: is there any scope for a retail trader to gain some insight through ML/DL/SL? Even if it is about getting better at identifying patterns, or finding probabilities of possible price direction, or better optimisation of regression analysis. Anything that makes it worth it?
Help with Machine learning applied to Forex Trading
Hello boys, i'm a python coder with 4 years of experience behind my back and a Forex trader. I know almost nothing about AI so i've come here to ask some guidance about an idea that came up in my mind.I've trained a lot in the Forex market until i got some positive edge in my entries( i don't use indicators or whatever) and i wanted to create a bot that can replicate my trading style. I am unable right now to put down specific rules that would allow me to automate the strategy so i was wondering if there is a way to teach a bot my trading style with AI. What i mean is: assuming I am able to feed the bot with a set of entry points (data)(like in the screenshot), how can I make the bot try to learn and replicate my trading style ? Where do i start? Thanks for the replies. https://preview.redd.it/tl0xlqfwm7d21.png?width=2238&format=png&auto=webp&s=72e3a7e999c1571989281baca696c9cd69f5079c
Anyone who's interested in algo trading and machine learning should check out this guy's channel. He's also got some other useful stuff for forex on there as well.
FREE: Build a Successful Ecommerce, MATLAB for Engineers, Neuro-linguistic Programming, Sıfırdan İleri Seviyeye PHP, Excel Basics [2020] + Advanced in Ms Excel 2019 & Office 365, Complete Guide to Pinterest, Vue JS and Firebase, Complete Modern JavaScript Firebase, Learn C Programming By Examples
FREE: Build a Successful Ecommerce, MATLAB for Engineers, Neuro-linguistic Programming, Sıfırdan İleri Seviyeye PHP, Excel Basics [2020] + Advanced in Ms Excel 2019 & Office 365, Complete Guide to Pinterest, Vue JS and Firebase, Complete Modern JavaScript Firebase, Learn C Programming By Examples
eToro: impressions, doubts and (ignored) lessons from copy trading
(no promotional content, no affiliate links) Hi, exactly four years ago, I started copying eToro investors / traders that I selected using the broker's built-in search engine (profitable in last two years, already being copied by others), followed by manual filtering, to take into account fluctuations in yearly returns, composition of their portfolios etc. With that, I got a list of 10 people whom I started to copy on a demo account: https://drive.google.com/file/d/1u52f0XHfr-LauIscKcFDYF0yGTTUr6VY/view?usp=sharing In the screenshot you can see that in case of the first two of them the amount invested was $10,000, while for the rest it was just $100. This is because I started copying the first two a couple of weeks earlier; eventually I changed this into $100 the same day I made the screenshot and this is when my calculations start - so this thing is irrelevant, I just cannot travel in time to make another screenshot. What I did after that? Well, within the next six weeks my profits oscillated between -$11 and +$9.50 (the biggest profit was on Nov 9, a day after US presidential elections). I found this "boring" and discontinued experimenting with copy trading. Today I looked back at those ten traders. Here is what I found. Firstly, seven of them are not with eToro anymore; investorNo1, Simple-Stock-Mkt, tradingrelax, 4exPirate, primit, Gallojack, xjurokx. The other three traders are:
toppertrader: not being copied by anyone and for a good reason: his loss this year alone is 61.16%!
Jean-marcLenfant: copied by only 67 people; his loss this year is -1.09% but in general he is quite successful, with yearly profits ranging from 3.57% to 7.32%.
Girem2: he has no copiers, his profit this year is 41.45% but in 2018 he experienced a loss of 83.15%!
My observations and thoughts are as follows:
Seven out of ten traders are not with eToro anymore, which makes me wonder why. I have no proof but my guess is they simply performed poorly, lost their copiers and closed their accounts. This is already alarming but what if they opened another account? Or, even worse, multiple accounts? They could be investing small money and try different risky approaches, hoping that at least one account will turn out profitable in the long turn, attracting potential copiers. (I'm not claiming that those 7 particular traders did this, it's just my general suspicion regarding some of eToro traders)
I'm unable to calculate what would be my profit if I never stopped copying them, because I cannot check at what day and with what profit those seven traders left eToro. I'm guessing this would be an immense loss. On the other hand, considering the three traders who are still with eToro, I would lose more than a quarter of my assets!
What now? I must be a quite adventurous person or at least an incorrigible optimist, because a month ago (exactly on Aug 26th) I started copying three traders with real money. Here is who they are. rubymza (Heloise Greeff)
invests in stocks, with GOOG, INTC, BLDP, MA, MSFT, AMZN, V, MU, IBM and NXPI making up 50.3% of her portfolio (allocation of each of them is in between 3.02% and 6.85%)
active since 2016 (only the year 2016 ended with a loss)
has 3044 copiers and $2M-$5M of copy assets under management
strategy (her own words): "My investing strategy focusses mostly on US indices, tech and pharma, promising future (5-10years) growth. My trades are based on technical analysis using machine learning to understand patterns and trends in the markets. I prefer to keep a diverse portfolio to spread risk while achieving great returns."
he is a Forex trader, making typically 21 trades per week; his favorite currency pairs are EURCHF (12% of trades), CADCHF and GBPUSD; the trades, however, typically make up below 5% of his portfolio (at least whenever I'm checking it), making most of my funds unused
active since January 2017: surprisingly enough, he has every single month profitable, though monthly profits are in the range of 0.03% to 3.34% only
has 8977 copiers and more than $5M of copy assets under management
strategy (his own words): "I monitor currency pairs all day to find the best entry. There is some management/scaling position for perfect entry. The risk control is a big part of my strategy," (quite vague, to be honest)
commodities compose 76% of his portfolio and his favorite assets are Gold and Oil (at the moment, Gold makes up half of invested amount)
active since July 2016, with the following yearly profits, starting from 2016: 6.56%, 10.05%, 13.09%, 32.26% and -2.03% (the current year)
has 1493 copiers and $1M-$2M of copy assets under management
strategy (his own words): "My system is based on patterns, and a variety of technical analysis tools and some fundamental analysis. I primarily trade in commodities. " (quite vague as well)
own experience: my profit with rayvahey is 2.56%
What was my strategy to hand-pick these particular traders? First I did some basic scanning using eToro's built-in search engine. The most important filter was that the trader was profitable within the last two years: unfortunately, eToro does not allow to reach details of earlier performance automatically. To know how the trader performed before 2019, I had to look at stats in the profile of each of them. I was also taking into account how often they trade (to avoid those who do only a couple of trades yearly), whether they were trading recently and whether they write posts regularly in their feed. With this, I got a list of fifteen candidates to copy:
olivierdanvel
marianopardo
rubymza
rayvahey
martidg97
overit
misterg23
knw500
jianswang
reinhardtcoetzee
miyoshi
alderique
jarodd76
benson9904
big-profits
As you already know, I finally chose three of them. Rubymza seemed to be the most trustworthy stock trader, based on profits, posts feed and regular trading, among other things. Regarding OlivierDanvel, his uniqueness is the ability to record continuous profits with the Forex market. Finally, with rayvahey I wanted to increase my exposure to the commodities market. Wish me good luck! Michael P.S. You might find those copy-trading related readings interesting:
Just some inspirations / reminders on strategy development
I just talk about really major pairs like EURUSD, USDJPY, etc. Forget about catching a trend, if you wanna trade trends, commodities, stocks, index funds trend way better, a lot more opportunities than forex. Major currencies range at least 70% of the time, if not more. Learn how to make money from ranging markets and hold a trend once you catch it. The biggest purpose of currency is for settling transactions, not for scalping profits. That's why it doesn't trend (aka remaining stable). Stability is why a currency being "major". Therefore most indicators don't work well with these currencies because first they are not designed for forex, second most of them only tell trends or overbought/oversold. Unless you are Soros or central banks etc no major currency can be overbought or oversold. Take advantage of "fakeout" (I still wonder if it's the right way to call it so, Trump's Tweets are one of the sources IMO). Accept the fact that it happens and think about how to profit from it. Market makers and big banks are also just market players, even though very much bigger, they are also profiting from each other. If you can't beat them, join them. Choppy market is still better than a still market. No market maker cares about support or resistance. Like no insider or institutional money (i mean human not machines) would spend hours and hours on charts drawing trend lines before they place an order. Why would you? Planning how to react in different scenarios after a position is opened is much better than trying to act like a crystal ball by looking at history when you trade something that ranges most of the time. The moment you observe a trend, chances are the trend is (almost) over. Even if things are against you, most of the time you can turn it to break even without using lots of margin. (Most news are just as big as baby's cough.) But still, very few news are really big (911, fukushima, brexit, covid, name it), don't ignore the news completely. Money management is very important. Most traders (of course including many of those on reddit) just talk about how to make an entry but seldom talk about how to manage an already open position or how to close a trade. The latter is way more important than the former. Besides japanese candlesticks, there are a lot more charting options out there. Be creative and know what you are trading to the deepest !
40 + Free Courses : Chinese Characters, Brand Awareness, German Language, Earn Passive Income, Ultimate Content Writing Masterclass, Content Marketing Strategy, Big Data on Amazon web services, Mommy's Baby Guide, Practical Database Course, Relative Strength index,Interviewing skills and more
40 + Free Courses : Chinese Characters, Brand Awareness, German Language, Earn Passive Income, Ultimate Content Writing Masterclass, Content Marketing Strategy, Big Data on Amazon web services, Mommy's Baby Guide, Practical Database Course, Relative Strength index,Interviewing skills and more
40 Free and Popular Discounted: Vue.js 3 , Facebook Lead, MS-Excel, How to play Piano, Affinity, Mastering Agile & Scrum, CCA 175 Spark and Hadoop Developer, Oracle APEX, Flask Framework, Learn Python, Adobe Creative Cloud , Learn Java by building 17 & More
Forex and stock trading automation what do you think?
Hello, the reason I'm opening this discussion is that for a while I've been thinking about building forex/stock trading automation system(something like a bot but in a more sophisticated way). I have a decent python experience with over 1 year of hands-on deep learning and general purpose projects so it is valid to say I'm quite experienced with tensorflow 2.x, keras, numpy, pandas, matplotlib and most of standard python libraries. I'm currently learning c++ for performance needs that cannot be met with python alone. I've been intending to use my background to create a trading system I have not thought of specifics but my general overview of a problem solution might be to create a combination of deep learning models and maybe some reinforcement learning techniques as well(which something I'm currently learning). The question is do you think of a criteria to get best outcomes in terms of prediction accuracy and be able to deliver something that might turn profitable at some point? I have not started working on the project yet but when I read about the topic in kaggle and general machine learning forums, the idea of forex bots/ auto-trading systems is sometimes met with skepticism: some argue it's nearly impossible to create something useful and some others claim that they were successful to create deep learning models (my best guess is RNN/LSTM architectures) with variable accuracy(60-90%). I don't think experimentation will hurt, I mean I can start working something out and figure out the best results for myself but I thought it's a good idea to ask for guidance from those who have similabetter background/experience as/than myself as well as others who might have tried already and hear the feedback first. What do you think?
Eduonix Sitewide: Buy 1 and Get 2nd at 50% OFF + 50% Extra Discount. Mix Match with Any of these Mighty Bundles or E- Degrees or Courses Sitewide Use Code : APPLY50
The installation of machine learning algorithms in the FoRex trading online market can automatically make the transactions of buying/selling. All the transactions in the experiment are performed ... machine learning and investment options. This knowledge is important for the following sections. Chapter5presents our algorithm and explains our framework, Learnstream, which as far as we know is the rst system capable of online machine learning in a streaming manor. In Chapter6we adduce the experimental results based on three datasets (two foreign MACHINE LEARNING FOREX TOOLS signal forecasting. Stop wasting time trying to find the perfect strategy. Start earning with machine learning today! Try our beta, it's FREE. TIME LEFT. 00. Days. 00. Hrs. 00. Min. 00. Sec. Use the power of machine learning to achieve your financial goals. Beginner-friendly. using machine learning (ML) to the prediction of monthly stock returns. Our intention is to implement machine learn-ing methods in a relatively unexplored asset class: foreign exchange (FX). The objective of this paper is to produce directional FX By Reashikaa Verma, ParallelDots “What we want is a machine that can learn from experience“ Alan Turing. There is no doubt that Machine Learning has become one of the most popular topics nowadays. According to a study, Machine Learning Engineer was voted one of the best jobs in the U.S. in 2019. Looking at this trend, we have compiled a list of some of the best (and free) machine learning ...
How to Build a Winning Machine Learning FOREX Strategy in Python: Feature Creation (3)
11 videos Play all Build a Winning Machine Learning Algorithmic FOREX Strategy! PythonParseltongue The Top 5 Technical Indicators for Profitable Trading - Duration: 7:50. Quick look at our Data: Machine learning for Stocks and Forex Technical Analysis by sentdex. 11:00. Machine Learning and Pattern Recognition for Stocks and Forex Part 3 by sentdex. 6:26. Percent Change: Machine Learning for Automated Trading in Forex and Stocks Part 4 In this video we are going learn how about the various sources for historical FOREX data. Primarily, we will be using data from Dukascopy bank. There are man... This is the first video in the series where we will start to tackle the creation of financial feature functions that we will use as indicators for a machine ... How to Build a Winning Machine Learning FOREX Strategy in Python: Feature Creation (4) - Duration: 19:04. PythonParseltongue 9,141 views. 19:04. But what is the Fourier Transform? A visual ...
