Search
× Search
Friday, February 7, 2025

Archived Discussions

Recent member discussions

The Algorithmic Traders' Association prides itself on providing a forum for the publication and dissemination of its members' white papers, research, reflections, works in progress, and other contributions. Please Note that archive searches and some of our members' publications are reserved for members only, so please log in or sign up to gain the most from our members' contributions.

The reason why forex Algo’s fail ........and a possible solution

photo

 Peter Brennan, .

 Wednesday, April 13, 2016

The reason why forex Algo's fail ........and a possible solution


Print

8 comments on article "The reason why forex Algo's fail ........and a possible solution "

photo

 Jon Grah, Trading Signals Automation Expert AwarenessForex.com

 Saturday, April 16, 2016



You have a point in the video. But the problem is that in real-time trading, you are going to have to deal with BOTH situations ('ranges' and 'trends'). A lot of people try to deal with losses after the fact. The algo or supervision of the algo has to deal with losses in the moment. You must have the ability to adjust for both scenarios in real time. The solution of copying signals and then firing them for poor performance is more of an after-the-fact solution. This is different than knowing why you lose or win (because you actually know who you are serving in the markets....like a market maker) so that you can adjust settings based on specific what if criteria. Without that, you are left guessing when conditions change.

The book Trading And Exchanges, Market Microstructure for Practitioners chapter 8 'Why People Trade' is a good start to know about the different categories of market participants.


photo

 Zubair Badar, Senior Quantitative Financial Analyst

 Saturday, April 16, 2016



most of ranges strategies are flopped except they should be refined in timing which is very difficult even for professionals WHILE trend strategies are successful in back testing not in real time, do u know why? bcoz spread, slippage is soooo threatening and don't allow trend strategies to survive. what's solution? again it's all about seasonal timing in medium term range. everything or most of approaches are failed without timing application having trade cost, spread and slippage issues


photo

 Zubair Badar, Senior Quantitative Financial Analyst

 Saturday, April 16, 2016



timing is a key and determining right time requires many logics and research background and to be honest TRUE words are never spoken and most of successful researchers keep it secret


photo

 Brian Beamish, The Rational Investor

 Saturday, April 16, 2016



ya, ironically imo what's really important in creating trading systems is NOT shooting for 100% winners. The key to long term success is to try and ensure your avg. winner is greater then your avg. loser. Indeed, CME floor boys I worked with in the past were only looking for a 5% edge (being right 55% of the time and being wrong 45% of the time). The key here is WHEN YOU ARE RIGHT, YOU GET PAID!


photo

 Erkan Sağlam, Private Trader

 Sunday, April 17, 2016



I like self training machine learning algos for adaptation.


photo

 Marc Verleysen, founder at TSA-Europe -systematic trading and money management

 Sunday, April 17, 2016



You can combine models with a different market approach (range, fast trend following and slow trend following) into one single position using a timing sequence. So, while you have three underlying models, the actual trading will be done based on the model that is having the upperhand at a specific moment in time. Works well over here.


photo

 Zubair Badar, Senior Quantitative Financial Analyst

 Monday, April 18, 2016



Marc Verleysen can u shed some light in your description "time sequence" ??? use of time has many diversified approaches and interested to know what you referred above.


photo

 private private,

 Thursday, April 21, 2016



What Erkan mentioned makes sense here: a machine learning algorithm which learns over time should be tried on this type (and any type) of trading applications where the dynamic undercurrent should be measured and taken into account to adjust the trading algo

Please login or register to post comments.

TRADING FUTURES AND OPTIONS INVOLVES SUBSTANTIAL RISK OF LOSS AND IS NOT SUITABLE FOR ALL INVESTORS
Terms Of UsePrivacy StatementCopyright 2018 Algorithmic Traders Association