Announcement

Collapse
No announcement yet.

Yen Fortress - 100% Automated Portfolio Trend System - NON-GRID, NON-MARTINGALE

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • siamfx
    replied
    Hi Falcon,

    Your myfxbook - http://www.myfxbook.com/members/falcon333 has a vouch from BluePanther - http://www.myfxbook.com/members/BluePanther , but you have never posted a single comment on myfxbook which is weird because people normally only vouch if you had something worth posting.

    Also, your systems don't have a discuss button which I've never seen - http://www.myfxbook.com/members/falc...rtress/1047922 , unless your are an autotrade system. Can you clarify any insight? Is there a way to turn off discussions on myfxbook?
    Last edited by siamfx; 11-09-2014, 08:37 AM.

    Leave a comment:


  • siamfx
    replied
    Originally posted by falcon View Post
    provide testing reports from 1993 to present (the entire length of the Metaquote history). .
    The fact that you are using metaquote data is quite worry some. There are significant data holes in this data. The reason why vendors show backtests from 2006 til now, is from the available tick data from quality sources.

    I wouldn't take any backtest seriously until it is uploaded to myfxbook with tick data from an external data source.

    Honestly, I wouldn't bother proving yourself with backtests, Just keep your Live accounts running and base all your knowledge off these forward test. Best luck with your Yen Fortress.

    Leave a comment:


  • peter5433
    replied
    I have just signed up for this one using a very small account, adopting the same approach as falcon, in that if the returns continue to be this impressive, compounding will take care of the rest..

    I was always amazed that no one previously ( to my knowledge) has come up with a decent system for trading the GY which, being one of the most volatile pairs in history, should have presented enough opportunities for those more gifted than myself in coding strategies..

    What impresses me most with this system, compared to other mutli entry systems, is how it converts equity to balance - ie locking in profit, as well as controlling drawdown - being very well thought out, avoiding the usual marti grid pitfalls..

    Good luck all!

    Leave a comment:


  • falcon
    replied
    Originally posted by Kisigal View Post
    All graphs and test results are undeniably impressive, but why testing was performed exactly since 2002? With MetaQuotes server, you can upload a longer history. 'll be much more confident in the strategy and very grateful to You if you do the tests on a longer period. Thank you!
    Kisigal, what you are asking exceeds the burden of proof of 99% of strategy developers out there. I have never to date seen any EA vendor/signal provider provide testing reports from 1993 to present (the entire length of the Metaquote history). Most only show from 2007 to present. Rarely do they show from 2002 onwards and thus I thought I was moving beyond the norm in providing a 12 year extensive report.

    As I have said in an earlier post, "history from 1993 to 2002 turns out be problematic to most strategies submitted for backtesting, revealing more instability and deeper draw downs. I could write this off as that history being so long as ago it now irrelevant. But is it? It is a challenge for sure and I'm still grappling with it. "

    Nevertheless, always willing to please, here are two strategy reports freshly generated that show the two strategy returns from 1993 to present (21 year history):

    Two Notes first: 1) This forum does not allow me to upload html files greater than 85 KB, so I have to take screen shots of the 4000 KB reports. Also, the Modeling quality is open prices only because it is the fastest method, and all my strategies open/exit on open prices, so don't need every tick. Trust me, "open prices only" produces the exact same result as "every tick" for my strategies.

    Yen Samurai-251:

    YenSamurai-251_Sept-1993toNov2014.jpg

    As you can see, from the period 1993 to 2002, there was not much draw down, but considerable stagnation (not making new highs). While you might think that is unsatisfactory,
    it isn't so bad really, considering how easy it is for a strategy to blow up during that earlier period.

    Yen Samurai-295

    YenSamurai-295_Sept-1993toNov2014.jpg

    Here there was some growth from 1993 to 2002. The strategy made $461 (or 4,600 pips) during that period.

    In sum, if you think that history can repeat itself and become once again the period from 1993 to 2002, I would advise you to look elsewhere for greater returns. My system is modest only on that earlier period but it is not a great earner.
    Last edited by falcon; 11-06-2014, 09:51 PM.

    Leave a comment:


  • falcon
    replied
    re: pta. I could sense an affinity with you, thus why I explained myself. I think similar to you. We've probably been doing the same mental tracks, encountered the same mistakes, learned from them and started over. I think you will do well in time. Proper method, lots of good ideas, lots of experimentation and some free time.

    I also have 2 daughters that take up far too much of my time (5-year old and 8-year old). I couldn't do much of anything when they were born, so I totally get that. I'm a doting dad. Now that they are in school I can work in peace and create from 9:00 am to 2:30 pm when they get loose and I hang with them. But do I think strategy development when I ride bikes with them, push them on the swings, swim with them in the ocean. Yes, I do. I cannot turn it off
    Last edited by falcon; 11-06-2014, 09:52 PM.

    Leave a comment:


  • pta
    replied
    falcon: thank you for an exhausting answer as well as your introduction. I was just afraid that you're "just another guy with StrategyQuant" trying to sell a set of luckily found and most likely curve fitted strategies Which is not your case obviously. Im glad that there are ppl like you trying to push the envelope and spend the time by doing the research the right way. I will keep an eye on your portfolio for sure and I will also try to find the time to read some of the works you mentioned.

    Be a programmer myself Im still looking for developing a portfolio of my own proprietary EAs but since our daughter got born 2 years ago the priorities are completely different

    Thanks once again for your answer and keep up the good work.

    Leave a comment:


  • Kisigal
    replied
    All graphs and test results are undeniably impressive, but why testing was performed exactly since 2002? With MetaQuotes server, you can upload a longer history. 'll be much more confident in the strategy and very grateful to You if you do the tests on a longer period. Thank you!

    Leave a comment:


  • falcon
    replied
    Snapshot of One Strategy in Action: Perfect Trend/Reversal Catches

    I wanted to present to you all a snapshot of one of my strategies in action, to show more than words can do how accurate it is in terms of keeping on the trend for its full length and then exiting it and reversing direction at the right moment.

    YenSamurai-251_Nov6.jpg


    There is plenty of incredibly accurate trades here. Notice it caught each trend when it happened, both long and short. It just recently exited its last long last night for 810 pips profit and is now in short with 134 pips profit. I'd like to see a manual trader do so well. I certainly can't.

    Here is the same chart with the history of trades since Oct 09:

    YenSamurai-251_Nov6_AccountHistory.jpg

    Notice how it is catching all the moves for the full length of them. Not bad. I had to put it on its own live account for curiosity sake.

    The strategy has been a workhorse for Yen Fortress since the beginning, having produced 2441 pips since May 15. In backtesting it has produced over 37,000 pips from 2002-2014.
    Last edited by falcon; 11-06-2014, 07:53 PM.

    Leave a comment:


  • falcon
    replied
    re: pta

    I am using Strategy Quant's EA Analyzer Pro 3 for analyzing strategy results and to blend strategies into portfolios. It is the best product on the market for that sort of blending, along with correlation analysis.

    Did I use Strategy Quant Pro (Formerly Genetic Builder) and its genetic evolution algo to generate these strategies? Only partly. I use it for brainstorming, to help me discover good ideas and their combinations, but the fine coding of each strategy is my own. I am an advanced coder, having coded and run managed strategies on tradestation easylanguage since 1997, mql4 since 2009.

    I am the author of the Forex Razor School, including its popular section on Automated Trading and MQL4 Guide to Building an Expert Advisor. I have taught hundreds of people over several years how to properly code strategies and backtest them. All for free. I have even helped dozens of people create strategies out of their own ideas on the Forex Razor forum as Steven Randolf (my real name is Blake Rodger), also for free: https://www.forexrazor.com/forum/tab...S/Default.aspx

    My street cred out of the way, I have to admit that Strategy Quant Pro is first in its class for brainstorming and backtesting combinations of ideas. I highly recommend it to anyone for the price. However, I am fully aware of its pitfalls, the same pitfalls for anyone developing and backtesting a strategy on any period of history. There is so many snares and illusions in the process that most people get it wrong. There are steps that need to be taken, steps which I have covered in my articles in the past but I have not yet written a book about. It is a scientific process really, a method in its own right. And then there are so many wrong methods that can get you so far in the wrong direction. Believe me, I've traveled in those directions.

    As Francis Bacon once said (and I'm paraphrasing): "Method is the path to truth and genius the speed of travel. A proper method takes you slowly in the right direction, but genius along the wrong path/method can move one further and faster away from truth."

    Let's take an illustration of this. You mentioned the problem you saw once you expanded the price history and discovered that your nice stable profitable strategies were undone by chaos and unexpected periods of draw down. This is a great point. It throws up the issue of how much curve fitting is done when developing, backtesting and optimizing a strategy. Most people get this terribly wrong. They backtest on limited years and bars and test it another set of limited years and bars and things fall apart. I try to minimize this as much as possible by making sure that I have large in sample/out sample sizes of plenty of years + bars + trades. I also make sure that the strategy can work on similarly correlated symbols for the same length of time. For instance, if a GBPJPY strategy, it should work on EUR/JPY as well -- at leas to a reasonable degree ; if EURUSD strategy, it should work on GBP/USD, also to a reasonable degree.

    However, it is true that history from 1971 to 2002 to more problematic to most strategies submitted for backtesting, revealing more instability and deeper draw downs. I could write this off as that history being so long as ago it now irrelevant. But is it? It is a challenge for sure and I'm still grappling with it.

    But I think things will be alright. I have added to each of my strategies my own proprietary set of indicators and filters that I have used for the last 10 years to make each robust and stable. I won't be giving them away for free. I love to help the community but I have to make a profit as well.

    Love your question - shows a insightful and critical mind

    Leave a comment:


  • pta
    replied
    Hi falcon,

    seems like strategies generated using genetic evolution algo (StrategyQuant?). Do you use any portfolio optimization methods to rebalance the portfolio or is it a set and forget balancing act?
    Also beware the seemingly robust results. However right the StrategyQuant does the backtesting job (in-sample / out-of-sample) it's still just a generated strategy using a limited set of price history. My experience with StrategyQuant is once I expanded the price history after I discovered a nice stable profitable strategies, the chaos stepped in and prolonged periods of DD suddenly appeared on the expanded history interval.

    But I wish you luck

    Leave a comment:


  • falcon
    replied
    re: Kisigal. Yes, I admit, I was too giddy with excitement on my system at first and broke my own risk rules in trading this system on a $400 account. If I had traded it with a $400 account back in 2004, and I had that one bad month of -400, I would have been wiped out. But I took a chance. Why?

    Because I had noticed that the GBP/JPY had been locked in a tight range of 500 pips from beginning of Jan-2014 to May-2014, when I started trading the system. I thought to myself, this puppy has been already five months in a range, and it is due for a breakout, and my system will exploit that breakout without incurring too much risk. I was ultimately proved right about the breakout but wrong about the timing. It continued to be locked in a range for the 4 more months (May, June, July, August) and it wasn't till September that it busted out. But the system survived that range for those long four months and my reasonable prediction and patience payed off.

    Yes, because I had traded with $400 and not $1000, the present stats are distorted in my favor. I would be sitting on 200% return instead of close to 500% return at the moment.

    However, I have been honest enough to steer folks to the 12 year performance reports that quite clearly suggest that it is wisest to trade 0.01 per $1000, and at that amount to expect close to 200% yearly return with 40% draw down. Or, alternatively, you can trade with 0.01 per $2000, and expect close to 100% return with less than 20% draw down.

    I'm currently deciding to go to 0.01 per 1K -- I'm just waiting for a settling period. My open positions are too high in profit.

    Leave a comment:


  • Kisigal
    replied
    The test results are certainly tempting but at the monitoring You started trading with only$ 400 which exceeds all acceptable risks and the result is distorted. You will continue to trade 0.01 lot per 1000$ or 2000$? I understand that when you trade with 0.1 lot per$ 1000 of historical drawdown was about 40%? Thank you!

    Leave a comment:


  • falcon
    replied
    Strategy Reports - Part II

    Yen Samurai-738
    Pips: 23,600

    YenSamurai-738.jpg

    Yen Samurai-133
    Pips: 23,300

    YenSamurai-133.jpg

    Yen Samurai-114
    Pips: 20,000

    YenSamurai-114.jpg

    Yen Samurai-101B
    Pips: 14,000

    YenSamurai-101B.jpg

    Yen Samurai-638
    Pips: 12,800

    YenSamurai-638.jpg

    Leave a comment:


  • falcon
    replied
    re: Fxtsunami , thanks for your input on your criteria, that you like to invest in signal managers with large account sizes. I can appreciate that.


    Personally, that was never my own criteria; I look at performance and risk stats, irregardless of size, particularly if it is being auto-traded by robots (emotions will not interfere with trading on a larger account).


    Did I kick myself that I started out this account with a small size? You bet. I wish I put in more.


    I am now approaching near 500% equity return on this account, and if I had put in $10,000 at the start, I would now have $60,000.

    But at 30% monthly return I should get me there soon enough.





    Leave a comment:


  • Fxtsunami
    replied
    Why so small account?After 12 years we should see at least 100k account.Easyer to invest in signal provider with big account.Thank you

    Leave a comment:

Working...
X