Yen Trader

Thank you very much, nonlinear, for running these tests.

RSI RVI CCI all true. I personally do not care for indicators, but just testing here. Happy testing..

General observations thus far on 4H. Results seem better when using the signal of the current TF (others may try to scalp with this but I think that may be very difficult). If one does not use a TS you get a much higher PF (lower win rate), but a greater draw down (the indicators may help here somewhat, have to test more, but I am concerned about curve fitting with the indicators - I like systems with only a few variables). With the TS, the win rate is very high, with a lower DD, at the expense of profits (but easier to trade from an emotional standpoint). Need to test more on the daily to see how that looks, but I suspect it will look good. On the 4H, I could see trading this on lower lots across maybe two or three uncorrelated pairs. Will post more test results if everyone wants, keeping in mind that these back tests do not really mean too much. I only back test to confirm the concept on different samples.

Implemented Abokwaiks trigger section into "my" expert base as it contains already many additions and different exit variants so it was easier to add and test his trigger section.

This is basically the Abokwaik YEN-Trader Verison 3.

addition to his trigger i added the daily trend EMA 288 on 5min chart as entry filter (see post#36 as reference), the price must be in a defined area near the EMA to trigger.

Exit is a bit different:
If there is only one position open it is trailed until stop hit or a counter signal appears.
If 2 or more positions are open the expert exits on "the other side of the EMA.

not fully finished yet but did some tests with averaging, semi martingale and finally got this mix, please note, it will also close with a loss when conditions are met.

Very Interesting System!

My brain is at full speed, I thing I will also code this system...

if you backtest from 2012 - 2016, you will find it's always losing until late 2015.
I am pretty sure i have the complete data sets.

It's fascinating to watch this EA in visual mode. I've attached some screen-shots that show extremely accurate
entries, but I wonder if there isn't something else going on, here. The fact that I'm not getting very many stop-outs
makes me wonder if my backtests are flawed. Modelling is 90%, so, perhaps, that's why my tests don't mimic
what I've seen from nonlinear's results.

Been playing around with not using the TS - using the indicators does reduce the DD. Really not too bad for 3% risk. Will load some older data to test on these settings later.

I'm using v.03, which doesn't have indicator options. I ran a bunch of backtests with, and without, a TS, and
the tests run with a TS were far superior. YMMV.

What are your thoughts on max orders? Better to use more orders with less lots or less orders with more lots? I have been using max of one primarily because of US brokers.

I have tested both max orders 10 and max orders 1. For max orders 1, I just multiply the lots by 10 so each trade is .1 instead of .01. Either one returns about the exact performance and draw down

I ran a bunch of test from the period jan 1 2015 thru current (aprox 1.5 years). I have run on 1 hour and 5 minute charts. Have not done 4 hour charts yet. Its late so Im not going to post all the reports but in summary: None of the indicator options work very well. They all hurt performance except using the MA option which does not improve or hurt performance more than a few $ over the test.

Trailing stop: Hurts performance by quite a bit in almost all the tests I ran. It does reduce the draw downs a bit but IMO its not worth it. The highest was DD was 16% not using the trailing stop. Using the trailing, the DD was usually around 6-10% but returns were sometimes 50% lower so not worth it. I can post up some of the tests tomorrow but one of the best was on 1 hour chart using no trailing stop. 38% gain with a max DD of 12%. Modding quality on my tester is around 90% so not sure how accurate all the test really are.

The base settings for tp and sl are pretty good and yielded good results. I tried higher and lower tp/sl settings and none were better than the original EA settings. One other thing I noticed was 2016 has been good. Early 2015 is when most of the DD took place but from late 2015 into this year this ea really made hay. So it may be worthwhile to run tests on past years or longer time frames to see just how good it is.

I cut max orders to 5, due to margin percentage hitting pretty low numbers during trades. My demos did great last night, which yielded around 10% return.
However, that was with max orders at 10, which pushed margin way too hard. So far, I'm liking what I'm seeing.

Agree on this. I am not going to bail on the TS yet. With fixed TP targets it might only do well when the pair is trending. Also, using the indicators reduces profits because it is taking less trades, but those trade look like better entries. That is why it may make sense to trade with the indicators on multiple uncorrelated pairs, which will make for more trades and hopefully balance out returns as well.

Congratulation abokwaik , very good EA.

it is a trader preference, but RSI and/or RVI true, less trade, better entries.

Ran it on 2014-2015 with basic 1:3 RR on 4H no TS -- it still made decent gains. Same settings on AUDCAD 4H not so great PF .93 with only 19 trades

hi,can you share your set,please?

Not bad for one day.

how are you trading that Rob? 5M with TS? What is your trigger TF?

H1 EA input, on a M5 chart, with TS. Pretty much OOTB, except LookBackBars=2, and MaxOrders=5.
I'm a U.S. trader, so the 50:1 leverage is something that has to be carefully considered, with regard
to margin. I also had to cut my lot size from .1 to .05. This is probably still too aggressive for 50:1,
$2000 balance.