New EA EA YUME2

[lyazi]

which settings ?

 

[candleman27]

which settings ?

Thats Mr Tanaka's gold set, and it's I think at 1 lot per 10k, and on h4, spread set to 4, wow, now it's 7 months into the test, which seems much faster but still slow on h4, must be some kind of serious indicator! Will have to try on D1 chart too just to see but h4 def has more accurate entries than h1 and less frequent trades which wont piss the broker off as bad lol!

1560% in 7 months!!!!! Very impressive

 

[candleman27]

My live acct spread is currently 16 pts, revised the setfile to 33 start 28 stop for trailing to accomodate for higher spread, rerunning test and it looks very promising still about the same return as above maybe little more

Heres the new h4 setfile im working on will post when it's done, that's much better too

Jan 3 22 to May 9 22

1040% return in 4 months time. Best Tanaka robot yet I think

 

[candleman27]

Setting spread to current is the killer, I'm working on the trailing function settings to overcome the spread maybe my broker just has high spreads (typically seems to float around 4 to 5 pips during volume hours)

Sorry for my mental handicap I'll blame it on being sick for a solid week lol

 

[candleman27]

Let's keep something in mind for future research

The huge winners always happen around 6am GMT every time that's possibly worth noting for adjusting trading hours to try to find correlation between losers and winners with particular hours

At least they all seem to happen during big volume, might be as simple as excluding the Asian session well see

 

[Joerg Kappler]

My live acct spread is currently 16 pts, revised the setfile to 33 start 28 stop for trailing to accomodate for higher spread, rerunning test and it looks very promising still about the same return as above maybe little more

Heres the new h4 setfile im working on will post when it's done, that's much better too

Jan 3 22 to May 9 22

1040% return in 4 months time. Best Tanaka robot yet I think

View attachment 156830

Did you test with tickdata (99,9%)?

 

[candleman27]

Man I don't have that data currently, I'm running a longer test on h4 on lower settings now starting 7 1 21 to present and so far it has run all night and only made it to may of 22 lol but it looks good, you are welcome to backtest it however you want tho

I think it's on 0.4 or something like that to avoid the bigger losses in the losing stretches

I really think fwd test is more important than bt

BUT if you would optimize the setfile to profit from 8-24-22 to 10-24-22 that would be great not sure maybe it's the spread killing it in that date range it doesn't margin call just fails to profit during that period

Optimization for trading hours seems like an important one to me too

And I would really like to see auto lot working, it doesn't work with Tanakas gold setfile for some reason

 

[Joerg Kappler]

sharing my settings and result since 1 June. I had test since mid May while constantly adjusting and observing the EA characteristics. Up till this moment, I think this has potential. But I need the 'Max Position' function to work properly before I can further analyse its potential.

I hope that Mr Tanaka can look into what I have brought up on the 'Max Position' possible error to make this EA works.

I have been testing most of his EAs on the background and I think this is one of the best that suites my trading style, at least till now. More observations needs to be carried out with my forward test.

June
View attachment 156810

May (settings testing phase)
View attachment 156812

 

[Joerg Kappler]

Man I don't have that data currently, I'm running a longer test on h4 on lower settings now starting 7 1 21 to present and so far it has run all night and only made it to may of 22 lol but it looks good, you are welcome to backtest it however you want tho

I think it's on 0.4 or something like that to avoid the bigger losses in the losing stretches

I really think fwd test is more important than bt

BUT if you would optimize the setfile to profit from 8-24-22 to 10-24-22 that would be great not sure maybe it's the spread killing it in that date range it doesn't margin call just fails to profit during that period

Optimization for trading hours seems like an important one to me too

And I would really like to see auto lot working, it doesn't work with Tanakas gold setfile for some reason

Only tickdate BT tells a littel the "truth" of the quality of an expert.

 

[candleman27]

What is your spread setting I noticed that it will have a few months of loss before finally making it up overall but yes it's been challenging for a good setfile for this but that also doesn't look like Mr Tanakas gold setfile you're using

I'm not arguing with you you might be right but I have done quite a few back tests using tick story data and compared with a broker data It really wasn't much different at least with the experts I've tested and found the forward test to be much more useful

 

[candleman27]

I have a monster quad xeon server with maybe 24 cores and 32gb ram but I don't think my processor is even using 20% on the desktop PC which is probably faster than the server not sure why it's so slooooow lol

another thing to note is how spread sensitive this is, prob why Mr T wants to try on vps, a zero spread broker or super tight spread at least would really help this one would make all the diff.

 

[candleman27]

Can someone else look into this with me?? Every one of these massive profit spikes occurs around 6am gmt, very very very interesting with a time filter could avoid lots of losses, but usually only one big spike a month or 2 sometimes before it happens

This is current progress bt 7 1 21 and 3 10 22

 

[candleman27]

now it's having a tough time overcoming spread, that's 7/21 to 4/28.... has a great winning run then somehow doesnt work

Thats the longest backtest ive ever run, it still isnt done has 1 more month but we have a better picture now

 

[lyazi]

forward testing as well here

 

[candleman27]

I was trying to but it interfered with another ea i use full time on xau, it added a stop to an existing order that shouldnt have been there, no other way it could have appeared.. even with diff magic number so I dont know... if you can catch one of those big spikes you will be happy! Always seems to enter just before frankfurt last 30min of asian when the big spikes always seem to hit

 

[candleman27]

One huge factor is your broker's spread, ive seen zero spread on gold from somewhere, that makes all the diff on this ea if you could get that or at least a low fixed spread.. im a US trader so my options are very slim and its awful, I can go to the local horse track or casino and bet my whole bank acct, but I cant trade 100$ with 1000:1 lvg... stupid

 

[Joerg Kappler]

forward testing as well here

Can you give me your setfile. I try to do a correlation wit tickdata BT

 

[candleman27]

Can you give me your setfile. I try to do a correlation wit tickdata BT

It's actually Tanakas h1 gold setfile defaults but on h4, spread is everything for it

 

[tanaka akiko]

Just want to add, what I meant was when you set max position to more than 1 (for example 3). The EA will not open any new trade after the first trade is opened.

Yes, maybe it is closed with basket profit before 2nd order is placed.
When I change Basket profit=false, I see 2nd entry.

 

[lyazi]

hello tanaka make this guy an ea with settings thanks