Indicators development for T101's not very simple method

I have not been available cos I have to concentrate on pips accumulation, and forum activities can steal a lot of time from a trader unaware.

@diceman,
yes you are correct, the two past weeks were harvest of pips, again you are correct PATIENCE is the essence, 1 or 2 or no trades a whole week on 4HRS TF

there is the saying that if the eyes are patience enough they will see the nose, that is the case with BTS method

@jl83
Yes I still trade like that only now I looked for MTF TrendMagic that gives daily values on 4hr chart, this reduced the trades but increased the hit rates and makes the pip more abiding

I have tried to make the zzbrm offline basket chart work, but I couldn't so I still use the GPBJPY as the mirror of the basket week in week out.

kos

Hi Kos,

Thanks for your reply, happy to see you again here!

If I can be of any help for installing basket charts, please let me know, what kind of problems do you encounter during installation? For me it's working perfectly, I have set the time between updates at 1 second, so my offline charts behave like any other pair, sometimes even more quickly!

Regarding MTF TrendMagic, could you please share your indicator, so that I have a try before giving up?

Thanks!

JL

hi JL

you can use the SuperTrend nrp Mtf-indi and set wich TF you want instead of the "Current Time Frame" setting...if you want to switch to the default settings - simply reset it...somewhere should be also a version with alert...
regards

ps: never give up!...

Thanks Cadolino!

I would like to never give up... but it seems I have a problem with the most important thing with this method: patience. Though I don't have a lot of experience yet, I clearly prefer short timeframes, because watching the screen all day long, waiting for an opportunity, is something I don't like... and moreover I always fear to miss that opportunity, so generally I jump in too early (don't want to miss the good one), or too late (didn't see it was the good one).

But this week I try to apply strictly the rules Kos gave us. So unless I'm totally wrong, nothing happened so far, today. I would appreciate if someone could post his entries this week, so that I can check if I'm wrong or right with it.

Thanks

JL

You may or may not want to read this "book" of a post. It is detailing for me and others insights from testing hopefully.

Just posting what my indicator shows the market is doing, since entries were asked for, and I happen to be able to see and tell you when market started moving and breaking out (easy in hindsight for sure).

As you can see by attached pic, the market didn't start moving and (at least on GBPJPY) break out of range (see horizontal red line on GBPJPY chart showing what looked to be a double bottom, on GBPJPY, but when tested failed and price ran?) until about 21:00 GMT (Kos' broker Alpari UK is GMT+2, so there at 21:00 GMT he broker's server gave time if 23:00). I marked a red vertical line at that point also, which helped me draw a horizontal line on my BuysVsSells. See the range breakout on GBPJPY H1? See the breakout of range just slightly earlier on BuysVsSells?

Note: MT4, as I have said before, gives last week's open time during Sunday evening and night in the U.S., so when I started my BuyVsSells (or an EA that read the values from it), the readings were really reading the gains from this week added onto last week's gains. This is not correct. The indicator (started over) in the middle of the night Sunday-Monday U.S. zz was kind enough to share code with me, and when implemented with my own code, looks to have overcome this problem in my indicator. This code has not been distributed because for one the fix was just implemented, and for another it is in testing, and finally because zz hasn't given me permission to share his code, so I won't do that unless he gives me the permission. Without him and Robert, I wouldn't be where I am with this.

My personal EA's testing results so far:
Notes: At first, was using my code for "start of week" at 21:00 GMT - but my code didn't work by itself. The times shown for startbartime for each pair in comments were way off. But Alpari UK and IBFX didn't open until 22:00 GMT (another hour). So I figured I had time to implement zz's code because mine was not working by itself. So I did, and found that on the brokers that were open from 21:00 GMT used the correct first bar of week on those brokers. For those that hadn't opened yet, we would see if zz's code fixed those as well to using the correct startbar for beginning of week.

Well, it turns out that the brokers that opened at 22:00 GMT, using zz's code, used last week's last bar on each pair for the open price for each pair. This I considered incorrect, and already had code to fix that, so I implemented later in the night. So for Alpari UK and IBFX, because of this "problem" (before fixed), my personal EA on those brokers tested much better than on the other brokers which went through the whipsaw because of different open prices.

Just some testing results, to show you what is possible week in, week out, even though we don't know for sure WHEN the "big run up" will start, or if we'll have one or two "big run up"s during the week. Note on Alpari UK and especially IBFX what can be achieved if the EA does not trade during the ranging period before the big run up. That's really just gains that have happened in the last day - 1 day. Available once or twice a week, every week.

Testing gains for week (all baskets risking 7% per trade, so 0.5% per pair, for approx. 560 pip stoploss/40 pips per pair for MM purposes):

Alpari UK: 13.54% (used last week's last bar as this week's startbar before fixed) (so about 560 pips x 2 = approx. 1120 pips gained, since 13.54%/7%=1.9)

IBFX: 18.26% (used last week's last bar as this week's startbar before fixed) (so about 560 pips x 2.5 = approx. 1400 pips gained, since 18.26%/7%=2.6)

MBT MT4: -19.52% (used correct start bar - first bar of week, and took trades during ranging period like the rest did that used correct start bar. Gained 3.61% in last day, but not near enough to overcome the losses of previous day. I will work on this usng MM hopefully, so that when I'm losing, I don't risk as much - lower Lots, and when I'm winning, I max out my risk - more Lots - during the big run up. At least that's the plan)

MT4 ECN powered by ATC Brokers: No trades taken although same indicator and EA on this broker's demo as others.

CitiFX Pro: 2.91% (lost -10.74% first day, gained 13.65% yesterday. Gains actually overcome the losses).

Now this method I'm trying to do is just one method. There are others that attempt to do the same thing as this one, just using different method to capture the big run(s) and somehow try to not trade during the ranging period. It's obvious what I'm doing is not perfected yet (for sure), and only testing will give me the info to improve it hopefully. It all akes time. But the pips will be there, so don't worry about it.

Looking at these results, it makes me want to test without using ANY fixes (even ones in my already shared versions of my indicators). It just might be that MT4 was engineered for the gain of the makers, and they see all this and decided for their trading purposes to use last week's open bar as this week's start bar (this is default that it gives the wrong bar), AND that it's not using this week's start time until Sunday-Monday and not Sunday night. These two "problems" combined actually (at least in this one instance) kept the EAs on Alpari UK and IBFX from trading during the ranging period. Interesting. Repeatable? Don't know.

Regards,

Don

I'll have to cosign what disbellj just wrote. i have tried all the variations in here and even though Kos entries are very good. theres nothing like good 'ol price action. and it's up another 1000 pips in this picture. you need only one indicator and it will be the one in this picture.

and a note about mt4. the platform itself is flawed because of the whole starting week bar problme and it wasn't really made for us little guys to trade like big guys. but tha isnt it's only flaw though. but so many programmers know the code and it's so popular that there are so many bridges for it. even banks are seeing the benefit of offereing it as a platform for retail and institutional clients. i for one hope to get to that level

Hi hazelj80,

Could you please explain which are your entry points based on "price action" and this indicator? When do you enter exactly, and how do you avoid ranging periods?

Thanks

JL

Hazelj80,

What is the indicator name? I am trying to find it in the attachments but can not find T101 Master Indic.
Thanks

you cant really avoid it because you cant really poedict it. you will win way more than you loose

you enter when one side goes above 0. if it switches then switch direction and take a loss. most likely less than -560. then it will pull back and profit lock will kick you out. wait for the high to get broken, enter again. place you stop loss on this indicator at the recent low. If the high never gets broken, dont enter. dont try to counter trend trade unless you know what youre doing but theres no reason to do so. julius never did. no fancy secrets here. this indicator just helps you see what julius sees.

simple as that

it's called buyvssells there are different versions of it but the newest one will be availible sooner or later

Thanks for such a quick reply.

May I ask you the same question I already posted a few days ago... I have installed the offline charts, which work pretty well, so the basket now looks very similar to any other pair.

Compared to GBP/JPY, I must say it behaves really similar, it's very strange to see the nearly exact same movements...

So my question: which difference do you see between trading the basket, and simply trading GBP/JPY. There would be far less risk to do so (less lots open, less spread, no pairs going in opposite direction, etc.).

I must say I look at both charts (offline basket charts and GBP/JPY) since a week or so, and there are so similar that I can't find a good reason to trade the basket... But I'm sure you will have one, at least

Thanks

JL

it does behaves simular. but it's not like that all the time. people have run into this problem before. i dont trust one pair to run the whole basket. the entries can be a little off as well i found out. just enough that you're either too early/too late.

and it's always better to trade 14. you're going to make more than you would ever loose if you're trading the correct way. as julius once said "would you rather bet on kobe bryant to win the championship or the whole lakers basketball team?"

I see what you mean.

However, when I say trading GBP/JPY, of course it's not "looking at GBP/JPY and trade the basket", it's rather "looking at GBP/JPY... and trade GBP/JPY" ! Because you're right, they don't always behave exactly the same...

I was considering this because of Kos using his indicators on GBP/JPY to trade the basket, so I wondered why I could not simply buy and sell GBP/JPY, with the same indicators and method...

Anyway, thanks for your reply, and look forward to the "newest" version you were talking about!

(BTW, I decided to concentrate on entries based on Kos strategy, and I must say I have success since this morning, over 1200 pips now!!!)

JL

brucech,

The Master indi is simply me taking the sorting of gains, re-association with pair names, and the drawing of objects that are all in the backtest indicator I uploaded, combined with the BuysVsSells indicator I uploaded.

Now although I coded using:
Broker_WeekOpenTime = iTime(NULL,PERIOD_W1,0);

... which should give me the open time of the week, and from that I can use this code to double-check on lower timeframes than D1 if the bar that MT4 is giving me that is associated with that week open time is the correct bar or not:
StartBar = iBarShift(NULL,0,Broker_WeekOpenTime);
StartBarTime = iTime(NULL,0,StartBar);

... I found through testing that I was given the last bar of last week by default.

So I made this code to subtract one bar if the StartBarTime is older than the Broker_WeekOpenTime:
if(StartBarTime < Broker_WeekOpenTime) {
StartBar = iBarShift(NULL,0,Broker_WeekOpenTime)-1;
StartBarTime = iTime(NULL,0,StartBar);
} else {
StartBar = iBarShift(NULL,0,Broker_WeekOpenTime);
StartBarTime = iTime(NULL,0,StartBar);
}

But in the indicators uploaded to this site by me, they don't actually "start over" every week like I want them to. So first thing is I had to take the code above and move it from int init() to int start(), so that it's not calculated once when indicator is placed, but calculated every tick. So at the start of the new week, the indicator should see the updated information and "start over" Sunday night. Problem is, while this Sunday may have been 2010.08.22, so MT4 should have given Sunday night at week open when using iTime 2010.08.22 00:00, it was giving 2010.08.15 00:00, which was old information and not updated. I've noticed that on Monday that the indicator has "started over", but on Sunday night at actual week open, it doesn't.

So using zz's code:
Broker_WeekOpenTime = calcWeekStart( TimeCurrent(), BrokerTZ ); //zznbrm

... I found that my indicator was "starting over" on Sunday night like should have been, but I found that on broker's starting at 22:00 GMT, that last week's last bar's open prices for all pairs were used, instead of this week's first bar open prices.

So I had to add back in the above "checking and correcting" code, and found as long as I used zz's code, and also my code, that a total fix was found.

So long story short: Anytime besides Sunday night, you should be able to download T101_BuysVsSells or my T101_backtest indicator (first one uploaded, since I haven't really tested the 2nd one uploaded), and when both are put on screen, you'll basically see what hazelj80 and myself are looking at and making decisions from. The objects drawn on chart and the pair slots or colors we are not presently using. That's just there for info and to see when testing if we notice something that is repeating and we may be able to make a strategy from in the future. For now, we use BuysVsSells and price action from that indicator to come up with rules to trade. I just know price action rules, and no lagging indicator can beat it. If we could tell trend from ranging (easy in hindsight, but would probably have to show just buyall line and not others, and maybe see if I could make an T101_ADX indicator, then just maybe we could tell), then we would simply know when not to trade and when to get in and enjoy the "big run up" (and maybe even the ride back down too if it run up high).

Testing EA's at this point. Gains are for the week so far:
Alpari UK = 15.92%
IBFX = 20.24%
MBT MT4 = -18.15%
MT4 ECN powered by ATC Brokers = no trades
CitiFX Pro = 5.31%

Knowing I'm risking 7%, correlated with approx. 560 pips stoploss, then 15.92%/7%=2.27, and so 560x2.27=1271 pips, and 20.24%/7%=2.89, and so 560x2.89=1618 pips. The other brokers aren't worth the trouble of doing the math at this point.

If I was able to either not trade, or at least lower my risk (choose large stoploss so my Lots are lower than normal, or maybe half Lots if last trade closed with StopLoss) when ranging, then I wouldn't lose as much when ranging. If I was able to increase my risk (lower stoploss, or larger risk percent) when trending (maybe adding Lots if last trade was closed by profit-locking), then I would increase my percentage gain during trends.

Regards,

Don

Well, knowing the way the basket moves like GBPJPY, we could look for the big trend in GBPJPY each week instead.

It's really just personal preference. if you use an MM equation, it wouldn't be much different.

Put it like this, with my MM, I could do this for one pair:
Lots = (Account Balance x (Risk Percent x 0.01)) / (StopLoss in pips x Pip Value)
Example:
Lots = (1000 x (10 x 0.01)) / (40 x 10)
Lots = 100 / 400
Lots = 0.25 (2 mini lots and 5 micro lots on a standard broker where pip value is approx. $10)

OR I could do this for the basket, to find out how many Lots to trade each pair:
Lots = (Account Balance x ((Basket Risk Percent/14) x 0.01)) / ((Basket StopLoss/14) x Pip Value)
Example:
Lots = (1000 x ((10/14) x 0.01)) / ((560/14) x 10)
Lots = 7.1428 / 400
Lots = 0.01 (1 micro lot per pair on a standard broker where pip value is approx. $10)

Either way, you risk 10%. On the basket you'll have to get a lot more pips just to equal the gain in dollars you can make on one pair.

BUT we're talking about automation here too. And if I can automate a single-pair strategy that is consistently profitable (well, I haven't found one yet, or I'd automate it), then as the account balance grows, each pip becomes worth more, no matter if trading a single pair or a basket. I see the basket as diversification, since the risk is over multiple pairs vs. one pair. We're not betting on a "single player" but on the "team". But "betting" really isn't the right word anyways. If you can trade, it doesn't matter if it's one pair or 14 pairs, although I'd much rather automate either one because I don't like sitting and looking at charts for long periods. If I didn't mind doing that, there are MANY pips on EVERY PAIR that can be banked EACH DAY. Go with the trend, and it's that much more. Even against the trend are many pips to be had.

So to conclude, it don't matter. I wish people would quit bringing up this question. If you want to just trade one pair, then this thread talking about basket trading will not interest you. Using an MM equation like mine will make sure the EFFECTIVE pip value keeps increasing because Lots is increasing as your account balance grows. If you actually want to trade a basket though, knowing that one day each pip will be worth a lot, and if it can be automated, then this thread might interest you. Either way you'll get to a million with compounding. But 1st rule no matter if trading one pair or a basket is what? Risk management, reduction, and elimination if possible. If you can get rid of the risk as soon as possible on any trade (moving stoploss to break-even as soon as possible, or profit-locking like we're doing here), then what's the risk to reward? Risk is gone. Reward is ? (that's right, there is no limit other than price action - take what the market gives)

On yeah, someone will say: "10%!? Why are you risking so much?!", to which I'd reply: "Risk is personal. If you want to risk 2%, risk 2%. If I have a profitable enough strategy, especially if I'm still testing it, why not risk 10% if it wins like 90% of the time? If it wins like 50% of the time, yeah 10% would be way too much risk!".

Regards,

Don

I don't know if someone can answer this question (though I know it's personal to each one).

Based on your experience with BTS trading, what would you consider a good level of gain:
- before introducing a breakeven (to avoid losses in case of return)
- before introducing a trailing stop (to protect profits, while allowing the trend to develop)

I know that Kosomolate has talked about such levels, and I don't manage to find good tresholds.

Thanks!

JL

if you have good entries. you should just be using the porfit lock EA's. they should be more than good enough. i thought i saw them floating around unexpried but don't know where. either that or someone needs to build one for the numbers that Julius provided.

The issue with the ProfitLock EA's is that once profit is locked, even a small pullback triggers the stoploss.

Kos said he was waiting for a fair amount of profit before triggering the ProfitLock EA, and I have experienced the same, so I wondered which would be the best level to do so.

Creating a ProfitLock EA is very simple, but based on Julius' rules it proves to trigger the S/L too easily... It's also possible to change the rules to apply a trailing stop (to avoid "steps" in the profitlock), but the question is which level to start trailing stop.

The same is true for the breakeven, which has a different use, a minimal profit (just to cover slippage when closing all) is locked once a gain level is reached... but again... which level?

well if it pulls back. just wait re enter. you should have re netry rules. also what numbers are you using? because one set is right and the others are wrong. these are right ones:

http://www.forexfactory.com/showpost...postcount=2121

and thats how it should be done. youd need to have a muliplier if you want to trade bigger lot sizes than .10 mini lots on the stardard account numbers and anything bigger than .01 mucro lots will need a multiplier for the mini lot locking numbers.

don mentioned he was up 1400 pips this morning using profit lock numbers. so they cant be off. indicators that lag will cause you to enter late and then profit locking with these very specific numbers will be out of tune with the market. they work well. trust me.

also, the market moves in dollars, not pips so thats why he never did a trail stop in pips. makes more sense because not every pair is worth the same per dollar. eur/gbp being one of them in the basket

I agree with you about dollars and pips, but of course when I speak about trailing stop, I mean in dollar. The principle is the same, but rather than having the ProfitLock threshold jump from 700 to 1000, it's smoother because the ProfitLock moves at every tick (in the same direction of course). So if you decide to "give back" only 100$ you can do so. This is useful, for example, during news... just before news, you can tighten the trailing stop: if news confirms the trend, it's OK, otherwise you are stopped out... but with better conditions.

About re-entry, I'm OK with the principle of re-entering... but the rules to re-enter are not always valid once you have been stopped out... so the question becomes "re-enter or not re-enter". If you wait long enough before triggering the ProfitLock EA, you don't have such a problem, and you can make more profit in one single trade (which BTW avoids adding spread).

well, if youre up to watch the trades, then yes you can trigger the PL whenever you want. that's what julius does. I'd probably read on what he says about that as for clues. i think he goes by feel most liekly and we know no 2 traders are the same. Also, using breakouts would be the best thing. i think even Kos is using the last low/high for a stop loss. i wouldn't try to give back the whole -560. there wouldnt be a need to. even if you do use indicators.


i think what julius tried to do in markets like these this week, is to have looser numbers so you give it some room to breathe. as youve seen. too close causes it to get kicked out too early and "get in at the top" which will cause uneccsarry losses. too far away and then you give back too much.

what this does is let it run up, pullback get kicked out. break out run up again pullback again, get kicked out again. pretty simple, it's all elliot wave and breakouts in basket form really