T101 Basket Trader with ONE of Nanningbob's Balls

John,
I let it run all day, no problem. All I changed were some BE & trailing stop numbers. I'l let it run and see what happens. I'm running a 5m & 15m timeframe.

What do you think about changing the "define 15m" to "int Period = 15";

Hi TD

not quite sure what the difference is.

MQL automatically assigns types to #define statements depending on the value assigned, so #define Period 15 and int Period 15 are exactly the same, the only difference is that the int Period can be changed within the program run (which we don't want to happen anyway)

regards, john

Sadly, VQzz repaints.

Also, once the VQzz changes colour, the EA only trades after 1 hour or more. Is that suppose to happen?

And once it open the trade, after 30 minutes, it changes colour. Wow.

Happily it repaints. That's how it was designed to function. It repaints the current upper timeframe bar because that's what it was designed to do. It was created to visually represent the direction of the upper timeframe per the VQ calculation used by Kosomolate's strategy (close of bar only).


What do the rules of the strategy (EA) say about entries?

If you specify how you want the EA to trade, I'm sure jmw1970 will point you in the right direction.


The VQzz indicator changes color when the underlying VQ values change. It is as simple as that. If it is not working per your specifications, then you should probably alter the code accordingly or simple go back to using the original VQ indicator.

Using the VQzz indicator, or an internal function, or the original VQ indicator in an EA should provide the exact same results.

Thanks for the clarifications zznbrm.

I know VQzz works this way as you mentioned before in previous posts. Its the same as VQ7.

For my case, the VQzz went green, then EA made a buy trade, but VQzz went back to full red again. So I have to manually close the trade.

Hi All

this is just my opinion so it can be discarded as necessary as I am not nearly as proficient in T101 Basket Trading methodology as others..

trying to find the perfect indicator fit based on the offline chart is almost the same fruitless endeavour as trying to find the perfect indicator (or combination of) for a normal chart.

simply does not exist..

however we do have 1 advantage in our favor in trading baskets as opposed to single pairs, and that is - we are trading baskets, ie. multiple pairs, so why not take advantage of that fact

why do we try to get for example a moving average cross on the offline chart - it is prone to whipsaw just as a moving average cross is on a normal chart

however if we have a total of 14 pairs, and we get a cross on x percentage of them, then it is less likely to be a whipsaw reaction, we can also take into account the xxx index values, as many people have noted - we are mainly trading against the JPY (although the CHF is starting to bring some influence this year at least).

I personally think trying to find one/multiple indicators that work correctly on a offline chart that can be automated is the same problem as trying to find such a combination for a normal chart.

Therefore, let's think outside the box, ok - we have the offline chart, which is nice to get a visual indication of the baskets, but do we really need to do our entry/exit analysis on it, would it not be better to do this type of analysis on the underlying data

Just my thoughts, discard as necessary..

Regards, John

John
I think we see a slower turn then a faster acceleration of pips, at least thats how I see it with the GVC 8 pair.

I agree on the individual FX charts. The attached screen shot are all 14 pairs. I use a screen that has Renko 15pip charts. The slate ones are the GVC 8, the olive are very close followers of the GVC-8. Then there is the other 4 which seem, at times,totally different then the other 9 or 10. I believe the CHFJPY could make it an even 9 pair. They turn slow then zooom. I keep the renko chart open all the time when using Basket 8. It's like Kos said I watch for the price to bounce off of the HA, so I hold my position, at this time I don't trust myself to double down, or if it goes through the HA I EXIT my position. If I could use the Basket-8 for RENKO I would.

I know this charting looks like a big winner BUT It takes back almost as much as it gives on the reversals also missing & bad data, hence still looking.

Hi TD

that's a little further out of the box, but i still think we need to go further.

A lot of methods that can work manually, just don't work automatically as they require a "visual" element, ie. go back to the HAS, how close does it have to get?, how do we know the turn has happened etc.

The original method had no charts, unfortunately there was also no complete instructions as I believe Trader101 was keeping that secret so that he could charge for it on his mentoring course.

This is why I am thinking the triggers just need to be raw data - calculated purely using math, and forget we even have charts.

For instance, in the original method one rule that was mentioned was that the IA has to have settled down, ie, all the buys & sells together, because if the buys & sells are mixed up then we are likely in a ranging scenario, now with pure data we can see this, on a chart people have been looking for the perfect indicator to show ranging market for ever.

Another for instance, when trading the basket we know what we are mainly trading against, people have been suggesting previously to use CCFp, however the xxx indexes show the information just as well - again with the indexes we have pure data - no visual element required

Just some thoughts, I am hoping that these posts will spur people on to think of solutions that don't rely on indicators on the offline chart

Regards, John

Yea, so we're back to square one.

John,
Have been thinking along similar lines for a while now. Seems like finding the ideal dependable indicator for fully automatic trading is an almost impossible task. They all just reflect price.
The main problem being as you say, the whipsaws and retraces.
In keeping the the thread title, the EA is already capable of tracking Market state and can get us into a trade in the right direction.
If it could also monitor price rate of change it would have the real thing.
Would it be possibe to have Two Take Profit settings?
In a slow market it would exit with the SmallTP, reentering (if entry criteria were still met) to take advantage of the whips and fakes.
In a trending market it would hold out for the RegularTP.

To me the key seems to be BE, SL, TP and Ts. Not sure if that is an area in which a Trade Manager would be more effective.
Thanks for all your work on this project.
EP

Hi EP

the question would then be, how would the EA know which TP to use

i still think the answer lies somewhere in xxx indexes (usd, chf, jpy etc) as well as the original concept of the IA

at the moment i am thinking of building an IA simulator, that would step through a chosen historical week, minute by minute, (one minute would be 1 sec) which would allow me to see how an actual IA moved around, (anchor jumps, tandem jumps) etc. during the week in about 2hrs,

something similar to A-Samir but with a historical basis.

Regards, John

Some random comments:

The idea, that a strategy traded against an aggregate of pairs (like the 8/14 pair basket) can outperform a single pair, is problematic for several reasons:

- There's no Support/Resistance on the basket chart. It might look like there is, but there isn't
- A very small percentage of traders around the world are trading against these charts
- Contrary to popular belief, these baskets are not true (or synthetic) hedges
- The basket charts are not smoother than single pair charts (compare your basket chart to the EU chart - without looking at the prices, can you really tell which is which?)
- The cumulative spread of the baskets puts you at an immediate disadvantage, especially on the lower timeframes
- The delay when opening/closing trades for 8/14 pairs is more-than-likely not going to be beneficial to the trader
- When disaster strikes (lost internet connection, huge spikes in price, lack of liquidity, etc), you have 8/14 pairs to worry about instead of 1

I'm not saying that trading these charts can't be profitable. There are many traders on the other T101 thread that consistently show profitable trades and say they are trading profitably on a live account. Now, whether you believe them or not is entirely up to each person to decide. But, with all things considered, it's got to be easier to profitably trade currency pairs individually compared to the composite basket.
-------------------------------------------------------------
The idea of using currency indexes, strongest/weakest currencies, and other currency-related concepts for profitable trading has merit and is probably worthwhile to look at. Those things can give a different type of edge that the 1-symbol strategies can not.
-------------------------------------------------------------
However, for my money, simpler is always better (particularly so for automated trading). Most systems are over-complicated. I spent some time this summer with a "guru" who taught a system which utilized 6 different timeframe charts and had multiple indicators and lines all over the place on each chart. In addition, there was still plenty of discretion involved. Now, that is a complex way to trade. I'm not saying it can't be profitable...but it is more complex than it needs to be.
-------------------------------------------------------------
There is nothing wrong with using an indicator or two on your charts. Indicators won't hurt you as long as you know how to correctly use the data they represent. Price is the king...but, indicators can help "summarize" price into more usable numbers. The challenge is finding the indicator that gives you the information you need to make an informed decision about entering and exiting a trade.

From what I have seen, the VQ indicator is certainly not the answer.
-------------------------------------------------------------
Anyway...I think jmw1970's idea of thinking outside the box is a great. Revisiting the IA concept makes sense also. I think I did a IA simulator awhile back, but it really did not do much for me. It was cool to see how the pairs reacted over extended periods of time. But it always seemed a little goofy to me to base trades on that. Once again...many folks seem to have had some success with it, so it seems like a viable option.
-------------------------------------------------------------
As they say...there are many ways to skin a cat. Forex is no exception. jmw1970 will be a successful trader (if he is not already). Why? Because he is very smart, has excellent technical/analytical skills, and a desire to succeed. Everyone would be very wise to keep an eye on his developments...I know I will continue to do so!
-------------------------------------------------------------
Looks like disbellj might be up to some interesting stuff over on the other thread...something to also consider.
-------------------------------------------------------------
Oh yeah...when evaluating a trend-following system, I think it wise to spend about 90% of your time figuring out how it performs when the market is not trending. That's where backtesting can be a real asset - finding periods of time where the charts look really ugly and finding a way to get your trend-following system out with minimal damage. You know during trends it will perform well, but putting your system up against nasty situations is how you determine its true performance capabilities.

OK

while waiting for responses to my "teaser" posts about thinking outside of the box, which did not have the desired effect (with the exception of TraderDuke & zznbrm) - I decided to read through the original T101 thread (which explains phase 1 - though not 100% clearly) and the less publicised second thread (which explains more phases 2 & 3 - with even less explanation).

After this marathon reading session I have come to the conclusion that the basket trader I designed is quite simply not up to the job - it is fine for trading complete baskets, however for trading the complete method correctly, it sadly will not do.

This is what comes with jumping in and creating something when you have no clue!!!

I still believe the EA I have provided produces benefits, and I still also believe that by creating a TLE that mirrors the original IA concept, coupled with using xxx indexes a capable robot can be built (that can at least trade phase 1 correctly / full basket), although this would be a longer term type trade of maybe a couple of days, not the scalping people have been looking at.

So as my time for private projects is now limited by actually having to do some real work for a bit, I have decided to hold off on further developments to this bot.

I will be still looking however to create a full basket TLE, but as mentioned, this would be more of a daily type trader, not intra-day scalping - once I complete this, I will upload it to the thread.

Regards, John

Hi John
I am another of the silent followers to this and the other thread. I can't thank you enough for the effort you have put into this project and for sharing it with us. I am currently testing the VQzz 15 40 Tle and really like the concept and look forward to future evolutions.
Ray

John
I have been using the bot pretty much since it came out semi-automatic

I have been placing buy/sell stops where I think PA will go and once in profit I move up the SL line to make it trail.

I also put the SL very close and if the trade goes against me I take the hit. Better a small loss and then get going on the next winner.

Thanks for all your hard work

As far as indexes go I have been looking at the DJ30 DAX and USX
nothing really jumping out yet

Lately the markets are news driven all based on events of the default in Europe. Good default news everything up bad news the opposite too bad there is not a default index LOL

Honestly guys,

T101 himself does not use the original concept anymore.

When I mean original concept, I mean using the IA to determine trades.

Lets just say this come from a reliable source.

He's now analyzing offline chart to make trades, probably not all 14 pairs, but a selected few.

Most of his trades are combination of technical analysis, but this time involves lots of fundamentals.

I'm not supporting or opposing both methods, am just saying.

I always enjoyed creative works here, and the other thread of course.

Wanna get the real basket knowledge, sign up for his course lol. If its still available.

I am not promoting and in no way afffliated to T101, obviously.

I agree whole heartly with most of what you say, zz. I think your comment above concerning currency indicies is spot on, and that is the missing link.

I also agree that techinical indies still have the same pitfalls as any single pair, although I think they may work a little better with the basket. Still, that is always going to be the case as long as indies FOLLOW the price, and they always will, unless we figure out a way to go back in time.

Also, have the same observations about supp and resistance. When the basket is ranging, it is merely the basket "regathering" while all the various whales deciding what is currencies are going to be strongest next. It then has a spurt.

Currency indexes may be the "thinking outside the box" that JMW is thinking about.

I have been looking for an index that might work, but with little result. I tend to over-think it, or not always see the obvious.

I always had pretty good luck, trading the OREST dashboard, and my thought is this; If the basket ranking on the BOT could tell us something. perhaps a basket balance of XXX combined with a particular combination of "leaders" could improve the odds. Or, revamping the OREST to an 8 pair basket. I have fiddled with this a little, but since I know little of coding, so no luck.

Also, think that the number of pips you can expect to make from any given trade is limited, since it is only 8 pairs.
Expecting or waiting for a trade to go 1000 pips on a 15 min chart is probably pretty unreasonable, since this would mean each pair individually moving 125 pips average. 400- to 500 (50 to 60) per pair is more reasonable. Of course, trading on a 4 hr or daily with larger stops would make a 1000 to 1500 pip trade pretty do-able)

A loss of 250 - 300 pips (including spread which is around 45) is pretty easy to overcome. Beyond that, the recovery seems too time consuming.

thanks for reading, Shake.

I like the HMA kos style and think it is a good plan. However, I am just not GOOD at that one.

Here is a trading plan with the NLBB that I have been playing with for the last 2 weeks. It has been successful for me, with about 950 pips last week. I trade this live on my little account.

I have been trading this manually, using the BOT to open and close trades. It dawns on me that this might make a TLE, if we can get the logic right. I have tended to use judgement and intuition for taking profits, and accepting losses, which don't usually code very well for BOTs.

Needed: NLBB settings 36 period (default)
NLBB setting 72 period (change colors to suit for recognition).
SYN APB candles (optional)
Time frame: 5 or 15, assume it would work with any.

A few rules:

1. As with most MA crosses, the cross doesn't happen until most of the move is already gone. Therefore, in this case you are opening on a color changes and price crossing both NLBBs.

2. ASSUME that when the price is between the two NLBBs, that it is ranging. Therefore, never open a trade when the price is between the 2 NLBBs.

3. Try to keep losses less than 250-300. If price moves back inside the 2 NLBBs before you are in profit, give serious thought to closing and accepting a reasonable loss.


ENTRY RULES:

1. Wait for price to cross and close both NLBB.

2. The 36 NLBB to change color, maybe both (strongest)

3. Wait for Price to retreat back in between the two NLBB's

4. Set pending BStop/SStop at the small high that was created by price breaking the two NLBBs

5. Wait for trade to open. (price to come to you, here kitty, kitty)

6. If the price eventually moves back below both the lines, it may create a signal the other way. In this case, you can either cancel the Pending Stop you have already set, or wait or place a pending Stop the other way, making a straddle. (I have actually had some luck with this)

Trade management:

1. Set stoploss just other side of 2 NLBBs, or at previous small swing low

2. If stoploss is taken out, wait for next signal. Might happen very soon.

3. Once in profit, trust the NLBB colors. Once the candlewicks quit making HHs or LLs, close half, or 2/3 of order, move Stop to BE to protect profit. Usually, for the most pips, you will have to live through a retracement back between the two NLBBs.

4. If price resumes to more profit, trust the NLBB colors. STAY IN until stopped out or color change.

5. Trail stoploss as you wish.

Here are a few examples, one loser, and a couple nice winners.

I hope this is not taken as off topic. Again, I am posting in case there is any interest in TLE, however, I would appreciate any ideas or tips concerning the system itself.

Thanks for reading, Guys,

Shake.

Hi John,
Wondered when you mention currency indexes if you meant something like the cluster indicators.
At first they look like a lot of squiggly lines but on closer study something remarkable is revealed.
Searched the other thread, but there was only passing mention, so not sure if they have been investigated in any depth.
In the attached chart, I have replicated period shown in Beshakey chart. Have blanked out other currencies to just plot USD vs EUR and the result is quite obvious.
If settings can be tweaked, could it be possible to use the turnaround at extremes to buy/sell baskets?
EP

well something similar, I was more thinking about the indexes available on LiteForex (USDFLX, EURFLX, FLXJPY) - but basically looking at it the same way

in our baskets we are trading the USD & JPY (plus CHF if 14 pair) against everything else

so my idea was to combine the "other" and the usd,jpy,chf to give a similar type picture ie. just 2 lines

then look for a definite coming together (possibly by combining the 2 lines total to then look for a retrace of a fib number - to avoid whipsaw)

the same with the IA, the IA tells you what the market is doing but the question was always in the thread when do you enter or more specifically when do you re-enter

don't get me wrong - i am not against normal indicators on the offline chart, it's just that as up to now a "normal" indicator hasn't been found to trade a standard chart, something tells me we are not going to come up with one just because we are using an offline multi pair chart

hence my thought - lets think outside the box abit