Nexas's Intraday Range Bar Trading

Hi Scjohn,

Thank you for your input!

All I can say is wow!!! I don't even have an 80% strike rate! As you guys are aware from earlier on in the thread,my new years resolution is to be more selective and increase the strike rate.

I completely agree with your take on prop indicators. It makes sense. Do you buy them directly from this guys as individual indicators? or do you have to sign up to a course or something to get them?

You have gone through your filters in previous posts but do you think you could summarize in a few posts here and now and then if you like I will update the first page to include your alternative?

What do you think?

Nexas

Yes, I will do that but it will take a few posts. First, I'd like to talk about expectations and tell you why I'm doing this.

Looks like I'm about to exit a short trade on EURUSD I took at 10:41 EST on a thrust signal. Oops, took another turn down. Runner's up about 50 pips as we speak.

First -- I'm doing this to pay back Nexas for turning me on to trading the 4 pip chart which is producing considerably better results than my old technique.
Second -- I'd like to develop a few more Barry Burns enthusiasts who can help me improve what I do. (As, you'll see, there is room for improvement).
Third -- At some point, I'd like to explore a partnership with someone in the UK. I miss a lot of pips in the name of sleep.

Runner just stopped out at +48 pips. Done for the day.

Re expectations. The trade journal control panel I showed in a previous post shows what someone with a lot of trading courage could have done with the 3 setups (Barry Burns, Nexas, and Thrust). However, being a bit of a wimp, I trade it slightly differently. I usually exit half the trade at about 1x risk which costs me pips but makes trading much more fun and less stressful. Plus, once I bank that first half, I'm pretty fearless with the runners. I usually trade multiples of 4 lots which makes it all easier.

Here's what it would have done (actual plus hypothetical) with the early exits I actually use.



As you can see, the total pips went down as did the reward to risk ratio. The win rate went up slightly. So, this is the wimp version.

Note that, although the results are nice, 43 winning trades got away. So about 3 and a half times a day, I watch winning trades sail off without me. Of course, it's considerable comfort to note that the filters that eliminated those 43 winners also eliminated 135 losers and that the eliminated trades, as a group had rather abysmal results. But perhaps there's someone smarter than me out there who can figure how to keep more winners without adding a bunch of losers.

Now here are the actual results with hypotheticals removed. Let's see if it will let me do two charts in one post. Has sporadically been a problem in the past.

This shows part of what wanting to sleep has cost me -- and my journal doesn't even capture the first hour of London trading (no way I'd ever get up that early).



Time for lunch but I will get to the filters later.

Hi Scjohn,

Thank you for all of that!

With regards to trading partners I think that is a very interesting idea and would like to hear how such a venture could work?

I would be more than happy to trade the London mornings for you and then you trade the US morning for me :-) You get more sleep and I get to go flying in the afternoons :-)

Look forward to your response and to the filters.

Nexas

I promise I'll eventually get to the filters but I think it's important to make some concepts and metric terminology clear.

What the data show are historical results only. Although we have a statistically significant number of triggers, there is no guarantee that the next 5, 10 or even the next 100 trades will have a win rate greater than 80%. However, in the years I've been trading, I've found that using setups that had excellent results in the past have almost always made me smile later. The Nexas setups are relatively new for me, but Ive been using the Barry Burns and Thrust setups (albeit on longer charts but also with excellent results) for quite a while now.

Although I show Reward to Risk Ratio on the model (calculated as average pips gained per winning trade divided by average pips lost per losing trade), be aware that I (like the 3 professional traders I trained with) consider r:r in isolation to be a totally meaningless number.

Suppose I offered you the choice of two trade setups. One has a 3 to 1 r:r ratio and the other has a 1 to 1 r:r ratio. Which would you pick? You'd be surprised to hear how many new traders would immediately pick the 3 to 1. But now I tell you the 3 to 1 setup has a win rate of 20% and the 1 to 1 has a win rate of 80%. Now which one do you want? The much more important metric on the spreadsheet is Expectancy (calculated at (Avg Pips gained per Winning Trade X Win Rate) less (Avg Pips lost per Losing Trade X Loss Rate). An expectancy of more than 0 means the setup is expected to make money (but low expectancy promises a wild ride with big drawdowns). Some pros say expectancy has to be at least one before they'll consider trading it. Wimps like me look for 3 or more.

Pucker factor is my own metric and shows how many average wins it takes to recover the maximum drawdown. Drawdown is one of the major reasons most new traders fail. No one told them that a setup that has a nice win rate of say, 65%, can be expected to string together 4 losers in a row about once every month and 7 losers in a row about once a year. And this is assuming it's working exactly as well as planned! Elementary statistical analysis tells us to expect this, but they assume their setup is not working any more even though it's a perfectly normal event.

Note that the model I trade has had a not-so-wonderful reward to risk ratio, but great expectancy and a low pucker factor.



So please don't come to me saying that "I've been taught I need a 2 to 1 reward to risk ratio." Expectancy is what counts to me (and most professionals).

Also note that the maximum draw currently shown is what actually occured during the limited time period of the data. With only 100 trades in the Nexas model, experience (and again, elementary statistical analysis) tells me there are bigger drawdowns in the future, perhaps as much as 2x the model's number.

Back with more later.

Definitely worth exploring but let's get through these analyses first and see how you like the setups. Having someone trade a joint account (perhaps owned by a LLC) during the London session with me trading the US session is exactly what I had in mind. I think it's too hard to try to trade two accounts with a short trigger day trading chart.

Trust me, I'm getting to the filters but I think it's important to understand why something works, not just blindly follow a setup. The "why" is what helps you stick with the setup even when it goes through its inevitable rough patches.

If you look at the upper left hand corner of the trade journal results page, you'll see a list of five factors. These are what we're trying to measure with our indicators. Think about what tends to make a trend trade succeed:

1) Are we with the direction of the trend?
2) How strong is the trend?
Are there significant impediments to the trend:
3) Support / Resistance?
4) Is a consolidation period likely?
5) What about momentum? Is the trend strength increasing or decreasing?

While it's possible to develop a profitable setup with only a couple of these factors considered, it's not likely to be a high probability one and will almost surely produce some bodacious drawdowns. On the other hand, trying to take too many things into account increases complexity, increases the danger of over-optimization (curve-fitting), and can be difficult to actually trade. So, as in most things in life, it's a balancing act.

Although I take five factors into account, the way I do it allows me to make a go, no-go decision in less than a second, so it's not difficult to trade. Five factors, as long as they're all measuring different things, do not pose a huge danger of over-optimization. Using one and only one metric to measure each factor for each setup helps reduce curve-fitting risk. Have you ever seen a new trader's chart festooned with Stochastics, RSI, and CCI? These are all cycle indicators trying to measure essentially the same thing. There's an old saying, "Person with one watch knows what time it is, person with two watches is never sure."

Another important point is that none of these indicators is ever actually "right" or "correct" or, by themselves, even useful. It's the agreement of the five factors (or at least most of them) that creates high probability trading scenarios.

Look at the following attachment. There were 286 Nexus and Thrust (I've eliminated the Barry Burns trades as these are filtered differently) triggers during the 13 days. The 105 triggers where 4 of the 5 factors said "go" made 713 pips and had a win rate of 82%. The 181 triggers where 3 or less factors said "go" lost 759 pips and had a win rate of 25%. Now try to convince me that was merely a coincidence.



Of course the next logical question is, "what if we allow two bad things in a trade?"



What if we allow no bad things at all? All five factors have to agree? It's a very expensive way to increase the win rate.



And finally, what if we use no filters at all and take all triggers? The ultimate simple setup.



More tomorrow.

I've been enjoying your posts scjohn. Keep them coming. I totally agree with your last post on filters. I believe that in order to be a successful technical analyst, you actually have to do analysis. Most people just want to wait for a bunch of shiny indicators to line up, and be successful without doing any analysis. 2 of my favorite filters are volatility, and S&R. Look forward to see more of your analysis.

Okay let's talk about filters. First we'll start with unfiltered Nexas trades on the 4 pip range chart with an 8 pip range chart guide. Entries and exits are as Nexas explained in his original post and in the clarifications he recently posted for me. In all cases, EMAs on both the 4 pip and 8 pip charts were lined up with the 20 on the correct side of the 60 and the 60 on the correct side of the 100.

Let's look at the unfiltered case first. As the Nexas setup is essentially a scalping setup (close targets), the model suggests the best approach is to take everything off if profit reaches 2x risk or 13 pips, so we'll start with that. Of course, if it hits the hilo and then has an opposing color bar, it will also take profits there even if before 13 pips. Results are net of trading costs (spread, commission, and slippage) of about 2.6 pips per trade (the one pip back of the hilo Nexas uses in his stop is also included in this number).



Although it had a heady day where it pulled in 63 pips and made money on 4 days, the overall results were not good. I did not do an analysis of the 8 pip setup but my perusal of the charts indicated moving up to the 8 pip would only produce a marginal improvement. The basic problem with any trend trading system that does not employ some kind of filter to provide early warning of potential trend changes (like momentum) is that they will do wonderfully on days when the trend does not change during trading hours and be massacred when trend changes happen during the trading window.

Now let's apply all the filters (I'll discuss them later) and first require that ALL of them be in a "go" position in order to take the trade.



A much nicer result. 10 winning days, a few breakevens, and no losing days. 89% win rate but very few trades -- less than 2 per day in 6 and a half hours of trading.

Now let's allow one (but only one) of the factors to be bad or "no-go." We don't care which one. (I'm sure there's some complicated formula out there that works better than this but I don't know what it is and it's probably too hard to apply in a manual trading environment anyway.)



While the win rate goes down and we pick up a couple of losing days, profits go up substantially and we get almost 4 trades a day. It also comes close to that magic "20 pips a day" per contract traded.

This post is getting rather long so I'll continue on another one.

Now, the basic problem for those of you who do not want to use proprietary information or indicators is that one of my main setups (Barry Burns) is totally dependent on his stuff and my thrust trade filters are heavily dependent also.

Although some of Nexus filters also rely heavily on Barry Burn's stuff, the historical model suggests there is a version of the filters that can be applied without the proprietary stuff that provides some improvement.

Three of the filters: counter to DAD, momentum patterns, and to some extent into strong resistance, would require access to his stuff.

If you apply the other two filters and require both of them to be "go" the result is significantly better than unfiltered and you may be able to improve it by applying non-proprietary momentum indicators.



Although there are some actual trades in the data taken by application of other filters, I've not actually live traded this setup as I have the other three.

It requires the 20 EMA on the 8 pip guide chart to be moving in the direction of the trade and that the signal not follow a "parabolic" move (about 30 pips from high to low). Consolidation periods often follow parabolic moves.

The setup above is about all I can do for you with non-proprietary information or filters.

Anyone interested in the other setups would have to purchase Barry Burns' first two Top Dog trading courses -- Foundation One and Two. I'm pretty sure I signed a non-disclosure agreement and, even if I haven't, it's not fair to give his valuable stuff away.

The Barry Burns setups would come with the course and, once you've completed the course, I'll show you how to apply the proprietary indicators to my thrust trade and the Nexas setup.

Taking the courses (last time I checked they cost $495 bundled) is not a casual undertaking. This is not some lightweight indicator shop. This is heavy training in market theory and market geometry. They come with 27 thirty minute videos and 2 manuals. Unless you're a lot smarter than me it will take you at least 2 passes through each video and each manual before it really starts to sink in. However, for me, the payback was enormous. I learned a lot more from this on-line training than I did in a one-week, on-site mentorship with two top traders. As you can see, in addition to his high probability setups themselves, I was able to significantly improve my other setups.

When I took the courses about a year ago, he would not sell his custom indicators to anyone who did not take the course. In the courses, he shows you how to program his indicators (you don't have to buy them to use them), but I found it much easier to spring for the indicators also which I believe were another 3 to 4 hundred or so.

I received a private message indicating several folks feel I'm a shill for Barry Burns or perhaps even Barry Burns himself.

I was afraid that would happen which was why I first tried to use off-the-shelf indicators to illustrate my points. Look at some of my earlier posts re momentum. But the problem was that they just don't work nearly as well.

I assure you I am not Barry Burns. I have no interest whatsoever in his company and I don't make a dime on any of his products.

But, given the credibility issue, I'm done here and I wish you all well.

It's great pity, it was very interesting.
Hope the thread is not dead
and also that scjohn will be back

Hi Tomlee,

It is indeed a shame about Scjohn. I can't help but feel a little responsible as I did send him a PM, but more as a heads up about how people may perceive his last post rather than an accusation. I did explain that to him and he did mention that it wasn't because of my PM but because of others who were not very polite in their approach.

I was trying to explain how participants of the factory can sometimes see posts like that as I have had similar experiences myself.

So I can only apologize to to participants on this thread as his input was indeed very valid. Hopefully he decides to return.

What we can take from his posts though is that there are a few ways to improve the strike rate of the system and although he has shown everyone one or two ways, there are a few others that can also help.

I have never been one for indicators as I have done just fine with out them but I am open minded enough to see the possibilities. In fact a conversation with Scjohn last week where he asked me how I filter my trades as he can't necessarily see how it can be profitable without, got me thinking.

My answer to him was yes I look at a few things but they are more discretionary and down to experience. I then understood that this is not something that new traders have the benefit of and hence, if I really want to help you guys I need to think of a more mechanical approach that doesn't include my experience and feel for the market.

I have a few theories which I am testing out and will let you all know once I have finalized things, possibly later this week.

I hope that all helps.

Nexas

Hi Nexas, glad you are here,
indeed it's always delicate here in FF with all this commercial stuff.

BTW, what is the meaning of "counter to DAD" (in the scjohn's filtering) ?

And I am also for scjohn decides to return.

Nice day everyone.

Dear Nexas,

Any progress yet on proposed indicators that can be used with this type of trading?

Regards

Ian

Hi Ian,

Sorry I have neglected this thread a little bit of late. It started to get a little ugly with accusations flying all over the place. Mainly to Scjohn which I didn't think was right so I took a step back.

I don't really want to get into it all again so will just leave you with this short post which is to say that there are indicators out that help you make your decision on when to trade and when not to, but you still need a firm understanding of price action.

I have looked at and tested Stochastics (7,3,3) and it works very well for filtering out cr*p trades.

Wait for the pull back to S & R, Wait for cross of stochs during a strong trend and look for the entry.

I hope that helps. But I won't really continue to post on this thread as there seemed to be a bit to much bad blood (not necessarily in the thread but via PM's etc)

Also before anyone asks I am still trading the same way using the 8 and 5 range bars :-) Love them!!!!

Nexas

Reliable Free Range Bar Indicators

Free Indicators listed at end of post - see Constant Range Detector. Scroll down to - Constant Range Indicators and Constant Range Information

http://codebase.mql4.com/c/codebase/2012/06/CR-2.jpg
http://codebase.mql4.com/c/codebase/...CR-4_small.jpg

Nexas, I daytrade dips in the trend off range/momentum bars on TS. I am also trading off the higher highs of the swings against the trend, sort of when the price coming back to the retracement to start the whole cycle again...would love to correspond with you a bit...how to get in touch?

Been profitable, but looking for more...

Hi all,
Does anyone have the s-constant range bars indi for MT4?
Cheers

See post # 656 for alternatives

s-Constant Range Bars

Cheers.
Coastie