GBP/USD hourly fib scalp ideas

I always have fixed spreads, I get hit with about a 30 second order entry delay in volatile times so if I'm already in a trade I'll only get kicked out if the news reaches my order and I'll be subjected to 30 seconds of slippage. One time I had a pending order triggered during news and the candle went way beyond my entry price (in favor of trend) within the 30 seconds, then my order was filled and I was in a panic of a snap back, but it kept going and then started to hover still way in profit, so I got lucky.

Even after it settled down it still kept going in a profit with support/resistance way beyond entry price. I just take the setups without worrying about news because if I get hit with a bad one it'll be balanced by those good hits just like with overall trading systems. And in my experience even placing orders around NFP news (the worst of the worst) I have found that support/resistance usually holds, sometimes you'll see the volatile activity on the chart but it won't touch your orders (assuming fixed spread) unless it was meant to go that way in a proper trend. I've never really been faked out by news glitches on the chart in other words. If Iv'e been stopped it's because my trade probably wouldn't have ever worked in the first place and the news accelerated the travel to my stop.

The issue on how to pad stops with pips is what lead me to the idea I started doing to use MT4 as an overall setup guide and then observe my broker's chart to see the exact price at those levels. So I may find a swing and want to trade it, then I check my broker and see their swing low is 3 pips lower than MT4 demo so I use that price plus spread plus up to 3 pips extra I think was the magic number I developed.

So with cable having 5 spread overall for me, I notice that my broker chart price is usually centered with 2 of those 5 pips to the down side and 3 to the up side. If the chart says 2.0005, then the broker price is usually 2.0003 sell, 2.0008 buy. If I'm looking for a value to give to a sell stop it will be in that case, 2.0003 minus 3 padding and since it's an even 2.0000 I may make it 1.9997 or 1.9999. I don't often pay attention to round numbers in methods like this, I think of it as superstitious because when it wants to cross the boundaries, it does but there's something about a fat 2.0000 that looks wrong.

I may add even more pips by observing how the broker price compares to its chart at the time...sometimes when moving fast, the quoted prices will heavily favor one side, a chart may show 2.0005 and the sell is 2.0005 with a buy of 2.0010 if the action is moving down rapidly. I think it's a bit of lag mostly. So I'll just work with what I see at the time and this may skew test results a little but overall with observed spread plus 3 I think it's a good estimate for success or failure if it blows through.

When working with longer term trading methods I just use a fixed number like 10 or 15 outside the chart - it's sloppy because the long and short sides of the quoted price will mean one side gets in later than the other but the aim is to be able to blindly cover cost of spread and padding without the effort of fine tuning since it's long term.

Even though I'm aiming to avoid sitting 12 hours a session looking for a break out scalp to babysit with manual stop management, I'm falling back into that habit with the volatile hybrid. I saw a trade setup in the last few days where it would have failed but the M1 management would have kept the stop near the entry for a minimal hit. I need to back test the hybrid with M1 to see if that would give too many premature stops before hitting 38.2% extension or if it would enhance this by allowing the extensions to be reached and even beyond with a manual stop while minimizing the damage.

I think the results should be as good as the original scalp with M1 stops because in hindsight, the hybrid method is also looking for swing setups to trade as breakouts except it looks for more pure setups to try to increase the success rate and minimize the tediousness of having to watch for the exact retracements buried within the hourly candles and all that, it's supposed to be easier and better to use. Time will tell what works best.

Depending what testing shows about how M1 s/r behaves in this method, maybe the trades can be nurtured in the early development stages until they are in free trade and then let it go. If a good trade will usualy make it to the extension with clean M1 activity rather than premature s/r failures, that would be ideal - a failed s/r would mean a most likely bad trade and worth getting out early.

Thanks for the reply Thexder, just out of interest, what broker do you use?

Questrade (Canada)

Here's a V-M Hybrid scalp from this morning. Looking at M5 bars and trailing a stop behind bars or pullbacks in between new highs, the trade could be taken up almost to 61.8% extension and then it fails the whole trade as it takes out the maximum profit permitted with the trail. Good for maybe +40 after costs.

I clicked the H4 button after that last trade and look what else I just found, another +30 pips hidden in the same trade zone with the original Volatile H4 Scalp. If I were paying attention to this I'd have easily allowed myself to double up on the orders in the same direction because they are independently configured and managed differently.

Hi Thexder,
I love the way that you document your systems and have gained many good ideas and hints from reading them My setup and Broker combine to restrict me from using your 1 minute based trailing stop method, so I have to specify targets and SLs in advance ( though I often modify them if I am around as a trade progresses). I find that it is usually safe to use T2 (the -61.8/161.8%) less padding as the target.
You must have noticed that most of the September (theoretical) trades in your document would have hit the 61.8% target. In fact I think that all of them (except for the one on 27th which I will mention later), which do not pull back within the 0/100 before hitting the 38% target, go on to hit the 61.8%. So if you are going to be at your PC for a while, you could set the initial target to the T2 and revise that to the t1 if you get a pullback into the 0/100.

In the case of the September 27th trade, on my chart it came to 2 pips from the t2 - so using my padding rules, it would have been a T2 win. Even if you just use the spread and no padding pips, it would have hit the t2 target if the Fib had been drawn excluding the small spining top candle which preceeded the Breakout candle.

Just my thoughts on what seems a very promising system.

Ian

Hi Thexder,
I just thought of something else. Have you done any testing on taking the reverse trade after the initial one has hit its target and retraced through the whole original Fib ?
I find this usually works,using the original Fib targets though I always set them up with a T1 =38% target rather than the the 62% that I use for the trade in the original direction.

I haven't evaluated it before but now that I have the head start on the month of trade examples I plan to pay attention to the observations such as which targets were reached in the trades, how much retracement had to be withstood first, what happened after targets were met, etc. And in a live scenario I think the way I'd enter trades is with a far away target and then manage with the smaller TF trailing stop so I can make it up there to the expected targets and see what happens. If I find that I don't suffer much retracement in these trades, then trailing the stop should be successful as price moves toward the targets smoothly. It makes sense that the trades seem to flow smoothly because I'm basing the setups around original smooth momentums that are well behaved or sudden volatile moves that should also want to go forcefully forward, and so there's no reason to expect that suddenly after break out, the smoothness will change right away.

I only chose 38.2% targets as an arbitrary safe starting point for the feasibility study Now we can turn it into a system. Today would have given 70 pips on the combination of H1 and H4 volatility scalps. I'm going to pay attention to more testing and observing since it looks worthwhile.

I did some visual scan testing of this month with H1 and H4 volatile scalps and I was getting worried because I kept seeing day after day with no trade setups on either method...then I remembered this month started out ranging inward and so of course the price action looked all chaotic and uncontrolled.

It's like having a noise filter by looking for these setups. Then when things cleared up the price action started flowing and setups came, giving good results in both scalps just like how today gave 70 pips between the two.

I think I'll do some testing with the screen captures of the normal time frame setups and then the lower time frames to see if there's an optimal way to set targets and stops. When trailing M5 stops on an hourly V-M hybrid I would have been stopped out way early for maybe break even last week if just trailing the bars but if I waited and suffered through a -20 draw down, eventually proper s/r waves formed and it would have been over +100. But in general I doubt most trades would last through that type of formation. That's what the analysis will exploit when there's a history of screen captures to work with.

While I'm doing the research I'll take any live setups because I have faith in the probability of success based on 2 months of observation now overall.

I just posted a new V-M Hybrid document with October added to the September testing. This time I showed the M5 along with H1 activity for the month so more analysis can be done with the price behavior.

There were 9 trades logged where one went wrong but I think it could have been avoided, the retracement had an inside bar and later a doji, two quirks in the momentum. Having one or the other could be seen as a normal thing - the inside bar forms a small symmetric triangle continuation and the doji was at the end of the retracement signaling a turn around to resume the trend break out but all at once it's too much doubt in a smooth movement.

Overall it still looks good. October 07 was a bad sideways month and the system easily dodged the weird trading days with the hairy candles. Most of the trade setups this time were volatile break outs from small consolidations rather than momentum sightings, which makes sense in the range environment. So it's good that this is a hybrid system, more robust in catching moves and preventing bad ones.

I count 202 pips for October 07 using only 1 pip entry and exit padding with 5 spread per trade (2 for sell and 3 for buy sides of the entry/exits). That's using fixed 38.2% targets. If the trades are pampered we can get more out of the wins but I'm not sure if some others would be cut short, I haven't analyzed the difference.

200 a month is a good return for my goals and this was in a bad ranging month. I haven't looked to see if there's any 4 hour scalps in October yet. As a comparison, the very first scalp method I came up with before the M1 trail with fixed TP and stops and specific rules for pulling fibs (the oct/nov 2007 trade note pdf)...that method pulled 310 out of October 2007 with the brute force trades every day but it takes a lot of effort to maintain, requires suffering up to a week of constant losses before seeing any hope, and those results were based on a full 12 hour trading day 5 days a week to catch everything. When I do November with the hybrid scalp I can see how that one compares to the first method. I only tested 3 weeks of November 07 on the old method because I was doing the test at that point in the month and I never bothered to go back and finish it when the month ended, but I can compare the first 3 weeks. I pulled over 500 pips in those 3 weeks with the brute force old method.

Now that I think about it, all of this work is done on the same cable hourly chart (except the 4 hour method) and I have a total of 3 different hourly scalp methods to use. If one provides 300 and another provides 200 in a bad month and then there's the 3rd hourly method and 4 hour method, it could be theoretically possible to run all 4 methods at once and manage multiple open orders depending on the techniques used and do well just with that. But I'd have to be up for the screen time.

Thexder,

Did you produce any results for the brute force 1 min scalp where you simply trailed as opposed to closing at fib targets? I just wondered so i could maybe make a backtest for the same month and compare to see how two traders may differ using the same strategy?

Edit: I just ask as looking at your results for Oct 2007, when you go for targets rather than just exiting on the swings, you seem to be discretional on the fibs that you choose to trade depending on what they look like, ie long wicks etc. I was wondering whether you still adopted this discretional technique. For instance, i noticed on Oct 01 you decided not to take the long trade on the initial breakout as you was expecting a pullback from the large move the day before. I understand this totally, but we never know what price is going to do, and it could have broken and continued so am i right in assuming that your strategy is mechanical in its execution etc but the selection of which trade to take is still up to the traders discretion, and will obviously lead to different results?

Steve

hello, i have a question about how you trade the line chart


do you wait for the closing time of the price line, or do you take descision while price line is forming?

because most of time, when you draw Support Resistance lines of the price line chart, no one can beassured if price line is realy breaking the Support Resistance of it.

( Linechart draws price on closing, and do not contain candle spike information, what will give mostly false information of support resistance, and can result in false descision making )

Thexder,

I have attached a question that i wanted to ask about identifying trading ranges within the 1 min timeframe. There is a trade example in my attachment to see if i am understanding this correctly. I would be very grateful if you could take a look when you get a chance.

Thanks for your help on this,
Steve

I never did do an official backtest month of the M1 trail revision because I liked it so much and then it turned out to be working in forward testing well enough so I have no history to go on. But I'm considering doing that if I ever get time and compare everything within the most recent few months where my historical M1 data is still available. If I COULD go back to Sept 07 I'd do a few months around there since I'm doing all this other testing there too.

The '07 scalp was discretional on what fibs to sometimes trade but the M1 revision became true brute force, it was unbiased in which trend side to trade as well as when to trade it, I just considered anything that came starting from the first dormant session swing setup. Even in a mechanical system people will still probably trade it differently and interpret the mechanics differently. Even on threads where you see allegedly straightforward entry rules like wait for this indicator to go red and enter on the next candle...you still see people posting charts of improper setups and asking obvious questions like "do you wait for a candle to close" (that has to be the most common question on this site).

Any time I have a line chart like in three strikes (the first few charts in this thread) there are bars on the chart too so I am seeing the whole price action. Also notice that I have 3 main lines - high, low, and close prices so even if I don't have bars added, I have the entire price action envelope. And I make decisions from the close of the bars.

Thanks Thexder,

Are you using Metatrader for your backtesting? There are some good sites that provide free 1 min data going back to 2001. They were recommended to me by someone. I am sure you are aware of these and probably just using a different broker?

I think it's a valid setup to consider for the following reasons.

The previous trade was complete with its 61.8% retracement and can be considered ancient history regarding whether to keep the first fib in play or allow a new one. Although the first fib information isn't ancient history in terms of it having formed S/R and established a long trend bias to keep in mind.

To reinforce the death of the first fib range even the hourly has pulled back, not just a buried 61.8% retracement on a single unclear candle that could be just a spike, so it's been dead for a while in this method.

The traded break is in the long direction which supports the bias of the original fib. When looking for buried setups I wouldn't want to trade the opposite way from the original fib unless it's at the 100% low of that first fib, not a "new higher buried low" within that first fib range - there would be resistance from the first fib low against the trend and with recent lows being established in the first fib range. But a break long from a buried fib would be in support of trend continuation and at a better price.

As an added bonus, look at the first hourly fib and the new traded M1 fib. Notice your new buried fib long entry price (M1 fib high) is 4 pips above the inner fib 38.2% retracement of that first hourly setup - this buried fib has found a way to trade the break of the inner fibs on the previous hourly setup and is a strengthened way to confirm the hourly chart inner fib bounce/break. Overall we have an hourly long, a retracement, then a trade at the break through the hourly 38.2% retracement level as price resumes upward after the big overall retracement.

That's the type of reasoning behind the scalp methods, but everyone will see it differently and it's hard to make it purely mechanical. If 5 people did a month of back test with the same method maybe it would be good to see different results and compare the notes. I am guessing that overall even if someone else took the failed setup in this example, their good trades would come out on top. Combining everything would yield hopefully optimized new ideas to watch for.


Edit: I think I meant to say Second hourly fib, not first hourly fib. The prices I'm talking about at 1.9940 hourly 38.2% retracement and 1.9944 buried M1 High

I'm using one of the MT4 demo feeds and I think this one has less history than my previous demo feed. I had to change because I was getting 12 hours of lost feeds while others were still running so after a few times I stopped relying on that feed.

I didn't look into finding long term historical data before, any back testing I was doing was just for the recent few months at the time I was investigating so I never had the need. Where can I find the feeds? I feel like the short circuit movie robot. INPUT. INPUT. need data.

Thexder,

I really appreciate your time and comments, they are very helpful. You have basically said what i have been thinking and that is good to know. I didn't really note the break of the 382 fib from the original fib, which i recall from my initial read of Bo's thread was the long fib and a place to initiate new longs, so yes, that adds strength to it all.
I agree that different traders will trade this method differently but i do like the mechanical approach of entry and exit. I am currently finding that there are small discretionary sides to this strat that can be used in certain situations and these are things that you only find out for yourself through backtesting, which you know yourself.
I like this, and i know it requires a lot of screen time but if it pulls in the pips consistently then i will sit here 20 hours a day, lol.
I am going to probably backtest a whole year at least so i really get to grips with it and get plenty of trades under my belt.
Will post as i finish each month. Is it ok to continue posting in this thread?

Steve