GBP/USD hourly fib scalp ideas

Research and development work in progress

Here's an example of fib scalping as a backup plan. I was watching for a different hourly setup based on the Three Strikes method but the initial stop for this long trade would have been too wide.

Pulling a fib across the latest swing with a stop maybe at the pivot and a fixed target of 1.382 would give some pips while waiting for a better setup, or alternately instead of setting a fixed target, the stop could be moved to break even asap and then maybe the longer term trade can establish itself and I could switch from scalp to long term.

Easy 40 pips with reduced risk compared to the longer term method I didn't want to go with at the time.

I had a short setup that was a valid trade on the Three Strikes method except I didn't like the increasing volatility of the bars leading up to the break out, ranging up to 100 pips and growing each time. The risk of a huge stop getting hit was not very welcoming.

So with an otherwise valid looking short setup I went for the safer fib scalp, pulling across the low I was trading the break of, and the high that price made before heading back down for a break attempt.

Price fell very fast straight down from break out point to the area between the fib 38.2 and 61.8 extensions within 5 minutes without even retracing until after that point.

Another hourly cable example of a valid short trade setup at the blue arrow from the longer term trading method but the initial stop (upper red line) would have to be too large so I pull a fib across the wave from the intended break out low peak to the high peak before break out, set a fib scalp order with fixed stop at fib pivot (50%, lower red line) and target at either 38.2% or 61.8% extensions would have been hit.

Note I don't just pull fibs across random waves in these examples, the break out areas are valid as determined by three strikes trade setups and are only avoided in that method due to the large stops relative to recent price action. In this case notice how far that break out bar had to drop from the fib high line down to the blue arrow to break out, compared to the size of the bars leading up to that setup. That's why I switch over to the fib scalp and set a smaller stop and smaller TP and it has a good probability of making the fib targets.

Another use for fib scalping is trade chasing. If a certain trading system is being used but a sudden volatile break out occurs and the main trading method doesn't catch the big move, pulling a fib across the individual bars of this surge can allow formerly taboo trade chasing with controlled stops and reasonable targets.

The bars should be volatile compared to those leading up to the missed trade - at least twice as big preferably. As soon as it is clear that a trade is being missed and the first big bar closes pull a fib across it. If the bar is broken again in the missed trend direction take a scalp with a 61.8% bar retracement stop and 38.2% extension TP, or optionally let it run if itseems like it might be stable. Maybe the trend can be captured and stops can be trailed hourly behind bars.

Here's an example of 2 days in a row for January 2008 where a sudden volatility came about and several fib scalps could occur. The bars prior to each break out were relatively smaller and then suddenly a single large bar thrust out beyond recent support/resistance wave peaks out of nowhere.

Another day within the same vicinity of the last example, Jan 10 started out with a minor consolidating sideways movement and then suddenly a break out with a volatile bar to the long side. Leading up to the long break the largest bars were in the 25 pip vicinity while the break out bar was over 50 pips. A fib across that suddenly volatile bar with a 61.8% stop and a 38.2% TP would give a scalp of maybe 15 pips after spread and entry padding.

Before all of that there is an attempted break of the consolidation to the short side but that bar is not a sudden volatile shock, it's also in the vicinity of 25 pips like several bars before it. Even if it were considered for a fib scalp there would be no trade because that break out bar was not further broken so an order would not take hold.

For now I'm somewhat cherry picking examples because I only just thought of this yesterday and haven't done formal back testing. Usually when I am on to something I'll do at least one month of straight back testing to get a feel for the probability of success and then continue forward testing but at this point I'm posting examples so I don't forget. I'm going to be doing some unrelated back tests soon so I'll keep this scalp in mind when I go over the charts.

Thexder,

Here is my intial document as mentioned in the private message.

Steve

Here's the first setups of the 2 first days of August 2007. I can only get a 15 minute chart to go that far back in history so I can't see what the trails would have looked like but I can see an early stop on August 1 that would cut profit short. But my fib is drawn as shown. I start the trades at 12am EST so I look for the last clear and sufficient ranged (at least 40) move to draw across. On that day it was the high left at the end of the last day and a low in Asia. The low break came around the 12am mark. Since I can't fine tune the lower time frame I don't know if I'd have a small gain or small loss but at least it would be small if it's a loss. There were sometimes entire losing weeks doing this but then suddenly a 100+ move comes to balance it out. I can't walk the fib since I only have M15 going back to August.

On August 2 I see a very bad coil with not just 2 but 3 levels of high and low coiling prior to any break out, and it just feels very ranging looking at the fib area - candles almost span the entire fib boundary after it is established. In my other Oct/Nov notes I saw in times where candles almost span the whole fib range, it may be wise to avoid the trade because noise can go past the high/low easily, nothing is well defined and worth trading.

Before 12am that's where I would pull the fib across the up and down clear movements, but I'd be probably staying out of that one. And notice there's a false small penetration long and short, as I would expect in such a time. I learned more as I progressed and learned times not to trade, same thing happened at the start of this month, 3 days of nothing but losses and then zooming out on the chart there's a big inward spiral. When it broke free it provided a big move trade.

That's one problem with subjective things like pulling fibs and interpreting discretions. When I made the Oct/Nov notes I had done one full month of testing, then learned some things and began to interpret it differently and I started all over with different results. And after it was all done and I was content, I even forgot my own rules and had to read my own notes to remember how to trade it.

At this point in time anyway, I'm becoming more and more interested in that new volatile fib scalp idea I came up with last week. It doesn't take much effort to notice a sudden large bar showing up, and I am seeing it seems to possibly work on 4 hour as well as the original hourly chart I spotted it on. The 4 hour chart is easier to work with and the risk/reward is the same since it's fib ratios (although the risk is bigger than the reward since I allow more retrace than TP, some will despise this but I only care about pips in the bank over time. There's more to successful trading than comparing your initial stop to your TP). But on the 4 hour chart, instead of getting maybe 10 pips per scalp with fixed TP across a typical bar, you might scalp 30 on average per bar and spend less time looking for it.

When I get time I will try to test this over time and produce a document if it seems worthwhile to consider. If so, I think that will become my main scalp method. Right now I'm not actively scalping at all, I'm working on the three strikes method.

Thanks Thexder for the quick reply.
I will post my August results when they are done and then progress to September. Just out of interest, why are you not trading your 1hr scalping method with the 1min trail? I thought it was working well for you?

Steve

It requires sitting for sometimes up to 12 hours from 12am to 12pm my time if I want to consider every setup in the trading hours (I would often end the day after +20 gain) and I have to be watching carefully so when the break comes I can start following it. I guess a few months of it started to catch up to me so I am trying to ease the time. If I must scalp I'd prefer to just have to check in once per hour at worst case so I'm trying to come up with something to let me do that.

Yes fair enough Thexder. I understand it will be very laborious, but a few price alarms to inform of you of when trades are triggered should help a little, but yes, you need to keep an eye on things.
Will be interested in seeing what you come up with in your latest findings.

Thanks again,
Steve

Here's part of something I'm working on with cable 4 hour. This is from July 25 2007. Not tested fully, just notes I'm making to see if I can get something consistent.

To start there is a consolidation of doji's at the end up an up move and so it's not at all clear, need to wait for something to show direction. A solid down bar comes to hint at what might occur but since everything is still unconfirmed, that could just be one of those fake outs. A pullback comes with an upward retracement and then another volatile surge downward. That's the one I'd be interested in, now we have a direction. Pulling a fib across that surge I see a retracement only as far as 38.2 and then an extension 38.2 TP.

During the trade while waiting for the TP there was a consolidation that will now be evaluated for the next setup. A decisive looking short movement follows the consolidation (which lead to the first TP) so the momentum is short still, then comes a pullback and another clear looking momentum toward the down side including a volatile surge again. That's the one I'm interested in again so I fib that and get a fast 38.2 TP extension without much retrace.

When I first caught on to this I was noticing the successful trades of volatile bar scalps were in line with "good" setups from other methods that I didn't want to trade for some other reasons like initial stops too wide. If I choose random volatile bars then I can't guarantee a successful scalp. So when trying to develop the stand alone fib scalp I have to look for a "setup" in lieu of another method's setup. So I'm looking for a momentum with a pullback as a stab at it. Putting the fib across a single 4 hour bar could be considered the same as pulling a fib across 4 one hour bars and my overall mentality is usually hourly charts so I'm ok with pulling across a single bar like this.

I just posted a preliminary study of one way of scalping volatile candles on cable 4 hour charts. I only did one month of testing and it's just a rough note reminder to work on later.

There were only 4 trade setups (3 considered and the other doesn't trigger an entry) in the month and the net result is somewhere under 100 pips but it's only a scalp. It could be investigated with other charts to see if it works and maybe more charts could be traded if it is successful. Otherwise it could be a supplement to other methods if it works over a longer period of testing. Maybe the month I tested was just a slow month also. More work needs to be done.

Already doing this testing has spawned a new idea for hourly charts again and I'm doing some tests. No time is ever wasted.

Nice one Thexder, sounds like you are really putting in the work. I have just completed my first months backtesting for August 2007 using the 1 hour scalping method and trailing the stops behind the 1 min chart.
Performing the backtesting really gets you fluid at this and it becomes quite easy to implement. Very pleased with the results, 447 pips in total and an increase of 31% of the account, trading 1 lot on a $10,000 account. I would probably start trading a smaller amount with an account this size maybe half that initially but that is more than i have mustered before so all good.
I have enclosed all my trades in the document attached, maybe you could take a look and see what you think. Would be good to hear your comments, but i know you are busy.
Will let you know how September 2007 goes and then i will try Feb 2007, which looked like quite a nasty month.

Cheers,
Steve

I'll look at the test notes when I get the chance. Right now I'm back testing Sept 07 with my semi-volatile hourly hybrid I just thought of after doing the 4 hour volatile scalp.

You might consider seeing how Oct 2007 also works on the 1 minute trail scalp method - that was another bad sideways month and I never tested it with this method, only the old scalp method with fixed TP. I was always curious how the "new" scalp would have done. Thanks for the effort, it helps me stay motivated also. Sometimes I jump around so much I lose sight of what methods were good or not so good performers.

Ok I took a fast look, couldn't resist. I like that 102 pip August 6 trade. That's the kind of moves I frequently saw missed in the Oct/Nov 2007 scalp original method, I'd take maybe 14 pips with a fixed TP and then see 100 go by so I started the M1 trail.

And the early stopped long on August 8 that ended up going far after an early exit, that is the kind of observation that is driving this volatile scalp supplemental research right now. I want a way to get something out of it if I miss it, even if it's only 4 times a month. These are those times.

Just read it now. So the average per week was 111 pips, which is in line with what I've found in my own live trading where I was getting an average of 150 a week in good times, sometimes over 200 but sometimes just maybe only crawling out with +30 after bad ranging. Overall I claimed a conservative average of 100 weekly pips in the long run and 150 average a week in good times.

I wonder if some anti-range discretional filters would help any, or if we would be messing with a good thing. Sometimes I would take trades knowing times are tough but sticking to the brute force rules, knowing the big one will come some time.

Well, if I get these new scalps going, they could help make up for the draw downs in this scalp. That's one idea I had, if I use a 4 hour scalp that only gets 3 trades a month and 90 pips, or around 20 a week average, those 20 plus the 100 to 150 a week on M1 scalp, and any other scalps added, it could be worth looking into. There's not much extra effort involved in a 4 hour scalp, when the setup does come it's a set and forget pending order entry.

I added a new Cable hourly scalp PDF - The Volatile-Momentum Hybrid Scalp.
It's preliminary and is inspired by the 4 hour volatile scalp. I did one month of back testing to get started, the same Sept 2007 period as the 4 hour scalp.
Assuming they were chosen properly in the first place and could be repeated in the future, all trade setups that were valid were successful from what I can tell.

Looking at the way some of those break outs moved including several that could have gone up to 100 pips on a single bar, I'm sure using a lower time frame for support/resistance stop management would greatly increase the return on this.

After spread and 1 pip for entry and exit targets I count 287 pips for September 2007 from the hourly volatile hybrid. For a lower maintenance scalp that can be just monitored once per hour with set and forget orders, that's good if the long term success is the same. Using lower time frame trade management could probably add another few hundred monthly pips possibly. Combining this hourly scalp with the 4 hour low maintenance scalp, there's another 90 pips added for that month.

The interesting thing is although the 4 hour scalp only produced 3 trades in the whole month, one of those was during a major up trend that the hourly scalp missed at the end of the month because there was no place to anchor a fib and get in.

It's all coming together. I wanted to create several methods that can complement each other and that's what's happening.

Nice work Thexder, i am really enjoying all this. I am probably going to test at least a years worth of data to see if this really works and i will publish all of my results. If it can produce the kind of results we have got consistently then this will be great, you can't catch all the moves with any system so i am not concerned too much with those missed moves.
The only thing that gets to you is if you take a day off or miss a move that would be vital to adding up the pips for a week but thats the way it goes with any system or discretional method of trading. You never know the outcome of a trade and this is what i have been eagerly waiting to develop, something where i am not worried on the outcome of a single trade as i know over time that the collection of trades will produce a profit, that is the key to trading.
However, running a separate strat alongside this that may collect those extra pips may well be the answer.
Anyway, i will post my monthly results as i finish them.
Cheers,
Steve

Thexder,
I was just wondering how far behind the swing points you place your stop loss. I am currently working with 1 pip behind the swing but it is quite often that we see price come right up to the pip of the swing before reversing back again. When this happens, the trade result is touch and go, depending on the broker you are using, and the tricks they can play on the shorter timeframes.
I was wondering how you have faired with the live trading in situations like these and whether or not it may be better to keep the stop 2 pips behind the swing to allow for this.
I don't think it would make too much difference to the overall results if we added another pip to all the losses we took but in live trading it may be the difference between a very nice move and a loss.
Currently testing with 1 pip but will keep an eye out for the number of occurrences that this happens.

Steve.

Another thing that springs to mind is the widening of spreads around news time. Currently backtesting a NFP day and i was already in a trade well before it took off, but i trade with Oanda and i know that they can widen their spread up to 15-20 pips on cable around this volatile times. I have no problem with this as this is just the market conditions, less liquidity as everyone pulls their trades and they are trying to avoid being hit by the news traders but the question arises of how to deal with this if you are already in the trade.
It would be very annoying if the widening of the spread takes your stop, not due to the actual market price but just because they widened their stop. Maybe widening the stop to allow for this 15 pip spread when you know its about to happen may work. Its usually only for a few minutes and then the spread narrows back to normal. You may get a hit for a few more pips on the losses but i think by ensuring you stay in for the wins, that this will more than make up for those losses. Its a hard one to work out though through backtesting on a different platform.