Swingspot

[mer071898]

I had put this strategy on the back burner a few months ago because I was still working on my Synergy CRS strategy and because I was stuck on how best to profit from it. Once I finished my Synergy CRS strategy, I decided to take another look at it and after about a week, it hit me.

The strategy originally was based on a ZigZag indicator that people were using to "attempt" to pick out the highs and lows on a chart. The indicator, as I already knew, was severely flawed because it grossly repainted and would redraw the arrows whenever a new high or low was made. Obviously most experienced traders wouldn't touch it with a 10 foot pole, but then I got to thinking. So I had a buddy code so at the close of a candle, it would permanently leave the previous arrows on the chart as each new high or low was made and I had him change the arrows into boxes. As I stared blankly at it over and over for a couple of weeks, I caught on that after a few strings of boxes, we would have a strong correction, as the market does (like clockwork). Then I thought, if we were to use a progressive lot sequence that would slightly increase as we added more boxes on to it, we could essentially turn it into a "basket" of trades in which we could code up an EA to place the trades, then monitor the overall profitability of the whole "basket". Once the equity in the "basket" reaches a certain dollar amount, the EA would close all the trades at once.

I've attached a pic of what the indicator looks like on the chart. As you can see, we'll have a string of boxes and then a correction. Most of the time the correction surpasses the whole string of boxes, and like the last set, we may only catch the majority of the string. But looking at the last sequence of 13 boxes on the chart, if the first trade started at say .1 lots, and we progressively increased it to 1.3 lots (adding .1 lot to each new box), we would have easily made profit because all the larger trades below that particular correction point would overshadow the smaller trades above the correction point.

entry and stop.gif

Here are some sell examples from back in October, All profitable!

entry and stop 2.gif

My original thought was to enter on each box with a progressively higher lot size and then close all the trades when an opposite colored box appeared. I soon found out that after some sequences that would go 40, 50, and even 60 boxes or more without seeing an opposite signal, that I was thinking along the wrong path and I scrapped the project until now. I hope that we can delve into this as I almost think that this could be as good as, if not better than, my Synergy strategy. I have attached the zigzag indicator (which I've modified and called it SwingSpot Conservative) along with an indicator called i-breakeven than monitors the total equity of all the open trades on the chart. Maybe it can give us some code that can be inserted in the EA code. I also had someone code up an EA that placed trades and closed them on the opposite signal as I mentioned above, so if Steve or someone else wants to see the code used in that one, just PM me with your e-mail and I will send it to you.

[r2997790]

Mer, this looks very nice! How do you pick the top of reversal to know when to exit the trade? The scaling up of lot sizes looks like a nice idea too. Do you just run on H1. Have you testing with many pairs of just EURUSD so far? Nice job.

I do hope Steve, Squalliou or one of the EA coding gurus picks this up ... it looks excellent.

The indicator might need to be stablized too, as I went dizzy when I shifted the chart and it did this

oops.gif

[Pat1]

Mer,

Applying your Synergy concept of a range to Swingspot, you have a sort of dynamic range with a predetermined trade direction. If the price emerges from a range created using the highest high and the lowest low of the of the last three points, or highest point and lowest point, for example, will 1. indicate a start of a correction, 2. establishes the point of entry using all disregarded points above the range, and 3. close all trades with a gain or loss once an opposing point appears or the price emerges from the end of correction range . Trick is to determine if the last (lowest / highest ) point has been reached to initiate a trade. Nanningbob's MACD overlayed on two stochastic lines tends to indicate extreme points well.

Similar concept can be established using the lowest lows and highest highs outside of the standard deviation bands or keltner bands where the range is established, for longs, using the lowest price outside of the BBands and the lowest low of the average used to determine the the BBands. or Laguerre indi where last three extreme points form the range if indicator is at an extreme and the price emerges from the range and Lag indicator has crossed predefined level to initiate a trade.

Concept is similar to turtle trading method if I am not mistaken. Buy at 60 period highest high with a trailing stop of a 10 period lowest low and opposite for sells, loss or gain, it will work out in the long run. The Blue and Magenta areas in the pic below shows the concept. Though the values for SL for uptrend and downtrend should be different as the market behavior in these two states different.

All of the above indicate a possibility that an extreme point has been reached. Since all of the above have arbitrary settings, only if the market conforms to the system design(setup), then and then only does one participate.

Cheers,

Pat

[garyfritz]

This feels a bit like a Martingale system -- it works fabulously, until it doesn't. And when it doesn't, it might eat you alive. You'd want to trade it with small enough size that you could survive a move that just kept on going. Long-term testing might give you an idea for how likely / how often those "eat you alive" events happen.

[mer071898]

Mer, this looks very nice! How do you pick the top of reversal to know when to exit the trade? The scaling up of lot sizes looks like a nice idea too. Do you just run on H1. Have you testing with many pairs of just EURUSD so far? Nice job.

I do hope Steve, Squalliou or one of the EA coding gurus picks this up ... it looks excellent.

The indicator might need to be stablized too, as I went dizzy when I shifted the chart and it did this

Well, I have not finalized the way to close out the trades. My first thought is to wait for the correction to reach up to the first trade in the sequence then close out all the trades, so the first trade would basically be a break even trade. But since the price correction does not always reach the first trade, we must look for an alternative closing scenario. We could look at the total equity in the basket and close out the trades once a certain dollar amount is attained. I personally don't like to trade this on anything more than an hour timeframe, otherwise swap starts to be a concerning factor, although you can always take the sequences that earn positive swap. The price correction occurs on any pair, so this can easily be used on other pairs. The indicator does go haywire, but I know nothing about coding, so save a template and when it goes berserk, just reload the template.

Mer,

Applying your Synergy concept of a range to Swingspot, you have a sort of dynamic range with a predetermined trade direction. If the price emerges from a range created using the highest high and the lowest low of the of the last three points, or highest point and lowest point, for example, will 1. indicate a start of a correction, 2. establishes the point of entry using all disregarded points above the range, and 3. close all trades with a gain or loss once an opposing point appears or the price emerges from the end of correction range . Trick is to determine if the last (lowest / highest ) point has been reached to initiate a trade. Nanningbob's MACD overlayed on two stochastic lines tends to indicate extreme points well.

Very interesting view, I will have to look into it. I'm just not a huge fan of having a ton of indicator to rely on.

This feels a bit like a Martingale system -- it works fabulously, until it doesn't. And when it doesn't, it might eat you alive. You'd want to trade it with small enough size that you could survive a move that just kept on going. Long-term testing might give you an idea for how likely / how often those "eat you alive" events happen.

Well, I'm not talking about a true martingale, sort of a semi-martingale that progressively gets steeper the farther you get into the sequence. What you actually doing is scaling in as the higher high or lower low is being made. What I've done is make up several lot progressions that run about 60 levels. They are very mild for the first 30 levels then increase faster on the second half (just in case we would ever get that far), I understand the concern and with previous testing, I have ran into very long sequences where I would see 60-70 plus boxes, but again, there was always a correction that would encompass the smaller sequences which is what I want to target.

[DragosDanescu]

Hi, guys,

Just 2 considerations:

1. Attach a Stochastic and open positions only when it's into deep overbought/oversold area....
2. Open pending orders at high of the previous candle, not all of them will be triggered, so DD will be a bit lower, but they are there, in case of correction.

Cheers

[stimpy]

just playing with SemaphorTrue2_mod for mer.mq4 from Miyagi http://www.forexfactory.com/showthread. ... ost5333573

how would one call those values for an EA?

[hesselvanderhoeven]

Hi guys. I am trading a pretty similar strategy, only using Triangular MA bands instead of the Semaphor indicator. This gives far less drawdown and a better entry position. I attached a pic with a few trade examples.
You can read the full strategy here: http://www.stevehopwoodforex.com/phpBB3 ... f=12&t=566

I am still looking for someone to code an EA of the strategy, I think it could be pretty profitable!

Schermafbeelding 2012-05-19 om 19.49.29.png

[garyfritz]

Do those bands repaint? They look too smooth, and they move just perfectly only when they have to, to do that in realtime. Is that an issue with the strategy?

[hesselvanderhoeven]

Do those bands repaint? They look too smooth, and they move just perfectly only when they have to, to do that in realtime. Is that an issue with the strategy?

They don't.. You can download the indicator on my thread and try it for yourself.
I use it on a 5 min-timeframe. I buy/sell when price pierces through the MA. When it moves 10 pips against me I double my position, when it moves 10 more pips against me I double my position again, until it retraces 10 pips and I'm in profit.