Standard Deviation Scalping Method

[chamane]

Hi,

I found a Scalping technique called "5-min Standard Deviation Scalp" (pattern #5 in the document enclosed). I would like to know if anybody has tried this technique.

Briefly, on a 5 min chart, you apply a 10 bar moving average and bands 1 std dev on each side. I am using Bollinger Band (10,1).

When the trend is up (slope of the bands), wait for price to dip inside the lower band and then enter long. Exit when price reaches the other band.

When the trend is down, wait for the price to touch the upper band and then enter short. Exit when price reaches the other band.

To make sure the direction of the trend is right when price touches the bands, I use 10EMA close and 10EMA open. If 10EMA close is over 10EMA open, the trend is up when price crosses the lower band. Reverse for down trend.

Sometimes price doesn't reach the other band, but crosses the middle line of the BB. The middle line can also become the exit signal. There are also false signals (fewer than good ones) and I'm not sure about the stop loss. I would think 1XATR can be a good guess since it's all based on volatility.

On the picture, I circled the entry signals. Sometimes price reaches the other band, but sometimes it doesn't, but most of the times it crosses the middle line of the BB.

Comments would be highly appreciated. If you can watch 3 or 4 low spread pairs, the result should be interesting at the end of the day.

Thanks.

 

[A Dashing Blade]

Standard deviation based techniques implicity assume that market returns are normally distributed.
They are not.

 

[everyonerich]

looks like bollinger band thou..

 

[TheBramble]

Yeah, StdDev has it’s uses, but this isn’t the best of them.

This approach to trading a trending channel (where the channel can be whatever you like, Bollies, MAs, linear regressions – whatever) is solid, and is certainly more sensible than trying to imagine every band touch is ‘the’ reversal to the primary trend which is what many do.

However, I’d suggest waiting until the price comes back into the channel from outside, rather than simply touching the channel.

“Exit on the Left; Enter on the Right – most do the exact opposite and lose”.

I’d also recommend using a longer TF and giving the old scalp a miss. You don’t need to scalp trends…

 

[chamane]

Thanks to everyone for your answers. I understand that price should not follow normal distribution patterns since it doesn't follow a random process. But when you look at charts like the one I posted, price seems to follow some kind of "rubber band" type of pattern, i.e. get back to the average with a relatively high % of success. I suppose Stddev is only one of the bands that can be used as trigger signals.

The method was originally thought for scalping 5 min charts where price moves sometimes pretty fast across the bands. If you wait for price to close inside the channel from outside to enter trades, you can miss a lot of pips because bands can be narrow at times. I would think the best option would be to set pending orders (buy limit on lower band for uptrend and sell limit on upper band for downtrend) with a 1Xatr stop loss and the other band for profit level. Using higher timeframes is a good idea as long as you get enough signals. It would give enough time to setup the orders at each bar, but I am sure this method could be easily automated.

 

[NetTecture]

Looks like a very interesting setup. I will try doing something with it the next days. Definitely one that may really CRY for automatism - one needs to be pretty fast sometimes, which otherwise makes it error prone.

As you said, running that on a couple of instruments may be fun

 

[TheBramble]

But when you look at charts like the one I posted, price seems to follow some kind of "rubber band" type of pattern.

The price isn't following a 'rubber band' type of pattern - the stddevs are a function OF the price, not the other way round.

The method was originally thought for scalping 5 min charts where price moves sometimes pretty fast across the bands. If you wait for price to close inside the channel from outside to enter trades, you can miss a lot of pips because bands can be narrow at times. I would think the best option would be to set pending orders (buy limit on lower band for uptrend and sell limit on upper band for downtrend) with a 1Xatr stop loss and the other band for profit level.

I think you're misunderstanding what scalping is all about - and probably a good thing too. Pending orders and stop losses aren't really part of the scalper’s arsenal.

Couple of other things.

What do you think it means when the ‘bands narrow’? It means volatility is decreasing or dead. You wouldn’t want to trade this situation anyway – scalper or trend trader.

And the whole point of waiting for the price to go outside the channel/bands and come back in is precisely to confirm trend. Again, not an issue for a scalper, but if you’re channel/trend trading – trade the trend. If you’re scalping, you just need to take a pip off the other side.

Be clear on what you’re actually setting out to trade. Be clear on your objectives.

 

[chamane]

Looks like a very interesting setup. I will try doing something with it the next days. Definitely one that may really CRY for automatism - one needs to be pretty fast sometimes, which otherwise makes it error prone.

As you said, running that on a couple of instruments may be fun

Thanks. There seems to be a nice win/loss ratio. Sometimes price doesn't reach the opposite band after crossing the middle line of BB. Maybe closing half the trade after crossing the line and the other half at the other band with breakeven at the line would be good. An EA would certainly help backtesting fast.

 

[chamane]

The price isn't following a 'rubber band' type of pattern - the stddevs are a function OF the price, not the other way round.

I think you're misunderstanding what scalping is all about - and probably a good thing too. Pending orders and stop losses aren't really part of the scalper’s arsenal.

Couple of other things.

What do you think it means when the ‘bands narrow’? It means volatility is decreasing or dead. You wouldn’t want to trade this situation anyway – scalper or trend trader.

And the whole point of waiting for the price to go outside the channel/bands and come back in is precisely to confirm trend. Again, not an issue for a scalper, but if you’re channel/trend trading – trade the trend. If you’re scalping, you just need to take a pip off the other side.

Be clear on what you’re actually setting out to trade. Be clear on your objectives.

My objective is to use a system I will be comfortable with and that will generate between 30 to 50 ips a day if this is possible. I have a small capital so I am looking for a good system on lower timeframes. This one looks interesting to me when I look at past data. But I am not sure I would be good at scalping because it's fast and can get to your nerves easily. I will probably look more at 15min or 30 min timeframes. Pending orders will be more appropriate then, but I will practice first before deciding to use them.