Study used to decide Stop Loss placement

[No Quarter]

Hi all

So, we all want to be able to place the SL at the right distance from where we enter the trade. I have with some success used ATRx2. But the ATR measures in both directions so to speak, so in a heavily upward trending market like we had yesterday March 1st, the ATR reflects the length of both the bars going up and down. And if I'm Long, the length of the bars going up don't worry me, it's the pullbacks down that I need to have a study that shows me the historic length of.

So, does an ATR exist, that is split into up and down going bars? What would that study be called?

I'm on ETX, and halfway found a solution, using the "Change" study, and then only looking at the Change bars in the opposite direction of my trade. But... Change only deals with Open and Close, not High and Low, so it will underestimate an environment with long tails.

I hope I explained well, if not let me know. And any suggestions are welcome.

 

[tomorton]

That's a good thought No Quarter, it does seem that the ranges of up bars and down bars in an uptrend and a downtrend might be different and we could benefit from more information.

That said, Entry plus/minus ATRx2 gives a random price level as far as the TA of the market is concerned. I mean that the market sees and takes (more or less) account of highs, lows, tops, bottoms, MA's, ranges, trends and trendlines, but it doesn't even see your entry point.

So a stop that's not related to TA is at a random price. So it will be hit on a random basis. And how can that be the best strategy?

 

[barjon]

That's a good thought No Quarter, it does seem that the ranges of up bars and down bars in an uptrend and a downtrend might be different and we could benefit from more information.

That said, Entry plus/minus ATRx2 gives a random price level as far as the TA of the market is concerned. I mean that the market sees and takes (more or less) account of highs, lows, tops, bottoms, MA's, ranges, trends and trendlines, but it doesn't even see your entry point.

So a stop that's not related to TA is at a random price. So it will be hit on a random basis. And how can that be the best strategy?

Aye, random if taken from your entry point. In a strong trend measuring the pullbacks/ retracements from when the trend started to where you entered (assuming you hadn't hit on or near the start) and setting a stop taking account of the worst seen (e.g.: 1.5 x the biggest pullback/retracement) usually gives you a tighter stop than ATRs.

 

[tomorton]

Aye, random if taken from your entry point. In a strong trend measuring the pullbacks/ retracements from when the trend started to where you entered (assuming you hadn't hit on or near the start) and setting a stop taking account of the worst seen (e.g.: 1.5 x the biggest pullback/retracement) usually gives you a tighter stop than ATRs.

Interesting variation, and maybe better than (Entry - ATRx2) but isn't it still random, just using an arguable but arbitrary constant number to calculate exit price, rather than "reading " it off the price chart directly?

 

[barjon]

Interesting variation, and maybe better than (Entry - ATRx2) but isn't it still random, just using an arguable but arbitrary constant number to calculate exit price, rather than "reading " it off the price chart directly?

No, it's "technical" by reference to the greatest degree of pullback so far (and dynamic as the trend progresses).

 

[tomorton]

No, it's "technical" by reference to the greatest degree of pullback so far (and dynamic as the trend progresses).

Granted yes its drawn from chart print, but how can you justify 1.5x, except as a rule of thumb. And why would it work in every trend this year and forever on even the same market, let alone different classes of markets like stocks to forex to metals? If you have to keep tweaking your constant multiplier, based on no hard evidence other than it stops working, that's randomness in there isn't it?

I'm taking a hard line on this stuff these days. Part of the ageing process I suppose.

 

[barjon]

Granted yes its drawn from chart print, but how can you justify 1.5x, except as a rule of thumb. And why would it work in every trend this year and forever on even the same market, let alone different classes of markets like stocks to forex to metals? If you have to keep tweaking your constant multiplier, based on no hard evidence other than it stops working, that's randomness in there isn't it?

I'm taking a hard line on this stuff these days. Part of the ageing process I suppose.

Can't , 1.5 just a suggestion. You might go for just a tick beyond the biggest pullback move, 1.5 x, or whatever takes your fancy. The point is that you are using a stop based on the pullback move - thus not random?

 

[tomorton]

Can't , 1.5 just a suggestion. You might go for just a tick beyond the biggest pullback move, 1.5 x, or whatever takes your fancy. The point is that you are using a stop based on the pullback move - thus not random?

Well, a calculation based on the pull-back move might be done in a standard way with a repeatable formula, but the next pull-back will not be the same: and all the pull-backs in say BARC will be different from all those in AUD/USD.

 

[barjon]

Well, a calculation based on the pull-back move might be done in a standard way with a repeatable formula, but the next pull-back will not be the same: and all the pull-backs in say BARC will be different from all those in AUD/USD.

No, but you are looking at the pullbacks that have occurred in the same instrument during the trend and using the worst case scenario one. There's nothing to say the next pullback will not be even greater - in which case you'd be stopped and have a new value to use as a stop if you got into the trend again.