How to handle, with SL and TP, from 2000 to 2009? Very Different Volatilities

Hi, I'm back testing some systems, and have come to a little problem, markets range are very different than the last 3 years, on years past 2005, the markets did not had 80% of the movements that we have today.

So I'm testing a system with SL and TP values from the last 3 years, wich are rarely hitten on years past 2005, on the other words, I should have a smaler SL and TP, but how do I Get those values? I could get the values by optimizing the SL by that year data.
But thats worth nothing, because I must have those values before I do the trades, and not after.

How do you people handle with it?

Does anyone use ATR (average true range) for that?
Parabolic SAR? or any other indicator that could help determine those values?

You could always use a set of indicators not to serve as a stoploss or takeprofit, but instead to serve as a signal indicating a potential move against the trade.

That's a Very interesting perspective.

I used to use Indicators as a stop loss level but that was when my son was 2 years old. His son is now almost 1 year old.

The thing about (most) indicators {as we all have heard over & over & over again} is that they move Following the Price Action (PA). This basically translates to a Floating Stop and will only be hit 'x' periods of Bars after the Optimum Exit {& that means Good Or Bad}

There IS a Reason that most of the Old Fogies around here discard that "Measure the Last 'x' Periods & report back to me where Price is now in relation to that Old data." concept.

The Reason is; It May not mean squat tommorow. Never mind tommorow, It may not mean squat Now.

I know you have seen it, perhaps even read some stuff on it but the Fact Remains... You have to learn to Read The Market NOW!

The fact that not one {major player} was willing to sell below x.xxxx should be a clue.{this would be exhibited by numerous Lows at ~ the same Price} Or Buy above x.xxxx... {figure it out...}
These incedents create Ceilings or Floors. Why would you fight them?
Between the 'Ceiling & the Floor' there may be many points where the {for the fight fans} descision 'Goes to the Cards'

That's when it is "Game On" That's when We get to Play. Get it wrong & you're toast. Get it right & who knows how far you can go.

Keep it real. Humans in the UFC Pit or Humans in the Trade Pit. Same size, same shape, just a different colour.

Apply a sport you like, if you like, if it will make it easier to understand.

BMc(spear)
btw: B stands for Bill Which Stands for William which stands for... err... shakespear...
can't be. he got lai@ a bunch o' times.

I was thinking along the lines of donchian channels but at works too

From a system perspective, theres a few ways of predicting volatility - ATR, GARCH models, option pricing...

GARCH models are hideously complicated, and I could never help feeling they are overengineered for a problem that can never have a perfect answer. Implied option volatilities look like they should work, but in the real world I couldn't find any solid evidence for significant correlations.

And so for me - it's ATR every time. Keep it simple.

ATR is a very good way of determining initial sl.

Just experiment with: atr(X)*Y and when you find good [and by good I mean suitable for your method] X and Y, you will be fine with almost any conditions.

my goodness, really shows how long you've been in this game.lol