Risk / Reward

[TheBramble]

as long as I risk 2% of my account on any trade and ensure that my risk/reward is 2:1 or greater I should be OK in the long run ?...

How will you ‘ensure’ your R:R is greater than 2:1? Is it already? It’s not like you pick these things off a shelf, a 2:1 system a 1.5:1 system a really expensive 5:1 system. Whatever your system has been designed to do to trim that edge off of some other poor buggers profits it what ever it is. It’s an average, a ratio that you will retrospectively look at and point a finger at and say – that’s it. But you wont know from one trade to the next which is going to give you a nice 5:1 and which is going to take your stop. The sequence of strings of consecutive winning trades and losing trades make a mockery of any attempt to ‘design’ performance ahead of placing the trades.

Providing your trading system has generated useful performance during testing and providing the transition into live operation carries no additional or unforeseen risks (LOL), you can probably hope to make around the same level of performance, which will swing relatively wildly around from trade to trade, but will, over time, quite obviously, provide a quite artificial and useless measure of how you ‘did’ or how you ‘do’ on any one trade.

Certainly stick to a small %-age risk per trade (be it 1% or 2%), but don’t get hung up on getting a designer fit of a specific R:R – it’s playing a pointless numbers game and is often the sign of aggravated delay in launching into the reality of trading itself. If you don’t have a string of data from demo trading which indicates you’ve got an edge – you probably don’t have an edge.

I mean I'd have to be extremely unlucky to blow my account and have a whole string of losers if I stick to this.

No. You’re likely to blow your account even with this sort of performance. Definitely if it doesn’t have this sort of performance, but likely even if you do.

I guess you just have to set the stop / limt and walk away

No. Managing the trade is as important as your initial analysis and assessment. Part of your trading methodology will (should?) indicate under what criteria stops and limits are adjusted. And of course, you wouldn’t adjust them or pull the trade unless those specific criteria were extant.

 

[Technically Fundamental]

Bramble, you think r:r is a decent tool if used to help identify whether a trade is worth taking after your analysis throws out entries and exits or do you think higher prob should be given more consideration?

 

[candles]

you've come into your own recently with some great posts TheBramble......its just a shame that the plane takes off.....
good stuff.

 

[zupcon]

So what do you use to calculate probability?

I'd recommend reading Evidence Based Technical Analysis by David Aronson. The only problem is you might actually believe the authors conclusions that markets are random, and therefore TA and price action are a waste of time, and that would be a pity becaue the book does contain some useful techniques.

 

[Technically Fundamental]

I'd recommend reading Evidence Based Technical Analysis by David Aronson. The only problem is you might actually believe the authors conclusions that markets are random, and therefore TA and price action are a waste of time, and that would be a pity becaue the book does contain some useful techniques.

Completely random? That's impossible.

 

[blancspa]

is it normally the case that price tests S/R, before breaking ? or is this random?

 

[TheBramble]

you've come into your own recently with some great posts TheBramble......

So all the others over the last six years were sh!t were they?

its just a shame that the plane takes off.....

Actually, I changed my mind about that. Not about the plane taking off - it can't, but about the people still waiting for it to do so. They're OK really - just a bit fuzzy...

 

[zupcon]

Completely random? That's impossible.

Maybe I'm doing the author a bit of an injustice and putting words into his mouth, as its been a while since I read the book. He applies a bunch of statistical analysis techniques, on a bunch of rather dumb hypothesis and gets nowhere fast

But Its still contains some useful ideas for people interested in quantatative analysis of markets, its one of the few books I'd recommend.

 

[poborsky]

Completely random? That's impossible.

Hope so or ive been trading completely random trends off non existent s/r in the belief that the trend was my friend!

 

[Technically Fundamental]

Might give it a go. I've always assumed that the market was comprised of too many variables that in turn are reliant on human action to be properly understood.

 

[Technically Fundamental]

Hope so or ive been trading completely random trends off non existent s/r in the belief that the trend was my friend!

Cant be random. If Iran exploded the oil price would sky rocket before price stabilised at a new supply/demand once the problem was resolved. What's random about that?

 

[RedGreenBen]

Cant be random.

I agree BUT try creating a random time series in Excel (price at t+1= price at t +/- a random amount) and plot it.

You will be amazed (or perhaps I am easily impressed) by the S/R, trends etc that you will see.

Ben

 

[poborsky]

Cant be random. If Iran exploded the oil price would sky rocket before price stabilised at a new supply/demand once the problem was resolved.

Sure, but as a trader just follow the price action.

 

[Calinor]

Maybe I'm doing the author a bit of an injustice and putting words into his mouth, as its been a while since I read the book. He applies a bunch of statistical analysis techniques, on a bunch of rather dumb hypothesis and gets nowhere fast

Yeah from what I remember he applied stochastics and a few other indicators on daily charts (if these indicators lag a lot, then on a daily chart do they lag a whole day?), applied it to different assets and found some strategies for stochastics made a profit, some made losses, but none of the results was statistically signifcant for any of the methods of using stochastics. It wasn't the most sophisticated test, no support and resistance, no combinations of indicators and I think he admits as much by saying it is just to start a debate on technical analysis and not really a conclusion.

I think he did mention some other papers in which it was shown that statistically significant profits could be made on patterns like head and shoulders. But these things are not easy to test, since a head and shoulders pattern is subjective.

 

[Calinor]

I agree BUT try creating a random time series in Excel (price at t+1= price at t +/- a random amount) and plot it.

You will be amazed (or perhaps I am easily impressed) by the S/R, trends etc that you will see.

Yeah just for my own interest I once just took a chart and drew some lines fairly evenly spaced out (not exact, just eyeballed it), 100 points or so apart without caring about turning points or any of that. Then when I zoomed in I was amazed at how many times price just bounced off these lines as if they were support or resistance.

 

[MrGecko]

Fokker-Planck?

 

[RedGreenBen]

Yeah just for my own interest I once just took a chart and drew some lines fairly evenly spaced out (not exact, just eyeballed it), 100 points or so apart without caring about turning points or any of that. Then when I zoomed in I was amazed at how many times price just bounced off these lines as if they were support or resistance.

I think they are called Fibonacci levels

 

[blancspa]

I am honstly very confused here.
Can I bulid a system without knowing, if TA, fib, S/R, etc work?

 

[simcom]

A useful resource for calculating the probability of a losing streak over a particular number of trades:

http://pages.stern.nyu.edu/~churvich/Undergrad/Handouts4/13-Streak.pdf

 

[intradaybill]

following on from my last post "positive attitude" I've been trying to simplify my thought process right down....

as long as I risk 2% of my account on any trade and ensure that my risk/reward is 2:1 or greater I should be OK in the long run ?...

I mean I'd have to be extremely unlucky to blow my account and have a whole string of losers if I stick to this.

I guess you just have to set the stop / limt and walk away

It depends on whether the 2% is on current bankroll or initial deposits. The former is an anti-martingale method that theoretically never leads to a blow up but practically limits your ability to trade at some point. The latter carries a probability of ruin that is very small for 2% but it is there. Worse case scenario is of course is 50 consecutive losers right from start. Very low probability but non-zero. Another scenario is 20 losers followed by 10 winners and then by 40 losers, and so on. The probability depends on the success rate of your strategy and R:R. I say most blow up even with 2%.