Market Prediction/Risk:Reward

[turtle trader]

The (in)ability to predict where a market is headed, and ensuring our trades have a suitable Risk:Reward ratio are 2 very well discussed topics on this board.

The consensus regarding market prediction seems to be that it is not possible to predict which way or by how far a market will travel, so we should not try to do so - just follow what the market is doing. Even more widely accepted is the principle of risk:reward ratios - the potential reward any 1 trade offers should be x times greater than the maximum loss we are going to potentially take on the trade.

I'm pretty comfortable with each of these ideas on their own, but am struggling to tie the 2 together:

It's easy to control the risk side of a trade, but if we really aren't forming an opinion on how far a market might go in our favour, how do we know what the potential reward is, and therefore how do we know that our R:R is suitable? When people say you can't predict where a market is going do they really mean 'I don't know which direction it'll go, but if it goes up then it'll go up by x points, so that's my potential reward'?

How do others view/manage these 2 seemingly contradictory principles - am I missing something fundamental here?

Turtle

 

[Finlayson]

turtle

the only way I can think of it is have it relative to the market u r trading....for example I trade the DAX & for it's typical moves & retraces I find 10:30 a good stop & target or 10:50 if u r feeling clever & the market has been tight.

open to suggestions if anybody thinks this is wrong

Jay

 

[Trader333]

In my view all you can do is play the percentages. Because I trade Nasdaq stocks and the ones that I trade have a good intra-day range (sometimes several hundred points), then when I enter with an 8 point risk and I get it wrong I lose 8 points. But if the market proves me to be correct there is a good chance of making more than 3 times this and in many cases 10, 20 and I have even had 25 times that.

This is why I focus so heavily on reduced risk at entry but the trade may well only give me 5 points, 2 points or break even and I just dont know on an individual trade basis. All I know is that I can stack the odds as best as I can in my favour.



Paul

 

[mmillar]

...

The consensus regarding market prediction seems to be that it is not possible to predict which way or by how far a market will travel...

'By how far' is the bit you have wrong. We don't know which way a market will go until it starts moving. However, once it starts moving you can have a pretty good idea how far it will go (using various chart patterns, s/r levels etc). If that weren't true then you would be right about the two principles being contradictory.

 

[oatman]

....which comes back to my favourite - momentum
It's still the most important factor for me.
But how to judge it? :cheesy:

 

[sidinuk]

The problem with risk:reward is that most people don't take into account the probability of the risk or the reward happening. They assume they have equal probability and therefore a Risk:Reward of 1:3 is better than 1:2. This is just not right.

Say your trading the dow. You set a stop of 50 pts and target of 50pts, that's only 1:1 risk:reward. So why not set the stop at 25pts instead. Now you have a R:R of 1:2, great. Unfortunately the stop is twice as close and you are, therefore, more likely to be stopped out. So in reality the 1:2 is probably no better than the 1:1.

So how do you incorporate the probabilities into the Risk:Reward calculation? The only way is through extensive backtesting of the particular setup that you are looking at. This is very difficult because you have to be very specific about what you are looking at, not a woolly oh double tops have a 75% chance of reaching blah target - how do you define a double top, in which timeframe, over how long, what about the patterns that looked like double tops at the time but broke the wrong way and you've ignored them because "you wouldn't have traded it anyway".

 

[jsd]

well its known that if demand outstrips supply then prices go up and conversley supply outstrips demand, they go down, how best is it that we can identify theses points in the market at the same time knowing people often want to con you one way and then prices tank off the other, i think volume ,heavy volume at s/r levels (SPIKE BARS)also gives weight to the formation of potential tops bottoms at that point in time...

JD

 

[oatman]

sid, This all depends on your method.
For instance this morning, just after the opg on Ftse Fut., there was a low of 24.5 a double top at 31 and a gap at 48.
So you had a "target" to fill the gap on a break of the high(31), with a stop at new lows.
Unfortunately, a bit early for me. I like to see it settle first. It actually managed 41, but that's neither here nor there. It was an opportunity(missed)!

 

[zow]

Hi TT

Personally I have never been very comfortable with the 'plain Risk:Reward' concept because it misses the whole 'probability of outcome' aspect. SIDINUK has mentioned it above.
Say you do identify a series of trades with a R:R of 1:3, but what use is that if the positive outcome ratio/probability is only 20% or 1:5 ? In the long run you will have the odd largish winner (of 3) but that will be more than off-set by all the small losses (of 1 each) minus all the costs of transaction of course.

So similarly, I think there is nothing wrong with identifying a series of trades with a R:R of 1:1 (or even worse) as long as you are confident that a positive outcome will occur, say, 75% of the time. Nor with a R:R of 1:100 even if it only works 10% of the time.

In practice, when I trade I only try to identify occasions when I am fairly sure the market is going up or down (on whatever time scale). When I enter the market I do not have a pre-defined profit target. If it goes against me I try to get out quickly and immediately. I will though use the charts/studies/tools etc to identify support/resistance/congestion areas and with that information together with a reading of how I think the market is acting I will get out when I think that on balance the probability of me continuing to be correct has gone against me. Here is also where scaling in/out is indispensible: will the market go through that line of resistance? Don't know, but sell some of your position there, lock in some profit, and see if it does go through. If it does then great, if not you still have banked some points.

This whole process may take only a few seconds to several hours/days later.

hope that makes sense.
z

 

[turtle trader]

mmillar

That is the way I try to trade at the moment - once it's started to head in a particular direction i'll look for s/r levels etc to see where I think it might stop & calculate R:R from there.

My trouble may well be that I can also appreciate what T333 says - take care of the risk & let the reward take care of itself (sorry if that is an over simplification) and as Finlayson says - find something that works with your market & stick with it. What I need to do is formulate a fixed strategy rather than trying to apply 1 or more of the above to the trade i'm considering at the time. I'm hoping this discussion might help me, and perhaps others, do that.

Turtle

 

[turtle trader]

Sidinuk & Zow

Agree absolutely that percentage of winners is vital part of this equation that I have not yet included here. What I would ideally like to get to is a 'hard & fast' rule that I can stick to, so Sidinuks thorough backtesting (although it involves so much more work!) is probably more suitable to me at this time than Zows market 'reading', based on all his experience, but that's just one of those personal things that we all do differently.

As you also say, Zow, I have scaled out of positions at a point where a target has been met, but when I suspect there is further to go. This has worked well for me in the past, but I would like to quantify it's potential results, as far as possible

Turtle

 

[stevet]

evaluating risk reward requires 3 assumptions and 3 uses of the "think" word

one on market direction, and two on exits points

one assumption is dangerous, basing a trade on three is armegedon

 

[Finlayson]

Turtle

R:R is important I also think it is important to lock in profit on the way.....as I have a new scanner I just figured out how to do this, here is my first ever real trading month's record's.....I want to stress I had 'no-idea ' about the trades & all results r down to how the trades were managed.........I was guessing basically

 

[oatman]

This is the difficulty. You form a strategy but can you carry it through under pressure? This will depend on your trading time frame. If you're day trading, watching the ticks, you've got to think quick and that's where you make mistakes. If you're trading a longer time frame, then it's easier to manage the pos.

 

[Finlayson]

just to show with the right direction a total 'numpty' can do it

Jay

 

[turtle trader]

Stevet - I was worried you'd get involved!! ( Only kidding)

Finlayson - fantastic, but if you start showing records like those you're going to get mma crawling all over here making accusations of falsifying evidence!

Turtle

 

[Finlayson]

Turtle

yes I know.....but who gives t*ss really......just showing them for the new Guy who wants a bit of belief

& I have a new scanner

jay

 

[zow]

What I would ideally like to get to is a 'hard & fast' rule that I can stick to........

As you also say, Zow, I have scaled out of positions at a point where a target has been met, but when I suspect there is further to go. This has worked well for me in the past, but I would like to quantify it's potential results, as far as possible

Turtle

IMHO I don't think it is possible to create decent 'hard & fast' rules because the markets are constantly changing, second by second, day by day, year by year etc. Ok, there are companies e.g. MAN, that trade alot of money systemmaticaly/computer trade, but they also have a room full of programmers/systems managers/money managers who are constantly tweaking and modifying those systems to adapt to the changing world.
Maybe you can consider using Neural networks and 'learning computers' (of which I know zilch about) but I think you quite quickly get to the point where instead of trying to create a 'silicon trading brain' it's easier (and more fulfilling) to learn and adapt your own brain to trade well.

Trying to 'quantify potential results' is what I, (and I think several others above) try exactly NOT to do.....

 

[oatman]

Indeed, the "easy" part is getting "in" :cheesy:

 

[stevet]

the hard and fast rule is to enter trades at the right time and then hold that trade until the elements that got you into the trade break down

over time this will produce your risk reward basis for trading - which is not about setting fixed out points - but about having a profitable p&l

that is, you have a methodology and over time contiuous use of that methodology will cause losers and winners and over time you will either be losing or making money and the losses to wins give you the risk reward basis of your trading methodology

if your trading is based on setting fixed out points - that is not the risk reward of your trading - that is part of the methodology of your trading and if that produces a positive P&L over a long time frame - then that is a methodology with a postitve risk reward basis and worth continuing

risk reward is an actual - not a theoretical value