Risk:Reward Mystique

All

This follows on from money management mystique. That threads still there should anyone want to take up the gauntlet and prove the truth in the common trading 'truism':

"Good money management will turn a losing system into a winning one"

The next mystique is "Risk:Reward".

A very common use of this seems to be something such as:

- Price just bounced off support. Next Support level below that is 100 ticks away. Next Resistance level above is 500 ticks away, therefore risk:reward is 1:5.

Obviously, there are other examples but what the above says is that you are entering a smart trade because your potential reward is 5x your potential loss.

There are a few issues with this, most are related to probability.

Let's first turn this upside down and consider an risk:reward ratio of 5:2. That's 5 points stop loss & 2 points target. On a totally random 'coin toss' entry, out of every 7 trades you could reasonably expect 5 to be winners and 2 to be losers just because of the fact your targets are much closer to your starting point than your stops. This gives you a 71% win rate. Fantastic, eh ? It also gives you a losing system when considering slippage and fees. In this example, if the entry you are using has an 'edge', then it would need to return more than 71% winners. Agreed ?

Now let's go back to the 1:5 risk reward ratio based on Support & Resistance. Again, we can say that over time, with no edge, such a system will produce winners 1 out of every 6 trades OR 16% of the time. So, if the entry has an edge, you need to show more than 16% winners.

Now though, comes the tricky part (considering S&R based entry/exits).

On a 100 pip stop, 500 pip target, would you allow the price to move in your favour 480 pips and then watch it retreat all the way past your entry point to -100 to see yourself get stopped out OR would you secure some profits on the way up, trail the stop loss or just get out when the damn thing runs out of steam ? Also, if the price moves a few hundred pips in your favour, would you leave the stop loss at -100 ?

If you get out anywhere before 500 pips, as would seem sensible in the above case, you are no longer dealing with 1:5 risk:reward. Also - if you adjust your stops, you are also no longer dealing with 1:5 risk:reward.

If you do have decent exit rules to lock in profits, you are also subtly playing with the probabilities of success and rightly so. Over time, you could look at your win/loss ratio for the way you trade and assess probability but initial target is not your potential reward in your trades.

So - I think the this "risk:reward" mantra is just like the money management mantra. It's something that people listen to, digest and repeat because it sounds logical but it is something they don't look into too deeply.

Now - I would admit that if you had some sort of rigid scalping system where there is no intra-trade management, then risk:reward PLUS probability would be an interesting factor. Still though - without probability, risk:reward is meaningless.

So - can someone show us WHY risk:reward is repeated so many times by traders without it being used within the context of probability of hitting the target before the stop and without it being used within the context of the adjustments made to stops and targets once in the trade ?

Can someone show how risk:reward has any relevance when used on its own ?

In my opinion, only initial risk can be defined. This risk:reward ratio on it's own is not just meaningless but dangerous if it is being used as a measure of the 'smartness' of taking a trade.

Your point about R/R not being used alone is a good one, just like Win % cannot be viewed in isolation. I would make two points about your arguments, though.

The reward side of the equation is (or should be, anyway) a measure of likely potential when looking at a trade to be entered, not some rigid "only there" sort of thing. And even there it shouldn't be applied in all trade methodology cases (trend following being a very good example) except in a hindsight sort of way when evaluating performance. I actually think that's where many traders get tripped up. They try to apply what's supposed to be an overall performance metric on an individual trade basis.

As for R/R having anything to do with trade probability, it really doesn't. You suggest that 5:2 implies a 71% win rate simply based on the locations of the stop and target, but in doing so you make the assumption that price action is random and normal, which isn't the case. Is a 100 pt stop more likely to be hit than a 200 pt target? Probably so, but by how much? Trades are meant to be entered when the probabilities are skewed such that the odds of success and the R/R taken together produce a positive expected result.

To me, R:R has, always, been a mystery in as much as I can never guarantee that I am, even, going to get 1:1.

I figure out whether the thing is going up and how much I want to lose, if it doesn't.

Money management is very important, that R:R ratio has to be taken seriously, but whoever said
that "with good money management a losing system can be turned into a winning one" was talking nonsense.

Yet, if a newbie comes onto "First Steps" wanting a tip, as sure as apples, that old saw will be trotted out by someone. I may be guilty of saying many things but I have to believe it before I say it.

its true actual market movement has nothing to do with R:R or any TP/SL

Exactly! Risk/reward is meaningless unless you're looking at the probabilities of the trade succeeding or failing.

I enter a trade not based on it's risk/reward ratio, but rather the probability of it succeeding, which of course is a discretionary judgement.

If you're just playing with risk/reward you are also playing with your win ratio, and all you're doing is deciding in what manner you want to lose your money. A gain in one will produce a reduction in the other!

Rhody Trader said:

As for R/R having anything to do with trade probability, it really doesn't. You suggest that 5:2 implies a 71% win rate simply based on the locations of the stop and target, but in doing so you make the assumption that price action is random and normal, which isn't the case. Is a 100 pt stop more likely to be hit than a 200 pt target? Probably so, but by how much? Trades are meant to be entered when the probabilities are skewed such that the odds of success and the R/R taken together produce a positive expected result.

Click to expand...

There are 2 possibilities here - either I didn't explain myself OR you are dead wrong :whistling

On a RANDOM entry = with no edge, with a 'coin toss' approach. Your %age of wins will be DIRECTLY in line with the ratio of stop loss to profit target.

Let's say the system is entering a trade based on a coin toss heads for short & tails for long at 10:30am eastern time trading the e-mini. Stop loss is 5 points & target is 2 points.'

Run this over the past 10 years. You absolutely will see that 71% of the time you hit your target before your stop loss.

The reason I know this is because I was working with people to develop an automated system, one of the team got excited because of a consistent hit rate. When I analysed it, I saw that the hit rate came about not because of any edge but because of the 'risk:reward' ratio. Not only did I conclude this but I then coded a test 'random entry' system to prove it, ran it in and got consistent results in all markets I tested in all timeframes and over many different periods.

R:R provides a baseline probability of success. In this case 5:2 provides a baseline probability of 71%. This win rate is break-even before fees & slippage. To use a 5:2 risk:reward AND make money, you need to exceed that baseline of 71%. The results over and above the 71% win-rate would represent your edge.

It's just maths.

virtuos0 said:

I enter a trade not based on it's risk/reward ratio, but rather the probability of it succeeding, which of course is a discretionary judgement.

Click to expand...

This is it mate. I want to learn the why's and then try to establish where price should go if nothing extraordinary happens.

pedro01 said:

On a 100 pip stop, 500 pip target, would you allow the price to move in your favour 480 pips and then watch it retreat all the way past your entry point to -100 to see yourself get stopped out OR would you secure some profits on the way up, trail the stop loss or just get out when the damn thing runs out of steam ?

Click to expand...

I think risk: reward is a dynamic property. At the point it reaches 480, your risk is now 480 and your potential reward if it reaches your original target is only 20 (i.e. 20 more pips), i.e. a new risk:reward ratio of 24:1. If your stop were still at -100 then the risk is even worse 580.

Never thought of it that way mont.
But... how would you gauge probability of initial target being hit as you approach? Failed reversal signals? Momentum?
Imagine you have a hard initial target at 500 and it smashes through and you miss another 150 points.

pedro01 said:

On a RANDOM entry = with no edge, with a 'coin toss' approach. Your %age of wins will be DIRECTLY in line with the ratio of stop loss to profit target.

Click to expand...

Sorry, didn't see the "coin toss bit" first time through. Was kind of skimming.

But that tends to affirm my point. Trading is not about entering randomly (despite the behavior of many newbies! :cheesy: ), but rather entering when the probability of a successful outcome, combined with the R/R (and basic terms) produces a positive expectancy.

Or you could flip that around an suggested that your exits are that which determine the performance of your trading.

Rhody

You might want to read the 1st post again.... We basically agree. R:R is meaningless on it's own & even with win rate, it is meaningless unless you rigidly stick to your original targets.

R:R is bandied around like it's a badge of trading wisdom. You don't really know where the price is going when you enter. What we should write is R:?

The point about the ratio of stop loss:target is just to show that the ratio DOES have a bearing on your outcome. If you have any fixed stop:target systems - try doubling the stop & see what happens to win rate and profit. It effects any system in the same way because it is an immutable law of probability.

Of course people want to have an edge when they enter but that does not change the fact that manipulating the distance to stop loss or target will impact your win rate but not necessarily your profits.

Cheers

Pete

virtuos0 said:

If you're just playing with risk/reward you are also playing with your win ratio, and all you're doing is deciding in what manner you want to lose your money. A gain in one will produce a reduction in the other!

Click to expand...

What he said !

pedro01 said:

Rhody



The point about the ratio of stop loss:target is just to show that the ratio DOES have a bearing on your outcome. If you have any fixed stop:target systems - try doubling the stop & see what happens to win rate and profit. It effects any system in the same way because it is an immutable law of probability.

Click to expand...

Surely the point in placing a stop loss is to put it in an area where price is unlikely to breach an area of supply/demand given the circumstances and reasons for taking the trade in the first place and to simply double your stop would not only break your own trading rules but render the stop in an area where price in all probability will move straight through creating larger losses.

pedro01 said:

..................You don't really know where the price is going when you enter............

Click to expand...

pete

Yes and that makes TA itself something of a myth . What your TA does do though is give you a "reason to enter" and then it's down to how you manage what you've got. As you say, the only reasonable certainty you've got at that stage is your risk.

In the same way that your TA has given you a "reason to enter" (presumably based on some tested expectation) so your TA based target gives you a "reason to exit" (also presumably based on some tested expectation) if it arrives at that level. As you also say, there can't be any certainty that price will reach the target.

Nonetheless, if you're gonna use TA, the the fact that you've got certainty (risk) on the one side and uncertainty (reaching target) on the other doesn't invalidate using the ratio as a guide to whether you will take the trade or not.

' Course you might not use TA for your trade management at all - maybe just a risk 5, take 10 approach, but that's a different story albeit that the certainty/uncertainty aspect still applies.

good trading

jon

"Yes and that makes TA itself something of a myth"

that makes everything you think you see 20 pts from target a bit of a myth to

agree it gives a reason to enter - excuse to trigger - illusion of control

if your trade does not come with a virtuosO plan for the occasion your in trouble in the long run imvho

http://www.trade2win.com/boards/gen...6-hunting-secrets-revealed-19.html#post919908

"only reasonable certainty you've got at that stage is your risk"

if trading via sb firms

no real reason you should not make downside risk absolute imvho via guarenteed stop if a higher tf trader (week /day)

I for one enjoy my .........

illusion of control
Three little words that should cause a lot of people a lot of thought

delusions of grandeur.

three more!

"Make haste slowly"

good advice .............................. that it ?

poborsky said:

delusions of grandeur.

three more!

Click to expand...

I always wanted a stalker!

Is that some type of hat?