Odds of winning are constant: Using market bias AND stop loss | Page 3 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Ea for take profit and stop movement based off risk(stop loss) 13 replies

Stop Loss vs. Hidden Stop Loss 26 replies

Buy Stop and Sell Stop orders with OCO and Trailing Stop 0 replies

Stop Loss and Trailing Stop 0 replies

Movements of Forex, and using a stop loss 45 replies

Attachments: Odds of winning are constant: Using market bias AND stop loss
Exit Attachments

Odds of winning are constant: Using market bias AND stop loss

  • #41
  • Dec 16, 2009 2:38am Dec 16, 2009 2:38am
Quoting calderone
No matter how good your analysis, how good your signal and your market timing -the only difference between a winner and a loser is your psychology.

When fear is in control you are blind to the obvious.

I once, hours later asked myself why I didn't close and reverse instead of taking a loser.

Sorry to hijack the thread on emotions ruining good trading but the psychological aspect can be the deciding player. IMHO

You can't trade with scared money.
Good point. That's why I use an EA.
  • #42
  • Dec 16, 2009 8:18am Dec 16, 2009 8:18am
Quoting FX Compass
Triphop, I mentioned nothing of Stop losses in my post.

Here is the Link to the four part webinar, each is approximately 45 minutes long. Very informative for traders of any level.

http://www.fxstreet.com/forum/showthread.php?t=53230

Your welcome to view it and then take the statistics up with guest speaker. Don't shoot the messenger.
Compass, relax, no shooting any messengers - I just don't get how he's come up with the figures. Thanks for the links - I'll indulge when I have a few spare hours as his bio does look pretty good.
  • #43
  • Dec 16, 2009 10:41am Dec 16, 2009 10:41am
Quoting triphop
Compass, relax, no shooting any messengers - I just don't get how he's come up with the figures. Thanks for the links - I'll indulge when I have a few spare hours as his bio does look pretty good.
It's all good triphop.

Webinars very interesting at looking at the market through anothers eyes.
  • #44
  • Dec 16, 2009 3:21pm Dec 16, 2009 3:21pm
The reason 95% traders loose is because prices jump in either direction madly specially in e/u pair hitting SLs.

After hitting SLs, prices reverse back.

We are trying to make money out of this madness.
  • #45
  • Dec 16, 2009 3:45pm Dec 16, 2009 3:45pm
Quoting hanover

3. MM (I mean lot sizing, and method - fixed frac, optimal-f, progressive, etc..... as opposed to exit/trade management) has no bearing on expectancy.
This comment bothered me somewhat, especially after reading hanover's phenomenal 15 point post here:

http://www.forexfactory.com/showthre...t=23526&page=6

I think 3. in the comment above is true when you assume that there is no autocorrelation (trends or reversion) in the market and that there is no autocorrelation in your strategy.

If either assumption is broken, as hanover mentioned previously, cutting losses and riding winners would help your expected gain for a trending market/strategy (positive autocorrelation), and Martingale type strategies would help your expected gain for a reverting market/strategy (negative autocorrelation).

If you want to be more specific, there could be autocorrelation, but if the trading costs (bid-ask spreads, etc) are too high, then you can't realize a greater profit just through MM. We are now venturing into the borderline of EMH (efficient market hypothesis) and arbitrage theory.

Also note that if there is no insider trading and if we believe that fundamentals move the market, then perhaps forex rates should trend to the new equilibrium as news comes out, but it may happen so quickly that not many could profit from the move. At least that's what the EMH guys would say. You may counter and say that for the prices to move, institutions need to place orders into the market and since even the forex markets aren't infinitely liquid, we have a chance to "frontrun" them, and that's where the opportunity lies in capturing these trends.
  • #46
  • Dec 16, 2009 4:56pm Dec 16, 2009 4:56pm
Quoting kfmfe04
I think 3. in the comment above is true when you assume that there is no autocorrelation (trends or reversion) in the market and that there is no autocorrelation in your strategy.
#3 says "In a random market." Any market with correlations, whether automatic or not, would not classify as a random market.

Quoting kfmfe04
If either assumption is broken, as hanover mentioned previously, cutting losses and riding winners would help your expected gain for a trending market/strategy (positive autocorrelation), and Martingale type strategies would help your expected gain for a reverting market/strategy (negative autocorrelation).
If the assumption is broken, you aren't talking about a random market. A martingale strategy has no bearing on any of this, it's a progressive lot strategy. The size of a trade has no bearing on the outcome of that trade (baring margin call). There's a difference between profile outcome and trade outcome, as I addressed in my post above. You are talking about profile outcomes, which is a vastly different topic all-together.

Quote
If you want to be more specific, there could be autocorrelation, but if the trading costs (bid-ask spreads, etc) are too high, then you can't realize a greater profit just through MM. We are now venturing into the borderline of EMH (efficient market hypothesis) and arbitrage theory.

The entire crux of the argument made thus-far is that MM does not provide an edge. It neither increases nor decreases an edge, it is edge neutral. Edge as in the odds of a single trade being more or less successful. But, as I said earlier, MM does play an important role to an entire profile of trades, since without it you cannot realize a statistically significant number of trades for the edge to manifest.

Generally speaking, MM doesn't effect trading costs for your average bucket shop trader. Most of us pay a spread, not a fixed commission. Naturally, people in an ECN might find this to be a little different.

Quote
Also note that if there is no insider trading and if we believe that fundamentals move the market,

Insider trading in the forex market? You think the minutes of the FoMC meeting get leaked out to a single special party and nobody else gets wind of it?

Quote
At least that's what the EMH guys would say.

Tdion, et all. "There is no edge" and the like. I disagree. Stuff gets priced in over time, it doesn't happen instantly. That's where edges are found, ex: making prices move more efficiently.
  • #47
  • Dec 16, 2009 5:26pm Dec 16, 2009 5:26pm
My fault - I didn't see the "random market" part of that post.

If it's a "random walk", there's not much you can do, strategy or MM-wise, that I agree.

FWIW, I consider Central Bank intervention to be an example of forex insider trading.
  • #48
  • Dec 16, 2009 6:10pm Dec 16, 2009 6:10pm
Quoting kfmfe04
FWIW, I consider Central Bank intervention to be an example of forex insider trading.
Ah. That's an interesting view. I can see that, sure.
  • #49
  • Dec 16, 2009 7:42pm Dec 16, 2009 7:42pm
Quoting kfmfe04
This comment bothered me somewhat....
By 'expectancy', I meant the expected outcome of a trade, as explained by Van Tharp in Trade Your Way To Financial Freedom. I've also heard it called 'edge' or 'profit factor'.

Expectancy = (win% / loss%) x (net avg win size / net avg loss size)
  • #50
  • Dec 16, 2009 9:26pm Dec 16, 2009 9:26pm
All trades are essentially 50/50.There is a 50/50 chance the market will go up or down. The only thing that determines which 50% prevails as a win or a loss is where you set a target for either a profit or a stop-out.

The market movement is chaotic; there is no guarantee that it will go where you'd like it to go. This sets the stage for the 50/50 odds to occur. (Actually, the odds are worse because you have to pay the spread).

So setting, say, 10 pips as the profit target could result in a successful trade while setting a 40 pip profit target may result in an unsuccessful trade if the market doesn't reach the profit target before turning around and going the other way. And then there's the element of time and account size. Most trading accounts are limited in capacity so you'll either hit the profit target or perhaps the market will move the other way and you'll run out of money or the market will move against you and just bounce around in non-fatal but negative territory, not getting any better or worse for who knows how long. In the latter case, at some point you have to decide to stop the trade because you can only watch a trade for so long before it becomes a waste of time.

Setting a profit target and a stop-loss doesn't change the basic odds. If you set the profit target and stop loss both to 10 pips the odds are the same as if you set them both to 100 pips. The results may be different in each case, but the odds are still 50/50. Changing the ratio of the profit to stop loss doesn't change the odds either. If you set a profit target to 100 pips and the stop loss to 50 all you're doing is doubling the odds that you'll hit a stop loss. So for every 100 pips you make in profit, you'll likely hit two stop losses for 50 pips each. Same difference.

Though you may be able to observe a market bias--i.e. a trend--that can help improve the odds for successful trading, there's no way to know for sure if you're at the beginning, middle, or end of a trend. And the only way to prevent a stop loss is to not set one. Of course, by not setting a stop loss you're opening up the possibility that a margin call will become your stop loss.
  • #51
  • Dec 16, 2009 11:22pm Dec 16, 2009 11:22pm
Interested to know what everyones thoughts are on an equal stop loss / take profit position.

In other words let stay a reasonable stop loss is set at 2 X ATR which happens to be 50 pips. A take profit is also set at 50 pips, is it the general consensus that this is a 50/50 outcome ?

What if the trade was placed in a clear uptrend or downtrend ?

Would this increase your odds ?
  • #52
  • Dec 16, 2009 11:22pm Dec 16, 2009 11:22pm
Quote
All trades are essentially 50/50.There is a 50/50 chance the market will go up or down. The only thing that determines which 50% prevails as a win or a loss is where you set a target for either a profit or a stop-out.

I don't believe that at all. The market might go up, it might go down, it might go sideways. But it might go up, then down, then way up again. I don't think you can say that just because the next tick may be up, down or sideways, that the cumulative odds of the market going up or down is 50/50.

If all that determined your result was the profit or stoploss, then you'd be working in a totally random market. You could not profit from that, as your average would be negative. That's no better than a casino.

Quote
The market movement is chaotic; there is no guarantee that it will go where you'd like it to go. This sets the stage for the 50/50 odds to occur. (Actually, the odds are worse because you have to pay the spread).

Chaotic is not random. Nobody is talking about a guarantee, we're talking about odds. It is mathematically provable that is impossible to turn a profit over time without an edge. MM does not effect edge. I can also demonstrate that there are times when the odds of a market outcome is not 50/50. The trick is figuring out when those times are. I have my ways of doing this, as I'm sure others do as well. Breaking the TP/SL parity problem... that's the trick. Not all that hard tho, really. There's a thousand ways to do it, it all boils down to market sentiment.

Quote
So setting, say, 10 pips as the profit target could result in a successful trade while setting a 40 pip profit target may result in an unsuccessful trade if the market doesn't reach the profit target before turning around and going the other way.

Lets say you have a 10 pip target and a 40 pip SL. That means for every 1 loss, you must make 4 wins just to breakeven without spread. In a perfectly random market, that would be relatively easy. But in the actual markets, it's anything but.

The edge in this system, then, would be determining when the market was least likely to trend away from you. If you could sufficiently guage when the market was going to move in a tiny range then you could make this system profitable. For every profitable system there is an edge, whether it's obvious or not.

If you were to trade open w/o an SL, with a 10 pip target, it might be convenient to think you'd never have a problem. But all it takes is 1 bad trade to wipe you out. Naturally then if you're trading counter to a major trend, this could be a serious problem. Thus the edge there depends on accurately guessing the long range trend enough that your account is rarely destroyed. As always, MM is only a factor in determining when that will be.

Quote
Though you may be able to observe a market bias--i.e. a trend--that can help improve the odds for successful trading, there's no way to know for sure if you're at the beginning, middle, or end of a trend. And the only way to prevent a stop loss is to not set one. Of course, by not setting a stop loss you're opening up the possibility that a margin call will become your stop loss.

Generally, that's what we tell people. "You can't know for sure if..." but truth is, you can. I do it all the time. You can easily tell when a trend is slowing down, it gives off massive signals when it does.
  • #53
  • Dec 16, 2009 11:31pm Dec 16, 2009 11:31pm
Quoting FX Compass
In other words let stay a reasonable stop loss is set at 2 X ATR which happens to be 50 pips. A take profit is also set at 50 pips, is it the general consensus that this is a 50/50 outcome ?
No, it most certainly is not a 50/50 outcome in the actual markets. I've done that exact simulation, it is not a 50/50 outcome. I can prove it. If it was 50/50 then the loss on the account would average out as the spread per trade. In reality, the loss is much much worse.

In a random market, yes it would be. In the actual markets, it is definitely not.

The thing to realize is sentiment. In the EURUSD, for instance, you have dollar policy and euro policy.

The ECB likes strong currencies, and dislikes playing tricks with monetary policy even if it's to stimulate a particular economy. This is not unexpected, the euro supports more than one nation and those nations produce a diverse set of exports, thus it's hard to reach a favorable consensus.

The dollar on the other hand is a single nation currency. We don't export a lot, we import most of our goods. But imports don't provide high salary jobs (low level support and retail jobs, at best). Reducing the strength of the currency makes our exports more profitable, and makes foreign investments look better on US balance sheets. It also forces ppl w/ established wealth to invest it, or risk losing it due to inflation. Thus a weak dollar is preferred by current monetary policy.

The underlying policies, then, favor a strong EUR versus the USD. Small blips along the way have messed with this, but it's still the big sentiment. Blips, such has economic collapse or large interest rates, are easy to factor in when they occur. This sentiment plays out most obviously on the weekly, but works it's way down to all the lower TFs.

It does not manifest in a straight bar ratio, however. The odds of the next bar being bullish or bearish is about 50/50, regardless. But the bullish bars (or sentiment-direction bars for any currency) will be much bigger, on average, than their counterparts. This is where the trend exists, in the average size of one bar over another. And that is why you can set a tight SL and a looser TP, and how you break the TP-SL parity problem. The rest comes down to finding a good place to put your stop loss (hang your hat, as I call it).
  • #54
  • Dec 17, 2009 12:02am Dec 17, 2009 12:02am
Good thread, guys!

Examples of inconsistencies that we can use to help us:

Inconsistencies is time: Market opens vs. afternoons; Tues vs. Fridays; Holiday seasons (BTW: Happy Holidays!) vs. non-holiday etc.

Inconsistencies of price: Relative difference in interest level of market participants in price reaching .42 vs. .00 levels.

Inconsistencies in correlations: periods of trading where a particular currency exerts a uniform pressure on the others (e.g. XXX/USD) vs. periods of less across the board influence.

Inconsistencies in "outside" factors: other-market influences or lack thereof.

Many more examples can be given. Obviously, not all factors are easily recognized or profited from but, hey - it's not supposed to be easy, right?

My point is that IMO your chances of winning/losing are constant - unless you do something about it by working on recognizing the inconsistencies that exist and putting them to work for you.

Thanks for the thread and, to all who celebrate them, have a great holiday season!

Brent
  • #55
  • Dec 17, 2009 12:36am Dec 17, 2009 12:36am
Quoting hanover
By 'expectancy', I meant the expected outcome of a trade, as explained by Van Tharp in Trade Your Way To Financial Freedom. I've also heard it called 'edge' or 'profit factor'.

Expectancy = (win% / loss%) x (net avg win size / net avg loss size)
Thank you, hanover, for the equation. One equation clarifies everything - it was my mistake for thinking that "expectancy" was mathematical "Expected Value" which is:

Expected_Value = ( win% x net avg win size - loss% x net avg loss size)
= SUM( pnl_per_trade x perc_chance_of_occurance )

FWIW, there are different degrees of mathematical randomness:

1. random, but stationary (eg rolling dice, casino) where if you take statistical measures like average or standard deviation over any long enough time period, it does not change much.

2. random, but characterizable as non-stationary (some long-term financial time series, random walk). Rather than repeat the various flavors, look at this:

http://www.investopedia.com/articles...stationary.asp

3. chaotic, where the system should be non-linearly deterministic, but is so sensitive to initial conditions that the results look random (weather)

http://en.wikipedia.org/wiki/Chaos_theory

There are many who try to characterize pricing time series like equity prices or forex exchange rates as 2. or 3., but that's even an approximation. The reality is actually something more complex and wilder.

But one thing is rather certain, chance of winning/losing is not 50/50 - that's an oversimplification. It's a little like saying, I don't know if it's going to rain tomorrow or not, so I say it's 50/50 since it can only rain or not rain. Well, if I have no other information or experience of rain, I might say that, but if you randomly put me on a spot on earth, chances are, that 50/50 would be way off.

At the core, price is an aggregation of people making decisions off fear and greed, and we can't model human behavior well, even on a statistical basis. But when there is a way to hedge underlying prices, we can price financial derivatives like options rather well - as long as "black swans" don't show up at the party - tail risk is ongoing research.
  • #56
  • Dec 17, 2009 9:11am Dec 17, 2009 9:11am
Quoting Rabid
I don't think you can say that just because the next tick may be up, down or sideways, that the cumulative odds of the market going up or down is 50/50.
I should have qualified this assertion by stating that this movement is in the absence of a particular trading method or attention to market bias. I'm not going to assume that anyone observing the market has a particular strategy from which they are operating. And the reality is that even if one has a strategy one has to make an arbitrary decision as to where to place a profit or stop target. The placement of a target in one place versus another will likely result in a different outcome because of the chaotic movement of the market. And you can't know for sure where the chaotic movement will take the market price.


Quoting Rabid
Chaotic is not random. Nobody is talking about a guarantee, we're talking about odds. It is mathematically provable that is impossible to turn a profit over time without an edge. MM does not effect edge.I can also demonstrate that there are times when the odds of a market outcome is not 50/50. The trick is figuring out when those times are.
Actually chaos is random. Playing odds has to do with trying to work a range of probability based on historical averages. The simplest target in trading, therefore, is looking at historical trade range. I'm sure that the consideration of trade range is integral to how you judge the market to gain the edge for successful trading. I did introduce the concept of an edge with the term market bias. The idea there is that it shifts the odds away from 50/50 and in a trader's favor.

And I disagree regarding money management. In my experience it is an important part of the successful edge. More about that in a minute.


Quoting Rabid
The edge in this system, then, would be determining when the market was least likely to trend away from you.
That is the real trick, isn't it?


Quoting Rabid
If you were to trade open w/o an SL, with a 10 pip target, it might be convenient to think you'd never have a problem. But all it takes is 1 bad trade to wipe you out. Naturally then if you're trading counter to a major trend, this could be a serious problem. Thus the edge there depends on accurately guessing the long range trend enough that your account is rarely destroyed. As always, MM is only a factor in determining when that will be.
"Accurately guessing" is a bit of an oxymoron, don't you think? Actually, one of the harder things to figure out and deal with is the long range trend because big sweeping changes in market direction mean the trade range across the trend change is large (the bigger the timeframe of the trend, the bigger the trade range).

In my experience money management is not just a factor in trading, it is equally if not more important than the entry strategy. Many times a trade taken from a particular market signal evolves into something entirely different when the market moves in such a way that the original signal becomes meaningless. It is MM that has allowed me to roll with the changes.

The most important aspect of money management though is actually very simple: Don't over-leverage. The success of most strategies depends greatly on the component of money management to help create balance in trading. In my opinion that is as integral a part of successful trading as searching the market for the highest probability trades.


Quoting Rabid
Generally, that's what we tell people. "You can't know for sure if..." but truth is, you can. I do it all the time. You can easily tell when a trend is slowing down, it gives off massive signals when it does.
I wouldn't doubt that you can read the market in such a way that you can trade profitably. I would however be cautious about massive signals. In my experience sometimes what appeared to be massive signals turned into massive mistakes.
  • #57
  • Dec 17, 2009 2:57pm Dec 17, 2009 2:57pm
Quoting natanas
And the reality is that even if one has a strategy one has to make an arbitrary decision as to where to place a profit or stop target. The placement of a target in one place versus another will likely result in a different outcome because of the chaotic movement of the market.
Yes, but there's far more to it than just placing a stop loss and a TP. If you just randomly trade over time, no SL/TP combo will work for you. Given enough time, the mean always wins out. There's a timing element to the problem.

Quote
Actually chaos is random. Playing odds has to do with trying to work a range of probability based on historical averages.
Chaos is not random. There's a mathematical difference between randomness and chaos. The market is a little bit of both. The thing about an edge, it's more than playing the odds. It manifests as odds, but works because of a fundamental inefficiency in trading.

Quote
And I disagree regarding money management. In my experience it is an important part of the successful edge. More about that in a minute.
I can mathematically prove that the tp-sl parity exists over all choices of lot size. Again, make sure you're talking about trade, rather than profile, expectancy.

Quote
"Accurately guessing" is a bit of an oxymoron, don't you think?
Not at all. Let's play 9 or greater. The game rolls a 6-sided die 2 times in a row. You can bet before the first roll, after the first roll, not after the 2nd. Return is even money, put $1, get $2 (your original $1, and $1 for winning) if you win, lose the $1 if you don't.

In this game, can you determine a strategy to beat the house?

Of course. There are 36 possibilities. The probabilities work like thus:
1 - 0 rolls for a 9
2 - 0 rolls for a 9
3 - 1 roll for a 9 [6]
4 - 2 rolls for a 9 [6,5)
5 - 3 rolls for a 9 [6,5,4]
6 - 4 rolls for a 9 [6,5,4,3]

There are a total 10 possible ways to reach a 9 or greater, then, right? That gives 10/36, or 5/18, or 27.78% chance of winning. Given even money payback, you can't possibly win with random bets.

But. If you confine your bets to when 6 is rolled and only when a 6 is rolled, your odds of winning go from 5/18s to 2/3rds. At that point, even money is more than enough.

This is a basic conditional probabilities game. The market is obviously more complex than that, but it works the same way.

In the market all trades are guesses, but they must be accurate to work. No given bet in the above game is guaranteed to win, but the law of large numbers will carry your edge to it's logical mean one way or another.

Quote
Actually, one of the harder things to figure out and deal with is the long range trend because big sweeping changes in market direction mean the trade range across the trend change is large (the bigger the timeframe of the trend, the bigger the trade range).
Sometimes. There are times within every trend where range contracts, however. But that's splitting hairs. From my perspective, a trend is nothing but a very large range. But if you're locked into smaller time frames then yes, you'll have to deal with larger time frames as you go.

Quote
It is MM that has allowed me to roll with the changes.
Yes, but that's not increasing your trade expectancy. Like I said earlier, MM allows you to trade a statistically significant number of trades to reach your mean. It provides no edge to an individual trade, but it does effect your entire trading profile. We are not talking about profile expectancy, we're talking about single trade expectancy.

Quote
The most important aspect of money management though is actually very simple: Don't over-leverage.
Yes, but that does not effect single trade expectancy (barring a psychological issue that prevents you from following your trade rules, which is tangential to the issue of the trade itself). Over-leveraging simply means that you can't survive variance the way you should, and thus break before the mean regression. That's a profile issue, not a trade-specific issue.


Quote
I wouldn't doubt that you can read the market in such a way that you can trade profitably. I would however be cautious about massive signals. In my experience sometimes what appeared to be massive signals turned into massive mistakes.
Go to the weekly, zoom out. Look for blocks of consolidation. Draw a top, draw a bottom. Range will consolidate tighter and tighter. Some markets, if you look around, go for years like this. But when the market broke out and found it's S&R, the new trend was established. That's a pretty big signal.
  • #58
  • Dec 17, 2009 3:38pm Dec 17, 2009 3:38pm
An inspiring response, Rabid. You've got me thinking a bit more about the finer points of mathematics.
  • #59
  • Aug 19, 2010 5:37am Aug 19, 2010 5:37am
Gspagon, 100 pip TP & SL you may like to consider Rabids game theory for entry. Scaling into and out of your positions {not excessivly due to spread costs} as you have the ability to change your max loss while the dice is in the air, or/& move your stop loss to say break evan plus spread [or commision for you ecn boys n girls] once PA hits 50 pip profit.
  • #60
  • Aug 19, 2010 8:08am Aug 19, 2010 8:08am
Was just reading this thread...It's interesting from the standpoint of analysis of the mathematics and probabilities of trading. However, it also assumes that most of the people on here talking about the percentage of sucess see most trades as random. That's because most on here still fail to be able to discern high probability trades. A great day trader can recognize those few seconds or minutes when the market "tips it's hand" and go in and extract those profits. They can see those moments when the momentum and volatility will propel their chosen instrument in a high probability direction . Those who rely on lagging indicators..fail to see the real time market movement and are unable to understand that unforseen events..or traders can easily reverse a strong trend will perceive the market as a game of chance. And for them...it definitely is. Let's not forget the impact of institutional algorithmic trading computers. So, trading the way the average retail trader does, yes, the market seems mostly random.
0 traders viewing now