Twelve widely believed forex myths | Page 2 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Twelve stepping stones for rookies 21 replies

Trading Myths and Some Forex Math 337 replies

10k (0.1) Position $300 to $64k In Twelve Weeks 17 replies

Forex Myths 29 replies

Anyone know most widely used EA? 1 reply

Attachments: Twelve widely believed forex myths
Exit Attachments

Twelve widely believed forex myths

  • #21
  • Dec 30, 2010 7:49am Dec 30, 2010 7:49am
Quoting nhct
#2 is not, in general, a trading myth.

If a given negative expectancy strategy happens to have a long-term observed trade distribution with a negative serial autocorrelation, then MM, all by itself, expressly designed to take advantage of that property, will either reduce the negative expectancy or turn it positive.

In plain English, negative serial autocorrelation means, in the simplest possible terms, that a loss is more likely than not to be followed by a win.

One good example on this forum -- although I would think there must be quite...
Correction!!! MM is not going to change the expectancy of the strategy, it will just alter the cumulative sum of profits and losses and thus might lead the strategy to yield either smaller losses or even in rare cases profits.

The expectancy of the strategy is only determined by the probabilities of winners and losers and the average winner and the average loser according to the formula of expectancy.

Expectancy = (PWinners x AWinner) - (PLosers x ALoser)

Since MM is not included in the formula, you can't assume that changing position sizing will have any effect whatsoever on Expectancy. It will only affect your cumulative P&L and that's all.

Quoting nhct
In plain English, negative serial autocorrelation means, in the simplest possible terms, that a loss is more likely than not to be followed by a win.
In the world of independent outcomes, this can never be the case. Independent outcomes means that every outcome does not depend by any mean of previous. That might be true for gambling games, however, might not be very accurate for markets. Yet, you will have to assess the risk of ruin.

Let me give you a practical example. A trend following strategy gets burned and experiences the losing streak during trendless periods of the market. If you were using a trend following strategy for the Dow Jones Industrial Average between 1965 and 1983, you would have been burned to death several times. Now putting the same strategy to work now, how can you make sure about the distribution of the winners or the losers for the next 10 years???!!! You really can't...Because as the sample gets bigger, from 1,000 trades to 10,000, the streaks become larger in terms of numbers.

In case of a 1,000 trade sample, if losers are 60%, then 600 trades are losers, but in the case of 10,000 trades and for the same percentage, then it's 6,000 trades. Out of 600 trades, what do you think the odds for a 30 losing trades streak would be? With the same in mind, out of 6,000 trades, what do you think the odds for a 300 losing trades streak would be?

Best Regards,

Nader

Best Regards,

Nader
  • #22
  • Dec 30, 2010 8:11am Dec 30, 2010 8:11am
Here's another myth. Trading the asian session is easier than others...
  • #23
  • Dec 30, 2010 8:19am Dec 30, 2010 8:19am
Quoting TJPLD
It doesn't matter if you compare nominal leverage or effective leverage. Both have absolutely nothing to do with your risk.
It could of course matter - depending on 'what' a trader uses as the calculation base for his/her positions.

Regards,
Daytrading

P.S. Happy New Year
Enter Signature
  • #24
  • Dec 30, 2010 8:58am Dec 30, 2010 8:58am
Quoting Intensity
Thanks for the explanation TJPLD !



I would say that 6:1 is more risky because an order can easily go 15pips against you while there are often retracements on a 70 pips move that would allow you to exit at breakeven....
Given a single trade, the risk is practically the same despite the difference in leverage, 1:1 and 6:1. Comparing the risk of 1 trade with 3 others might not really be that accurate.

It goes back to what you level are you looking at. If you are looking at the risk of individual trades against each other, then for sure the risk is the same.

However, if you are looking at the overall risk of a series of trades, then your assumption must be true.

Best Regards,

Nader
  • #25
  • Dec 30, 2010 9:03am Dec 30, 2010 9:03am
Here is an another myth:

* Good MM: Enter two lots. Take profit for one after X pips, move the other SL to breakeven an let it ride.

- MM has nothing to do with the above sentence. What you have here are essentially two independent strategies. First enters on the signal and takes X pips. Second moves the SL to breakeven after X pips and is hoping to capture a big move. You should independently evaluate eachother to make sure that both are profitable. If one of them is not you are actually loosing money won from the other strategy. If both of them are profitable there is a good merit to evaluate them only by their performance to see why they are performing as they are and to try to optimize them more if some solution should rise.
  • #26
  • Jan 1, 2011 1:47pm Jan 1, 2011 1:47pm
Quoting Rikers
Here is an another myth:

* Good MM: Enter two lots. Take profit for one after X pips, move the other SL to breakeven an let it ride.

- MM has nothing to do with the above sentence. What you have here are essentially two independent strategies.......
Agree with you. This again highlights two principles in my original post:

1. A method (entry+exit) must somehow be grounded in market behavior and probabilities, i.e. exploit an inefficiency, in order to be profitable. MM alone can not create an edge.

2. In any form of scaling in/out, each component position must be profitable in its own right, in order to contribute positively to bottom line.

Happy New Year,
David
  • #27
  • Jan 1, 2011 3:00pm Jan 1, 2011 3:00pm
Quoting hanover
5. That it's impossible to forecast price direction.
Every trading decision (e.g. entry, exit) reflects an implicit forecast, or at least a forecast on the basis of probability. The more accurate one is at adjusting one's position relative to market turns, the more pips one banks. If it's impossible to determine directional bias, or at least some kind of behavioral bias, at some points in the market, then it's impossible to have any kind of edge.

Of course it's possible to have an edge, but be oblivious to the fact that...
Hanover, you are one of the only people on this forum I have any interest in listening to, as you are always intelligent, logical and from what I can see, correct. However in this instance I have to disagree or at least question the meaning (hopefully in a constructive way).

I am a firm believer that it is impossible to forecast price direction. To do so would mean the ability to know the thoughts, feelings and beliefs of every participant and waiting participant, and also know how those thoughts, feelings and beliefs will make those individuals act. Maybe I'm wrong, maybe not.

But the main point is that I have a profitable strategy, traded live over several years, and I 100% categorically believe I have NO edge other than my trade management, i.e. how I exit winning and losing trades. I don't make any attempt to predict the market direction for my entries.

In brief, I only enter trades during certain intra day market hours, and if the price is moving higher, I go long. If the price is moving lower, I go short. If the price changes direction and goes against me, I get out ASAP. If price continues going in my favour I let it go as long as possible with a specific type of trailing stop.

I have a very tight stop loss and therefore get whipsawed horrendously and suffer some massive draw downs. My success rate of profitable trades is around 37%. However, I have no fixed take profit target and can have individual wins which return massive percent gains on my account and ultimately compensates in kind for the many small losses. My strategy is based almost entirely on cutting losses short and letting winners run, and it proved extremely profitable in my testing over multiple instruments without any optimisation for individual instruments, and also extremely profitable over the long term in my live trading.

My edge is in the trade management once the trade is open. I have no edge in my ability to predict market direction.

I'm not saying I don't have an edge, simply my edge is most definitely not in my ability to predict the market direction. I get the market direction wrong more than 60% of the time.

I may have misunderstood your meaning on this particular issue, and if so I apologise, but I hope I've made some points for your (and others) consideration.

Happy New Year to all and best trading!
  • #28
  • Jan 1, 2011 3:37pm Jan 1, 2011 3:37pm
Quoting flotsom
Hanover, you are one of the only people on this forum I have any interest in listening to, as you are always intelligent, logical and from what I can see, correct. However in this instance I have to disagree or at least question the meaning (hopefully in a constructive way).................................................(snipped)
And flotsom, you have so eloquently outlined the simplest, and most effective way to trade any market!

The axioms get thrown around so many times, but it is only when you really start to apply them in their truest sense - buy when it's up; sell when it's down, and cut your losses short and let your winners run, that the pips, ticks, or whatever financial market it is you trade starts to finally make sense.

Happy New Year to you Sir!
  • #29
  • Jan 1, 2011 3:46pm Jan 1, 2011 3:46pm
Hanover for President

Very nice thread, H-Bone.
  • #30
  • Edited 8:10pm Jan 1, 2011 6:44pm | Edited 8:10pm
Quoting flotsom
I am a firm believer that it is impossible to forecast price direction. To do so would mean the ability to know the thoughts, feelings and beliefs of every participant and waiting participant, and also know how those thoughts, feelings and beliefs will make those individuals act.......
Flotsom, I likewise respect your views enormously. What you've described is, at least as I understand it, the very basis of 'trend trading', and I have satisfied myself that it can work very profitably (especially in the longer TFs where macroeconomic factors play themselves out), hence I continue to use many of these same principles myself.

I don't want it to seem that I'm adamantly defending my position, but I'd like to take the time to elaborate on what I meant in my original post, where I was trying to keep each of the 12 explanations brief. I likewise agree with you that to forecast direction with 100% certainty would require an omniscient view of the sentiment of all participants, an obvious impossibility. However, the key words in my post (which I did try to think through carefully at the time, LOL) are "exit", "implicit" and "basis of probability".

When I enter in the direction of a trend, I believe I'm making an implicit forecast, that price is more likely to continue in that direction than not, i.e. on the basis of probability. Otherwise I might as well toss a coin: heads = buy, tails = sell. In your post you say "if the price is moving higher, I go long. If the price is moving lower, I go short". The fact that you're both experienced and profitable suggests to me that there's some advantage in doing this, i.e. it's more than just an unfounded, random guess.

A win rate being lower than 50% could be the result of two things: either (1) the trader is entering in the wrong direction (e.g. he should be buying when price is moving lower, and selling when price is moving higher). But, like I just said, I seriously doubt that that's the reason in your case. Instead, I suspect it's (2) your tight SL that's the cause, and hence I think it's much more likely, just as you say, that your edge comes from your trade management. This may sound unreasonably pedantic, but I see exits that use a trailing stop as being based on an implicit forecast: that once price starts trending, it's better to ride the trend than exit too early. The forecast, again on the basis of probability, is that the a strong trend is more likely to continue, than reverse. I see that happening, especially in the longer TFs: for example, I made profitable trades (albeit in demo) by simply shorting the Euro around the time when Greece was in serious trouble last year. That involved some kind of implicit forecast: Bad news in Greece = probable continued fall in the value of the EUR relative to USD. And so on.

I guess I come from the standpoint of a discretionary trader, where every decision I make (buy, sell, stay out of the market; hold, or exit) as being only as good as price subsequently proves it to be. If I'm wrong more often than I'm right (in terms of both entry and/or exit) then my equity falls accordingly. But for a more mechanical trader, who simply and rigidly follows his entry and exit rules, being able to read the market's next move with any kind of expectancy is irrelevant. That expectancy has all been pre-built into the trading rules, the result of prior testing.

I hope that didn't sound too defensive or pedantic. I'm not really too adamant about anything, trading-wise, especially as there are many different roads to profit. My intention was to provoke some lively discussion, and it goes without saying that "forecasting" can mean different things to different folks.

Keep those pips rolling in 2011, my friend.

David

[EDIT] In "Myth #5" I also talked about what I call "directional bias", and "behavioral bias".

The former is obvious, i.e. a certain chart pattern in a certain context creating a bias immediately following which, price will likely spend more time moving in one direction than the other.

One possible example of behavioral bias might involve S/R. If price reaches a key S/R level, there might be a 50/50 probability of a breakout versus a bounce, but if one or the other occurs, price will likely move a (relatively) long way. Hence a straddle strategy might be in order, i.e. using a tight SL and then letting any resulting move run with a trailing SL. Another example is touched on by nhct, or specifically his link to Craig Harris' method. Mr Harris uses a technique whereby if his pattern (based on fractals backed by the angle of the stochastic plot) fails, he anticipates a move in the opposite direction that's strong enough to cover his immediately prior loss, and (come what may ) he increases his position size in an attempt to capitalize.
  • #31
  • Edited 4:31am Jan 2, 2011 3:14am | Edited 4:31am
Thanks for the reply hanover. As I suspected, my interpretation and/or understanding of your original post was slightly different to your intentions.

In the sense you described above, my strategy could easily be described as making some level of forecast so far as if the price is moving higher, I assume (forecast?) that it will continue to go higher, and there's my reason for entry. I'm not sure I would personally view it as a forecast of price direction, but as you said; "forecasting" can mean different things to different folks!

I still don't see my entry or my forecast of direction based on probability as giving me my edge. My low win rate is absolutely attributable to my tight stop loss, but even with a wider stop I suspect (though not certain) that my win rate would only increase to 50% which I would not be able to describe as an edge.

For me I genuinely view my entry direction as irrelevant. My annual profits are determined by what I do "after" I enter a trade rather than in "which direction" I enter the trade.

Hopefully I've put my opinion across in a constructive way which will help give some pause for thought to others. Certainly not trying to argue merely for the sake of argument.

Good trading to all. I forecast this year will be a profitable year


EDIT: I've realised after re-reading hanover's posts that I am concentrating too much on the issue of forecast of price direction at the point of entry which is not hanover's main point. I can accept the argument that once in a trade which goes in my favour, I am making a "forecast" that the trend will continue for some distance (based on probabilities) and I am going to ride it as long as possible.

I think my issue was that I am a purely mechanical trader with extremely rigid rules which I follow without any care or concern about "predictions", "forecasts" or "reading the market", and this was somewhat distorting my understanding of what hanover was trying to get across.
  • #32
  • Jan 2, 2011 3:49am Jan 2, 2011 3:49am
A Happy New Year to everyone!

flotsom, I bet that if you had used a 100 pips TP for every trade you would have won almost the same amount in 2010. Your edge is actually your skill to time and enter when the market has momentum. You either take a very small loss or you let the trade running but after the momentum is over your trade has a 50% chance of retracing to your entry point or going the same amount of pips in your direction.
  • #33
  • Jan 2, 2011 2:14pm Jan 2, 2011 2:14pm
Quoting flotsom
.......I think my issue was that I am a purely mechanical trader with extremely rigid rules which I follow without any care or concern about "predictions", "forecasts" or "reading the market"........
This is of particular interest to me personally, because I've not been able to find a strictly 'mechanical' approach that's profitable.

If it's possible to be profitable by simply picking any pair, any timeframe, making a truly random entry (by 'truly' random, I mean toss a coin to determine buy/sell, i.e. ignoring trend direction, momentum, volatility etc) and using simple, rigid rules to exit, then writing a profitable EA should be child's play. We could all retire rich, and effortlessly so.

Like I said, I use many of the same ideas (tight initial SL, trailing SL) but I also use multi-TF correlation, S/R confluences, climax volume candles, and currency strength (pair the strongest against the weakest to find the strongest, cleanest trends). Rightly or wrongly, I've assumed that these additional components contribute significantly to whatever edge I might have. My win rate is likewise < 50% due to the tight SL causing many trivial losses, but the low-risk entries plus longer term trends create a growing accumulation of high RR trades.

I'm in no way doubting that you're profitable. However, if being profitable is as easy as what you appear to be saying (random entry; rigid exit rules), then I'm envious , because (like everybody else) I'd love to have a profitable EA manage every aspect of my trading, generating a risk-free, stress-free passive income, leaving me free to get on with the rest of my life!!
  • #34
  • Jan 2, 2011 3:20pm Jan 2, 2011 3:20pm
Quoting hanover
If it's possible to be profitable by simply picking any pair, any timeframe, making a truly random entry (by 'truly' random, I mean toss a coin to determine buy/sell, i.e. ignoring trend direction, momentum, volatility etc) and using simple but rigid rules to exit, then writing a profitable EA should be child's play. We could all retire rich, and effortlessly so.
I'm not 100% sure that my own psychology could cope with a purely random coin toss entry as opposed to entering a trade in the direction of the immediate price movement. But there are many studies which seem to suggest that random entries are profitable with a good exit strategy. I believe Van Tharp attempted to show this, and also the following study:

http://www.compuvision.com.au/Exampl...GrailPart1.htm
http://www.compuvision.com.au/Exampl...GrailPart2.htm

As for my mechanical system, it can be automated which is what I am working on now via an API FIX for my currenex platform. The problem that makes it less easy than your expectation is down to the draw down. Can many people either manually stick to the rules, or sit back and watch a "robot" continue to trade while your equity depletes in a 30, 40 or 50% draw down?

From what I've seen of the vast majority of the Forex Factory population, taking 4 or 5 losses in a row is unacceptable and unmanageable (I've taken 34 losses in a row in the past). Of the hundreds or even thousands of systems available on this site, the instant they hit a 10-20% or more draw down, they are scrapped and called failures, or "tweaked/optimised" beyond recognition.

I think that is where there is a huge difference between mechanical trend following and other types of trading strategy. For my system to be profitable, I have to be able to ride out some genuinely massive draw downs which most would interpret as a failing/losing strategy. For me, the draw down is just a part of my absolute return strategy. Although I expect this draw down, it certainly isn't stress free watching it happen, and I would put money on it that the majority of people would pull the plug on any robot they watched that did this to their account........hence it's never as easy as it sounds.

Like you hanover, I try to incorporate other aspects to the system to improve my odds. Volatility and momentum play an important part. I'm an intra-day trader and I only enter trades during specific hours which give me better odds of catching big moves on the occasions I'm in the correct direction. My exit strategy is also not a standard, or at least not a well used form of trailing stop. So this isn't as simple as a random/careless/poorly thought out entry with an inflexible trailing stop exit. But, my method is strictly mechanical, can (from my testing and experience) be applied to just about any pair or in fact any instrument successfully over the long term, AND (moving back to the topic of our discussion) of all the components, the initial entry direction, is by far the least important element. For me, forecasting direction is irrelevant because my rules will make the best of which ever side of the market I find myself on after the entry.

I do hope I'm not detracting too far from the original intent of this thread.
  • #35
  • Jan 2, 2011 4:32pm Jan 2, 2011 4:32pm
Quoting flotsom
From what I've seen of the vast majority of the Forex Factory population, taking 4 or 5 losses in a row is unacceptable and unmanageable (I've taken 34 losses in a row in the past). Of the hundreds or even thousands of systems available on this site, the instant they hit a 10-20% or more draw down, they are scrapped and called failures, or "tweaked/optimised" beyond recognition.
This comes back to the trader psychology. Most traders usually don't have faith in their systems when it comes to losing streaks. That's basically because they haven't tested the system well to have that faith in during bad times.

A trading system must be tested for robustness, not only profitability, drawdown, losing trades, winning trades, etc...Almost 95% of the trading systems are not tested for robustness, not to mention that a lot of technical traders don't know about that in the first place.

When the system looks just great on historical backtesting, usually the trader will praise this and go ahead using it immediately, especially if it was a system with a +ve edge or +ve expectancy. However, if history says that the longest losing streak was 10 trades, the traders' mind immediately believes that 10 consecutive losers is the max he can get in the future and he starts to panic when he hit 8 in a row (Sometimes after 4 or 5 as you mentioned).

Very very few traders will wait and stick with the system even if the consecutive losses went up to 15. Those few actually know that the system is robust enough and will work and they know very well that historical results are not an indication of the future. The future IS not going to be EXACTLY like the past. It has to be different. The question is, HOW MUCH DIFFERENT?

Quoting flotsom
I think that is where there is a huge difference between mechanical trend following and other types of trading strategy. For my system to be profitable, I have to be able to ride out some genuinely massive draw downs which most would interpret as a failing/losing strategy. For me, the draw down is just a part of my absolute return strategy. Although I expect this draw down, it certainly isn't stress free watching it happen, and I would put money on it that the majority of people would pull the plug on any robot they watched that did this to their...
You can try to optimize your system so that it yields a lower drawdown and consequently a lower profitability. You will have to do so though without destroying the edge of the system and without curve-fitting the system of course.

Optimization is good but it must be done wisely and scientifically.


Regards,

Nader
  • #36
  • Jan 2, 2011 4:49pm Jan 2, 2011 4:49pm
Quoting narafa
A trading system must be tested for robustness...
And how would you go about that? Testing about exceptions in strategy - distribution of wins/losses is directly tied to the market behavior. ex. every strategy would fail tragically if the market went random.
  • #37
  • Jan 2, 2011 7:38pm Jan 2, 2011 7:38pm
Quoting narafa
This comes back to the trader psychology. Most traders usually don't have faith in their systems when it comes to losing streaks. That's basically because they haven't tested the system well to have that faith in during bad times.
Why put faith in a test? My confidence stems from understanding my philosophy, being able to apply it, and knowing why it is profitable.

If we are to test something, I'd prefer to test positioning of entries rather than wins and losses, because a poor entry can profit, and a good entry can lose. A good trader could see statistical significance within a couple dozen trades.
  • #38
  • Jan 2, 2011 8:04pm Jan 2, 2011 8:04pm
Quoting flotsom
I'm not 100% sure that my own psychology could cope with a purely random coin toss entry as opposed to entering a trade in the direction of the immediate price movement.......
I'm not sure that the tests Dr Tharp was referring to were specifically forex (see Trade Your Way To Financial Freedom, pp 200-201). FWIW, I posted on this topic here.

And can we assume that all markets 'trend' in exactly the same way? The psychology is probably very similar, but I believe that the other drivers, and market structure, are/is very different. FWIW, I posted what I saw as being some of the differences between stocks and forex here.

You still had the faith to persevere with your system after 34 consecutive losses? Wow, that's impressive! I recall a post by Joel Rensink where he said his edge was still valid, even after 24 losses in 26 trades. It's interesting how there appear to be two schools of thought: those who claim they're consistently making 10%-30% consistently every WEEK (e.g. here), and those who claim that forex is more difficult, that 30% per YEAR represents an excellent return, and that equity fluctuations, including long strings of losses, are occasionally inevitable. From my own experience, and given the returns of funds who supposedly have some hotshots and experienced vets among their trading personnel, I suspect that a result nearer the latter is a more realistic long-term target to aim for.

I apologize if I made it sound like your system was as simple as random entries and rigid exits. It's clearer to me, from your last post, that your entries aren't random, even if (as you say) they are the least important element.

This is a worthwhile discussion, IMO, it's definitely not straying from my original intent. When you start a thread, you never know where it will head. In any case, I don't see myself as 'owning' any thread that I've started. Intelligent discussion is always welcome.
  • #39
  • Jan 2, 2011 8:12pm Jan 2, 2011 8:12pm
Quoting Mr J
Why put faith in a test? My confidence stems from understanding my philosophy, being able to apply it, and knowing why it is profitable.
There's an interesting post by Merlin, where he talks (in the last paragraph) about starting with a premise based on one's understanding of market behavior, and then using backtesting to prove its validity; as opposed to concocting a premise from the results of one's testing.
  • #40
  • Jan 2, 2011 9:34pm Jan 2, 2011 9:34pm
Quoting hanover
There's an interesting post by Merlin, where he talks (in the last paragraph) about starting with a premise based on one's understanding of market behavior, and then using backtesting to prove its validity; as opposed to concocting a premise from the results of one's testing.
Yes, it's basically the argument for developing a hypothesis grounded in skill and experience, rather than developing one through datamining.
0 traders viewing now