Long term Profits within Negative Expectancy | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

What is Expectancy? 13 replies

System Expectancy - Monte Carlo Simulation 5 replies

Short Term / Medium Term / Long Term? 3 replies

Random Market vs Set Expectancy 8 replies

Expectancy is always zero 41 replies

Attachments: Long term Profits within Negative Expectancy
Exit Attachments

Long term Profits within Negative Expectancy

  • #1
  • First Post: Aug 17, 2009 12:30pm Aug 17, 2009 12:30pm

Long Term Winning in a Negative Expectancy Game



***** Caveat***** Many will disagree with what I am about to write. If you cannot accept mathematical certainty, please do not waste your time reading this.


Trading retail forex is a negative expectancy activity. I hope all readers understand what negative expectancy is….but I will give a very brief description. Every trade taken opens with a negative in the profit/loss column because of the spread. This spread must be overcome on any trade before profit can be posted. If there were no spread, forex would be a game of neutral expectancy, meaning that overtime random entries could be made with equidistant profit and stop loss points. Over a large number of trades you would expect to break even. However, the spread makes that profit and stop loss points non-equidistant and puts forex trading in the realm of negative expectancy. If you go long E/U with a 2 pip spread, you must go a full 10 pips higher for a 10 pip gain, but you need only 8 pips lower for a full 10 pip loss. I hope that everyone can see that, from a probability standpoint, forex is negative expectancy.

So, how to overcome this negative expectancy? You must have a system that correctly predicts the direction of a currency pair at a rate of greater than 52% (give a 2 pip spread) while keeping profit and loss points equidistant. You may manipulate your system up or down this scale. For instance, you may decrease your win %, but you must proportionally adjust your profit point away from equidistant to compensate for the lower win rate.

Is there a system that can produce positive expectancy with equidistant profit/ loss points? I do not believe there is such a system available to small retail traders. I have seen many claims of such systems, but none that I have reviewed hold up to long term scrutiny. Please do not buy a retail forex system. If any of these systems could produce a positive expectancy, even a small one, the developer would be rich beyond their wildest dreams and not interested trying to get you to send them your money.

I have a winning system—how long until I am rich? Maybe you have access to inside information and can predict currency direction at a winning rate. You are still not home free. You must overcome randomness in the market and manage your money so as to not let randomness and volitility wipe you out. Consider the following: You have a system that produces 60% win rate given equidistant profit loss points. How many trades must I enter to be 99.9% certain of a win? N= log(1-.999)/ log(1-.60). This yields N=7.5. This number tells us that every 1000 trades you will hit at least 1 stretch of at least 7 or 8 losers in a row--- even with an incredibly high win rate of 60%. Don’t be surprised when these losses happen, they will happen to a mathematical certainly of 99.9%.

Some people make big money in forex. Absolutely some do. Some take accounts of $500 and turn that into $20,000 in a month. Some blackjack players take $500 and turn it into $20,000 in a few hours. Same for craps, roulette, etc. All these people are experiencing short term volatility in a positive way. It cannot be sustained and if they continue to trade or play for an extended period of time their account balance will be zero.

Conclusion. Please do not take this to mean I am a kill joy! I have traded forex for quite a few years and have had periods of great success, along with great disaster. I still trade from time to time, but I do it for fun, somewhat akin to casino gambling. If you are trading retail for any other reason that this, you will be disappointed. You may be successful for a few months or even years, but in the end if you trade long enough you will be wiped out.
  • #2
  • Aug 17, 2009 12:49pm Aug 17, 2009 12:49pm
All perfectly true, assuming your belief that an edge is impossible to gain. Which is what people will have gripes with, nothing wrong with the maths.
  • #3
  • Aug 17, 2009 1:03pm Aug 17, 2009 1:03pm
Quoting FX-Petra
So, how to overcome this negative expectancy? You must have a system that correctly predicts the direction of a currency pair at a rate of greater than 52% (give a 2 pip spread) while keeping profit and loss points equidistant. You may manipulate your system up or down this scale. For instance, you may decrease your win %, but you must proportionally adjust your profit point away from equidistant to compensate for the lower win rate.
A flaw in your logic is the assumption that the movement of price in a given forex pair is "random"; nothing could be further from the truth. There are patterns and tendencies of price movement resulting from recurring conditions in the fundamental data set that drives price, and these can be reliably exploited, perhaps predicted, and traded profitably.

Also, you seem to suggest (above) that an adjustment of "win rate" necessarily is accompanied by a corresponding adjustment in the ratio of profit/loss risked. I see that as a non-sequitur.

Just my 2 cents. Thanks for raising an interesting topic.
"If The Fool persists in his Folly he will become wise." - William Blake
  • #4
  • Aug 17, 2009 1:21pm Aug 17, 2009 1:21pm
This is COMPLETELY WRONG!

Only kidding. This didn't jar at all, and when looking at any kind of system, a binary R:R (1:1) is the first thing I look at. Changing of R:R parameters is then a tweaking exercise rather than the foundation.

I wouldn't share your pessimism though, but then that entirely depends on your expectancy, as you rightly point out.
  • #5
  • Aug 17, 2009 1:21pm Aug 17, 2009 1:21pm
Quoting The Fool
A flaw in your logic is the assumption that the movement of price in a given forex pair is "random"; nothing could be further from the truth. There are patterns and tendencies of price movement resulting from recurring conditions in the fundamental data set that drives price, and these can be reliably exploited, perhaps predicted, and traded profitably.

Also, you seem to suggest (above) that an adjustment of "win rate" necessarily is accompanied by a corresponding adjustment in the ratio of profit/loss risked. I see that as a non-sequitur....
"Random" is a loaded word. Understanding randomness is a topic phd dissertations are built on. I don't care if it is random, I just want to know if it can be predicted correctly more than 52% of the time, give equidistant profit/loss points. If anyone can do that over a large number of trades, let them step forward. I do not believe it can be done.
  • #6
  • Aug 17, 2009 1:34pm Aug 17, 2009 1:34pm
Quoting FX-Petra
"Random" is a loaded word. Understanding randomness is a topic phd dissertations are built on. I don't care if it is random, I just want to know if it can be predicted correctly more than 52% of the time, give equidistant profit/loss points. If anyone can do that over a large number of trades, let them step forward. I do not believe it can be done.
Well, I don't share your disbelief but I don't have hard data to offer in any kind of rigorous way so we might have to leave it at that, for now.

Of course, the linchpin of "proper money management" as it is often presented around here is that you need to have a risk/reward ratio of 1:2 or higher, not with equidistant profit/loss points, as you suggest, implying risk/reward of 1:1. If one is disciplined about trading with a risk/reward profile of 1:2 or better, then of course one may realize consistent profits with a "hit" rate of <50%.

Back to my original point - I did not understand why you insist on stipulating equidistant profit/loss, nor how an adjustment therof necessarily implied an adjustment of "hit rate". The two are not necessarily dependent. I can have an 80% hit rate with a risk/reward profile of 1:1 or 1:5 or whatever.

Now, from personal experience I can tell you that my own various failures have been mostly due to breakdowns in the discipline required to stick to simple rules of money management....
"If The Fool persists in his Folly he will become wise." - William Blake
  • #7
  • Aug 17, 2009 1:42pm Aug 17, 2009 1:42pm
Quoting FX-Petra
"Random" is a loaded word. Understanding randomness is a topic phd dissertations are built on. I don't care if it is random, I just want to know if it can be predicted correctly more than 52% of the time, give equidistant profit/loss points. If anyone can do that over a large number of trades, let them step forward. I do not believe it can be done.
Hi,

What would be 'a large number' of trades that would convince you?

regards
daytrading
Enter Signature
  • #8
  • Aug 17, 2009 1:43pm Aug 17, 2009 1:43pm
Trading is negative expectancy to a very tiny degree when trading larger trends. 2 pip spreads, when a SL and TP are 200 pips, isn't really that big of a factor. Interest rates are more than likely going to determine to a greater degree whether one is profitable over a long period. My advice to those with limited resources will always be to zoom out and go slow. If spreads are holding you back, then you are trading wrong... plain and simple.

The Fool, agreed in some sense, nothing is "luck" or purely "random". However, this doesn't mean that a person can spot patterns or circumstances that will give them 50%+ win ratios. Those are likely only to occur when the trader actually creates their own probability by manipulating all the aspects of trading, entry, MM, exit, etc.
The market is my nation. Traders, my family. Hello, brothers and sisters!
  • #9
  • Aug 17, 2009 1:47pm Aug 17, 2009 1:47pm
Quoting The Fool
Well, I don't share your disbelief but I don't have hard data to offer in any kind of rigorous way so we might have to leave it at that, for now.

Of course, the linchpin of "proper money management" as it is often presented around here is that you need to have a risk/reward ratio of 1:2 or higher, not with equidistant profit/loss points, as you suggest, implying risk/reward of 1:1. If one is disciplined about trading with a risk/reward profile of 1:2 or better, then of course one may realize consistent profits with a "hit" rate of <50%.

Back...
I have simply stated the MINMUM success that a system must achieve to produce long term profits. You MUST over come the spread +1 pip to be profitable. Stated another way, you must be able to achieve 1 more pip in profit than you are willilng to risk in a loss. Any system that can produce an 11 pip win vs 8 pip loss (and 2 pip spread) more than 52% of the time will make you rich beyond your wildest dreams.
I can change those numbers in any way you wish, but they must remain proportional for an system to achieve minimum profitability.
  • #10
  • Aug 17, 2009 1:56pm Aug 17, 2009 1:56pm
Quoting daytrading
Hi,

What would be 'a large number' of trades that would convince you?

regards
daytrading
10,000 trades @ 52% win rate would not be significant enoungh, from a mathmatical persective, to draw any conclusions. However, from a practical view point, if I could verify a system that could be a winner after 10,000 trades (or trials) I would start to get excited!
  • #11
  • Aug 17, 2009 2:15pm Aug 17, 2009 2:15pm
Quoting FX-Petra
10,000 trades @ 52% win rate would not be significant enoungh, from a mathmatical persective, to draw any conclusions. However, from a practical view point, if I could verify a system that could be a winner after 10,000 trades (or trials) I would start to get excited!
What about a person that can be a winner after 10,000 trades?

Systems do not trade. Traders do. Traders are people. Mathematics is not big enough, not deep enough to provide the measure of what a person can do.
"If The Fool persists in his Folly he will become wise." - William Blake
  • #12
  • Aug 17, 2009 2:19pm Aug 17, 2009 2:19pm
Quoting FX-Petra
10,000 trades @ 52% win rate would not be significant enoungh, from a mathmatical persective, to draw any conclusions. However, from a practical view point, if I could verify a system that could be a winner after 10,000 trades (or trials) I would start to get excited!
hahaha...

If you have a model that trades once a day only (which is not uncommon) and has a statistical advantage, you are talking approximately 38.5 years of data.

Hedge funds employ statisticians that get excited about statistical edges that are far more short-lived than that. The difference lies in the ability to know for how long the edge will last - or when it is about to fail due to a paradigm shift that is unaccounted for.

I have developed a model which has such an advantage (trades 5 times per day, with 1300 trades per annum).

It takes full trading costs (spreads, rollovers, etc.) into account.

By the time the models behaviour starts to shift, adjustments will be made.

Over the last 3 years, the annual avg return was +36%. Worst real equity DD was -7.5%.

I can live with the knowledge that the model will not work over 38 years profitable....because it does not need to.

It is very old news that inefficiencies disappear - and it does not matter since new ones can be exploited. Funds pay people good money to look constantly for new inefficiencies because they know - and so do professional retail traders. Only fools believe that trading models work over unlimited periods of time.

regards
daytrading
Enter Signature
  • #13
  • Aug 17, 2009 2:40pm Aug 17, 2009 2:40pm
DT, thought 10,000 might make your eyebrows raise...
Can I ask, slight diversion, is there anything other than actual losses in your system that'd inform you that it was past its sell by date?
  • #14
  • Aug 17, 2009 2:50pm Aug 17, 2009 2:50pm
Quoting triphop
DT, thought 10,000 might make your eyebrows raise...
Can I ask, slight diversion, is there anything other than actual losses in your system that'd inform you that it was past its sell by date?

Many factors, Triphop -

Sequence of consecutive losses, distribution of different size losses over different time periods, distribution over selected time periods of outsized winning trades and many more.

It works a bit like in a surgery when you have to check the patients blood pressure, heart rate, blood sample, colesterol, liver etc, etc. before you know whats wrong.

If its just a cold, then there will be no need for surgery -

regards
daytrading
Enter Signature
  • #15
  • Aug 17, 2009 3:52pm Aug 17, 2009 3:52pm
Quoting The Fool
What about a person that can be a winner after 10,000 trades?

Systems do not trade. Traders do. Traders are people. Mathematics is not big enough, not deep enough to provide the measure of what a person can do.
You just haven't done enough maths my friend.

After all a trader IS a system, so a person being a winner after 10,000 trades is the same thing. All we have to do now is find a journal with 10,000 calls in it
  • #16
  • Aug 17, 2009 4:47pm Aug 17, 2009 4:47pm
Quoting daytrading
Many factors, Triphop -

Sequence of consecutive losses, distribution of different size losses over different time periods, distribution over selected time periods of outsized winning trades and many more.

It works a bit like in a surgery when you have to check the patients blood pressure, heart rate, blood sample, colesterol, liver etc, etc. before you know whats wrong.

If its just a cold, then there will be no need for surgery -

regards
daytrading
Ah, that makes sense, and may your patients live a long and fruitful life. I have been a bit simplistic in the past by looking at raw % DD. I will investigate.
  • #17
  • Aug 17, 2009 5:03pm Aug 17, 2009 5:03pm
I must be doing some thing wrong then. I just made over 5000.00 the past 3 weeks. The last EURUSD swing was approx. 206 pips, from which I collected 198 with just 4 trades -- very nearly 100% win ratio.

But that's way more than 2% over neutral, so I must be doing something wrong. Right?

/snark
" Average traders look to the left, great traders look to the right."
  • #18
  • Aug 17, 2009 11:42pm Aug 17, 2009 11:42pm
@FX-Petra, great post, and it has elicited some very thought-provoking replies. Should be manadatory reading for newbies, and a reality check for the complacent among us.

As you also correctly point out, a small edge (plus conservative MM) is all that's (potentially) needed to grind out a life-changing profit. We always seem to be reading about the importance of discipline, but for me finding a genuine methodological edge has proved to be the most difficult element. Does a (TA-based) edge, and I mean one with the potential to last a lifetime, actually exist? I have no idea.

I don't think the argument as to whether it's possible to attain positive expectancy in trading can ever be proven. Those who believe that it's impossible must concede that their current level of knowledge is finite. While those who claim an undeniably profitable track record can not prove that it will last forever.

There's an interesting post by Darkstar here. I note his final remark: "You’re either exploiting inefficiency in the market or you’re providing the liquidity for those that do. Everything else is a roll of the dice."

One small point: it is (or at least, was) possible to gain positive expectancy playing Blackjack, by counting cards, although it depended on casino rules and conditions. Which of course is why casinos introduced countermeasures.

@Magicchance, an excellent result over 4 trades! However, consistency is key. If you can sustain that level of performance over several months, then you might yet make a believer out of me.
  • #19
  • Aug 18, 2009 12:21am Aug 18, 2009 12:21am
Quoting FX-Petra
..... ..... Trading retail forex is a negative expectancy activity........

....... You may be successful for a few months or even years, but in the end if you trade long enough you will be wiped out.
What a great post! what a great post! Thanks for sharing views......................... However, would like to express these two differing opinions in view of the last sentence (bold above).

1. Forex trading is not just mathematics, statistics and patterns - real economic factors support mega moves and trends. Therefore positive expectancy is most probable for informed and "smart" traders. By smart I mean traders who are aware of short-term abrupt behavior of market and accordingly manage their positions and funds.

2. If a trader is "consistent" in its true sense, this market cannot eliminate that trader. If this was untrue, there would have been no bookmaker left in horseracing - my third generation family business. Kindly note that factors affecting currency prices are far more visible and effective in currency price speculation than those that indicate or affect performance of runners in a horserace.
The Thief of Wall Street
  • #20
  • Aug 18, 2009 12:24am Aug 18, 2009 12:24am
@Hanover. I know who you are here and I respect you very much. I'm not a programmer like you, I'm just a small time trader with much to learn.

The results I just posted are 100% true and accurate. I will be the first to admit that the 198 pips out of a possible 206 sounds extreme, and it is. It was a good week and it surprised even me. But I must also say that my normal win ratio is a SOLID 75-80%, and has been since last year.
" Average traders look to the left, great traders look to the right."
0 traders viewing now