Are you a consistently profitable trader? | Page 14 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Anyone consistently profitable using Expert Advisors? 45 replies

How to be consistently profitable... but I bet you won't do it 65 replies

How long did it take to become consistently profitable? 117 replies

Your consistently profitable trading style 37 replies

Consistently Profitable Signal Journal: Simple, Low Maintenance Gains 0 replies

Attachments: Are you a consistently profitable trader?
Exit Attachments

Are you a consistently profitable trader?

  • #261
  • Apr 3, 2012 6:55am Apr 3, 2012 6:55am
3. Persuade a gullible casino to change the rules, so that the odds are in your favor. Apparently that's what Johnson did. It's hard to believe that a big name casino would allow themselves to be so easily suckered. But that being the case, one could win easily enough by employing a simple playing strategy, and consistent and conservative MM.

There are definite parallels there to trading. Edge + conservative MM + consistent execution = $$$.

David[/quote]

hi hanover

this was the interesting bit for me,he realised that overall casinos were showing declining profits over an 18 month period due to new establishments gaining licences.this in turn led to the casinos extending there risk to increase profit,greed.he was able to negotiate favourable betting terms to bring the odds to 50/50,he never doubted his skill .i can compare this extended risk in a similar way to the banking industry leading up to the collapses.

yes i also found it interesting.
  • #262
  • Apr 3, 2012 10:14am Apr 3, 2012 10:14am
Quoting FX4Newbies
Thanks PhilipK.

I expect that you'll be getting a reply the next time they're on.

Take care and thanks for posting. Good observation and question.

FX4

I also wanted to get a bit of clarification if possible on Custos comments regarding a lucky streak. I agree that it's important to know if you have simply been lucky, because I look to this business as just that, a business. (or a potential business and profession).

Let's say that a given novice trader has a personality where the right method for him/her would be a fairly simple strategy using a pretty much bare chart.

now, he or she is using a specific entry, say an inside bar on a certain time frame and is not profitable yet.

now, he or she keeps plugging away and after a while this approach to trading yields some consistent results. After all, he or she has heard that a simple approach can be profitable - no need to over-complicate things.

so now this novice says to themselves, "ok things just may be turning around. I used some discretion in which inside bars to take and I'm profitable over x trades. (whatever x trades may be, say 1,000) I'm building my screen time and if anything, I should be able to make it slightly better, it looks like I can hold the trade for x more pips in xyz situation"

Then the performance heads downhill.


Was it a lucky streak?

Did the person have an edge which was lost?

Is there a difference?

What now?


this is from a trading educator who I get emails from fairly regularly. He is mostly selling a futures daytrading "system" I do not own it.

============================================
Number of Trades 50 100 200 300 500
<>Margin of Error 14% 10% 7% 6% 4%

The more trades you have in your back-testing, the smaller the margin of
error, and the higher the probability of producing profits in the future.

=============================================

What is your opinion on this?

All this was basically to ask, what are the possible ways to handle this situation?

this person seemed to be on track, maybe this inside bar approach doesn't work when the mkt gets too choppy. But now there is potential for confidence to be shaken.

and I think it's important for the person to know what he did right versus what was wrong in order to continue with his trading and education

This person might think "Damn that was just a lucky streak" But maybe a lot of progress was made and he/she now needs some assistance adapting to the new market.

Thanks for reading, any thoughts?
  • #263
  • Apr 3, 2012 10:39am Apr 3, 2012 10:39am
Quoting PhilipK
so now this novice says to themselves, "ok things just may be turning around. I used some discretion in which inside bars to take and I'm profitable over x trades. (whatever x trades may be, say 1,000) I'm building my screen time and if anything, I should be able to make it slightly better, it looks like I can hold the trade for x more pips in xyz situation"

Then the performance heads downhill.


Was it a lucky streak?

Did the person have an edge which was lost?

Is there a difference?

If you can't quantify discretion, how can you know whether the failure of the system is just temporary or permanent ? Discretion is an adaptation to changing market conditions. Some can adapt pretty quickly, others can not. Also, there could be several reasons for the failure:

1) Previous success was just a lucky streak
2) The system may be working under trendy high-volatility conditions and now the market may have become ranging and low-volatility.
3) Trader has lost his discretionary ability

Or a combination of the above.
  • #264
  • Apr 3, 2012 11:43am Apr 3, 2012 11:43am
Quoting micmic
If you can't quantify discretion, how can you know whether the failure of the system is just temporary or permanent ? Discretion is an adaptation to changing market conditions. Some can adapt pretty quickly, others can not. Also, there could be several reasons for the failure:

1) Previous success was just a lucky streak
2) The system may be working under trendy high-volatility conditions and now the market may have become ranging and low-volatility.
3) Trader has lost his discretionary ability

Or a combination of the above.

thank you for your comment micmic

I'm glad you defined discretion - Don't think I've seen a definition


1) Previous success was just a lucky streak
How can I determine this? If an approach works over 1000 trades Im lucky, over 2000 I'm very lucky...?


thought provoking post - I'm interested in ways of analyzing this type of stuff while trading a couple

simple strategies on a demo (simulator) just not sure how to determine 1) 2) 3) or a combination of the above
  • #265
  • Apr 3, 2012 12:23pm Apr 3, 2012 12:23pm
Quoting PhilipK
How can I determine this? If an approach works over 1000 trades Im lucky, over 2000 I'm very lucky...?
You have to set a failure rate that will cause re-examination of the system. That could be a percentage of lost trades in a specific number of trades, a percentage of lost trades over some period of time, a specific drawdown, etc. Nobody can do that except you, because only you know the specifics of the system. Some scalping systems may be a failure if you lose money one week, some longer term systems may be deemed as failures if you lose money one year. You will have to define yourself what failure is acceptable and what is not. Sometimes a system may be only good for as long as it works, ie you have found an inefficiency that you may exploit until it is noticed by too many people. There are systems that may work for a couple of years and then fail, but for some people these two years may be enough to make them millions. It all depends.
  • #266
  • Edited 12:49pm Apr 3, 2012 12:38pm | Edited 12:49pm
Quoting PhilipK
...so now this novice says to themselves, "ok things just may be turning around. I used some discretion in which inside bars to take and I'm profitable over x trades. (whatever x trades may be, say 1,000) I'm building my screen time and if anything, I should be able to make it slightly better, it looks like I can hold the trade for x more pips in xyz situation"

Then the performance heads downhill.

Was it a lucky streak?

Did the person have an edge which was lost?

Is there a difference?

What now?
Then the performance heads downhill.

Downhill indefinitely? For how many trades?

I think it is just as important, even more so, to qualify the "downhill" behaviour.

Will take it from there when the reply comes in .

In addition, it might be a better idea to focus on "qualifying" a successful system by the failure rate instead of by the winning rate. The lower the failure rate the better the system. Not sure how that would work...just a thought. Psychologically, if we want a system to work, we tend to see all the possible success and may not be too keen on seeing any failure in it. But, if we seek to disqualify it, if it can hold up, it will pass with flying colours. Then we'll "know" that we really have something that's been "through the fire" so to speak. Get my point?

FX4

P.S. Just saw Micmic's reply about selecting a failure rate, so I guess the train of thought is along the right path.
Trade what you see, not what you think or feel. Let the market come to you.
  • #267
  • Apr 3, 2012 12:55pm Apr 3, 2012 12:55pm
Quoting micmic
You have to set a failure rate that will cause re-examination of the system. That could be a percentage of lost trades in a specific number of trades, a percentage of lost trades over some period of time, a specific drawdown, etc. Nobody can do that except you, because only you know the specifics of the system. Some scalping systems may be a failure if you lose money one week, some longer term systems may be deemed as failures if you lose money one year. You will have to define yourself what failure is acceptable and what is not. Sometimes a system...
thanks again - I'd like an opinion on one last thing - The trading educator I mentioned stated that one should test their strategy on at least 200 trades.

That seemed low to me in general and I was wondering if micmic, FX4Newbies or anyone else would care to give their opinion /experience on roughly what number of trades would be considered a good sample, if, say, the approach was taking 1 to several trades per day based on a 5 min chart
  • #268
  • Apr 3, 2012 1:01pm Apr 3, 2012 1:01pm
Quoting FX4Newbies
Then the performance heads downhill.

Downhill indefinitely? For how many trades?

I think it is just as important, even more so, to qualify the "downhill" behaviour.

Will take it from there when the reply comes in .

In addition, it might be a better idea to focus on "qualifying" a successful system by the failure rate instead of by the winning rate. The lower the failure rate the better the system. Not sure how that would work...just a thought. Psychologically, if we want a system to work, we tend to...


Yes, downhill for how long... I was vague
I have a guide that got from FF, not sure which thread, that gave the % chance of having x number of consecutive losing trades, given a winning percentage. Don't know how numbers were arrived at but it was interesting ....
  • #269
  • Apr 3, 2012 1:05pm Apr 3, 2012 1:05pm
Personally I think testing a robust un-overoptimized system over 200 trades is a fair start, however there are guys on this forum that say at least 1000, 10,000+. They claim that anything less (presumably would mean) you're being fooled by randomness like a coincidental sub 1000 trade lucky streak...then wonder if those people are just being fooled by a 10,000 trade lucky streak haha, why not test it over 1 million trades, why not every chart since rice trading in Japan!

The way I see it is that if you have a clear defining edge that makes money, and holds up in the worst choppy periods you can find without having to change it, then that's pretty darn good.
  • #270
  • Apr 3, 2012 1:05pm Apr 3, 2012 1:05pm
Quoting PhilipK
thanks again - I'd like an opinion on one last thing - The trading educator I mentioned stated that one should test their strategy on at least 200 trades.

That seemed low to me in general and I was wondering if micmic, FX4Newbies or anyone else would care to give their opinion /experience on roughly what number of trades would be considered a good sample, if, say, the approach was taking 1 to several trades per day based on a 5 min chart

Depends on various things:

1) Are those 200 trades on different market conditions ? Because if they are all during a very bullish and volatile EURUSD month, the situation may change dramatically next month.
2) What is the risk:reward ratio ? Some systems may be fine with losing 20-30 trades in a row in order to catch a very big move. Other systems may have a very small take profit and a large stop loss. They may even win 200 trades in a row but a single loss may wipe out all profit.
3) The system may also have different behaviours in different pairs

I'd say that for certain volatility and trend conditions, and a r:r close to 1:1, and 1-2 pairs maximum, 200 trades should be ok. But the evaluation should be repeated when market conditions change. A recent evaluation will always be more valid than an older one.
  • #271
  • Apr 3, 2012 1:09pm Apr 3, 2012 1:09pm
Quoting jeanlouie
Personally I think testing a robust un-overoptimized system over 200 trades is a fair start, however there are guys on this forum that say at least 1000, 10,000+.
un-overoptimized is the key word here... curve-fitting waits just around the corner (or should I say curve ? )
  • #272
  • Apr 3, 2012 1:13pm Apr 3, 2012 1:13pm
thank you, be well!
  • #273
  • Apr 3, 2012 1:16pm Apr 3, 2012 1:16pm
Quoting jeanlouie
Personally I think testing a robust un-overoptimized system over 200 trades is a fair start, however there are guys on this forum that say at least 1000, 10,000+. They claim that anything less (presumably would mean) you're being fooled by randomness like a coincidental sub 1000 trade lucky streak...then wonder if those people are just being fooled by a 10,000 trade lucky streak haha, why not test it over 1 million trades, why not every chart since rice trading in Japan!

The way I see it is that if you have a clear defining edge that makes money,...
A common sense approach for sure. If these systems are supposed to have about 2 good years (I'm sure some last longer) and you test so many trades (10,000+), you might not have time left to receive the benefit. Never went through such an extensive back-test. I did mine for 90 days. Not sure how many trades it was. Anyway, it still works for me. 200 trades would be a decent start I'd imagine, but the more the merrier...better stats. Do as many as you need to do, until you feel as though you "know" what the system would give: what makes you feel comfortable.

I like your answer JeanLouie.

FX4
Trade what you see, not what you think or feel. Let the market come to you.
  • #274
  • Apr 3, 2012 1:52pm Apr 3, 2012 1:52pm
Quoting FX4Newbies
Hanover, here's a picture of the "Blackjack Beast".

FX4
Thanks for the link. Seems that it wasn't the Don Johnson from Miami Vice.

Quoting mister me
.......this was the interesting bit for me,he realised that overall casinos were showing declining profits over an 18 month period due to new establishments gaining licences. ......
Actually, what he did wasn't that clever, at least mathematically; it seems that he simply persuaded the casinos to offer him a positive expectancy game. Hence anybody who used a tight strategy and MM would eventually win, under those playing conditions. So either the casinos were gullible, or (as you say) they intended to use high rollers like him to promote the game, and rejuvenate their flagging business, and were willing to suffer some losses as some kind of marketing expense.

Obviously what made Mr Johnson's feat newsworthy was the amount of money that he won, and the fact that the casinos banned him. But (and as any savvy forex trader knows) if you're going to employ prudent MM, you'd need several times that amount of money to start with, and slowly grind out the fortune from your low risk per hand. Either that or he took undue risks and simply got lucky.

Agreed, it's a good parallel with the banking industry.

David
  • #275
  • Apr 3, 2012 2:19pm Apr 3, 2012 2:19pm
Quoting hanover
Thanks for the link. Seems that it wasn't the Don Johnson from Miami Vice.



Actually, what he did wasn't that clever, at least mathematically; it seems that he simply persuaded the casinos to offer him a positive expectancy game.

David
He did not persuade the casino to give him a positive expectancy game;

"When Johnson and the Trop finally agreed, he had whittled the house edge down to one-fourth of 1 percent, by his figuring." And the 20% loss discount only kicked in after $500k in loss.

On another note, I absolutely love hearing about several casinos losing millions on a single night!
  • #276
  • Apr 3, 2012 2:24pm Apr 3, 2012 2:24pm
Quoting jeanlouie
He did not persuade the casino to give him a positive expectancy game;
Then he was merely lucky.
  • #277
  • Apr 3, 2012 4:01pm Apr 3, 2012 4:01pm
Quoting PhilipK
Yes, downhill for how long... I was vague
I have a guide that got from FF, not sure which thread, that gave the % chance of having x number of consecutive losing trades, given a winning percentage. Don't know how numbers were arrived at but it was interesting ....
Philip,

The probability p of x consecutive losses within an n trade sequence, where a system has a win rate of w, can be calculated using the Excel formula
p = 1-BINOMDIST(0,n-x+1,(1-w)^x,FALSE)

The formula assumes that all events (trades) are independent events, i.e. that there is no serial correlation between wins and losses. Hence it's a binomial distribution.

David
  • #278
  • Apr 3, 2012 4:52pm Apr 3, 2012 4:52pm
Quoting hanover
Philip,

The probability p of x consecutive losses within an n trade sequence, where a system has a win rate of w, can be calculated using the Excel formula
p = 1-BINOMDIST(0,n-x+1,(1-w)^x,FALSE)

The formula assumes that all events (trades) are independent events, i.e. that there is no serial correlation between wins and losses. Hence it's a binomial distribution.

David
Hanover,

Glad to see the math/programming behind the "madness". What is the ^ in the formula...what is the instruction associated with it?

Also, would you please explain the difference between a robot (bot) and an EA that auto-trades. Maybe that info. is already in a thread somewhere. Any good leads?

Thanks,

FX4
Trade what you see, not what you think or feel. Let the market come to you.
  • #279
  • Apr 3, 2012 7:08pm Apr 3, 2012 7:08pm
Quoting FX4Newbies
Hanover,

Glad to see the math/programming behind the "madness". What is the ^ in the formula...what is the instruction associated with it?

Also, would you please explain the difference between a robot (bot) and an EA that auto-trades. Maybe that info. is already in a thread somewhere. Any good leads?

Thanks,

FX4
^ means "to the power of", i.e. exponentiation.

I'm not aware of any fundamental difference between a robot, an EA ("expert advisor") and an automated trader. Just different names for the same thing. Maybe somebody else has an answer?
  • #280
  • Apr 3, 2012 7:08pm Apr 3, 2012 7:08pm
Quoting hanover
Philip,

The probability p of x consecutive losses within an n trade sequence, where a system has a win rate of w, can be calculated using the Excel formula
p = 1-BINOMDIST(0,n-x+1,(1-w)^x,FALSE)

The formula assumes that all events (trades) are independent events, i.e. that there is no serial correlation between wins and losses. Hence it's a binomial distribution.

David

thanks for the information
0 traders viewing now