Do you think statistics work in trading? | Page 5 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Statistics combined with system. Profitable? What do you think 71 replies

If you haven't made it yet, what makes you think you will? 212 replies

Do YOU think EA's work (POLL) 70 replies

Do you really think a broker will let you get away with that? 4 replies

Trader Statistics and Broker Statistics 0 replies

Attachments: Do you think statistics work in trading?
Exit Attachments

Do you think statistics work in trading?

  • #81
  • Sep 19, 2013 5:44am Sep 19, 2013 5:44am
I think it works. It's a game of probabilities.
Statistics can help you to find your % of profitable signals, e.g. 30% ---> Your R:R at least must be = 3:1 (TP:SL). One must put strict rules under the percentage of successful trades.
  • #82
  • Sep 19, 2013 6:52am Sep 19, 2013 6:52am
Quoting 21vs7
{quote} You said mathematics dont work. I said It works. Statistics is only a little sister of the mother Math.
I still don't understand what you mean. There must be some misunderstanding. Show me where I stated mathematics doesn't work with all the letters, cause I clearly do not remember making that statement. If I was a mathematician maybe I would make such a bold claim, but my math knowledge is actually quite limited. As a rule of thumb I avoid talking about things outside my realm of experience/expertise, especially using absolutes.
"To hold, you must first open your hand. Let go." - Lao Tzu
  • #83
  • Sep 19, 2013 7:08am Sep 19, 2013 7:08am
Quoting vox dei
{quote} I still don't understand what you mean. There must be some misunderstanding. Show me where I stated mathematics doesn't work with all the letters, cause I clearly do not remember making that statement. If I was a mathematician maybe I would make such a bold claim, but my math knowledge is actually quite limited. As a rule of thumb I avoid talking about things outside my realm of experience/expertise, especially using absolutes.
eeeh, I see I qouted you as my response to the other guy who said math is nothing. My baad
  • #84
  • Sep 19, 2013 7:21am Sep 19, 2013 7:21am
Quoting wookey
{quote} Forex is a playground of big banks.

Completely different situation is in Forex. It is dominated by big banks including central banks. Which more often than not do trades for other than simple trading purposes. They do not care that you have a perfect butterfly pattern and the market should go down.
There you have it. Forex is dominated by a relatively small number of large players. They determine the prices and accumulate or distribute large positions which takes time to complete. Therefore, since the big boys don't trade using the same methods as small players e.g. TA, the type of statistics that could prove useful will be those that can identify what the big boys are doing. Studying statistics on chart patterns is a waste of time since none have proven to be reliable i.e. the cause and effect relationship is poor for all chart patterns.

Focus on the behaviour of the big boys (e.g. Commitments of traders report) and then follow them or risk being trampled by the elephants in the market. They trade long-term positions and maybe so should you. Remember they are still around while 95% of small guys have gone broke so just maybe they have a superior method?
  • #85
  • Sep 19, 2013 7:37am Sep 19, 2013 7:37am
Quoting 21vs7
{quote} eeeh, I see I qouted you as my response to the other guy who said math is nothing. My baad
Now, everything falls into place. All good bro
"To hold, you must first open your hand. Let go." - Lao Tzu
  • #86
  • May 13, 2014 10:07am May 13, 2014 10:07am
What do you think of the statistics from this article?

http://samuraitradingacademy.com/7-b...tion-patterns/

Anyone knows where I can find other statistics for traditional patterns to compare the results?
  • #87
  • May 13, 2014 10:39am May 13, 2014 10:39am
Statistic will not work in the long term because they can't factor in the unexpected.
They help... but they will not protect you from the "black swan" event, the uncertain thing the burns the account.

If you are a mathematician you might enjoy these books: "Fooled by Randomness" and "The Black Swan" by Nassim Taleb.

Mr Taleb is an author, a long term wall street survivor, and very much into mathematics and statistics.
His books will have you reasoning on why statistic can not be trusted blindly.
  • #88
  • May 14, 2014 7:27pm May 14, 2014 7:27pm
Does statistics work ? Very good question.

Answer : Short term yes . Long term No.

You must have an edge to survive.

If statistics really worked, more than 50% of traders would be profitable consistently. Unfortunately its not so.
  • #89
  • May 14, 2014 7:40pm May 14, 2014 7:40pm
You guys are spending FAR too much time analysing the market and not enough time trading it!!
Analysing is looking back, trading is looking forward. You make your money looking forward.
Are you here for the mental exercise or to make money. You decide.
Develop a system, develop a strategy to trade that system and then start looking forward.
Your account will love you.
Cheers, H
  • #90
  • May 14, 2014 8:21pm May 14, 2014 8:21pm
Quoting Aich
You guys are spending FAR too much time analysing the market and not enough time trading it!! Analysing is looking back, trading is looking forward. You make your money looking forward. Are you here for the mental exercise or to make money. You decide. Develop a system, develop a strategy to trade that system and then start looking forward. Your account will love you. Cheers, H
How is that at all helpful? In one line you tell people that they need to "develop a system" and in the next that they need to "look forward". How does one develop a strategy by only watching the hard right edge of price? Learning from the past is critical, and a thorough analysis and study of past price action is key to building any understanding of the market.
  • #91
  • May 14, 2014 8:33pm May 14, 2014 8:33pm
Quoting Kanzler
{quote} ...... Learning from the past is critical, and a thorough analysis and study of past price action is key to building any understanding of the market.
Absolutely.
But there is a point when you must stop looking back and start trusting the knowledge that you have developed.
That is when you become a trader and stop being an analyst.
That is when you will start making money - if your intelligence is good enough.
Trading is all happening in your mind. If you believe that you are a mathematician (as the first post says) then you need to stop being a mathematician and become a trader. mathematicians analyse, traders trust their system, have a strategy and act.
Your trading strategy is much more important than your system. Please do not ask me to explain "strategy" - that is a very big topic.
It encompasses risk/reward, trading times, attitude to risk, choosing currency pairs, conserving capital, compounding, exposure,trading thru announcements, market activity, observing price action/activity, etc.
Have a great day.
Cheers, H
  • #92
  • May 14, 2014 8:41pm May 14, 2014 8:41pm
Quoting Aich
{quote} Absolutely. But there is a point when you must stop looking back and start trusting the knowledge that you have developed. That is when you become a trader and stop being an analyst. That is when you will start making money - if your intelligence is good enough. Trading is all happening in your mind. If you believe that you are a mathematician (as the first post says) then you need to stop being a mathematician and become a trader. mathematicians analyse, traders trust their system, have a strategy and act. Your trading strategy is much...
I'm not trying to pick on you or anything like that, but I still have a few things to say here. Intelligence has little to do with it. I know there are many very intelligent people, with IQs probably in the 140+ range, who would never get this game without the help of others. I certainly wouldn't have.

Analysis and study of past market moves is a great way to avoid losing your shirt. And you can't trust a strategy you haven't tested before, usually with historical data to start with. Forward testing can be done after you've taken a look at how your ideas fared in past markets, but just jumping in without historical study is just gambling. The mindset between being an analyst, mathematician and a trader is rather irrelevant when you're still learning.
  • #93
  • May 14, 2014 8:54pm May 14, 2014 8:54pm
Quoting Kanzler
{quote} ...... The mindset between being an analyst, mathematician and a trader is rather irrelevant when you're still learning.
We do not disagree. It is simply a matter of self belief. If you view yourself as an analyst you will always be looking back. If you view yourself as a trader you will have done all of your analysis and be looking forward to placing appropriate trades.

You can analyse the market till the cows come home. It will not make you any money. Only committing to live trades will test your knowledge and, over time, make you money.
If you are in this business as an intellectual exercise, you are in it for the wrong reasons.
This is a high risk business and the only reason to be in a high risk business is to make high returns.

You are always learning in this business. One of the most powerful lessons I got was to "enjoy your mistakes".
This is all about
1) acknowledging that it was YOUR mistake.
2) analysing how it occurred.
3) putting measures in place to make sure that it does not happen again
4) saying to yourself "now I do not need to make that mistake again"
Cheers
  • #94
  • May 14, 2014 8:55pm May 14, 2014 8:55pm
Quoting Aich
{quote} We do not disagree. It is simply a matter of self belief. If you view yourself as an analyst you will always be looking back. If you view yourself as a trader you will have done all of your analyst and be looking forward to placing appropriate trades. You can analyse the market till the cows come home. It will not make you any money. Only committing to live trades will test your knowledge and, over time, make you money. If you are in this business as an intellectual exercise, you are in it for the wrong reasons. This is a high risk business...
Fair enough.
  • #95
  • Edited May 15, 2014 1:03am May 14, 2014 10:04pm | Edited May 15, 2014 1:03am
Options trading is built on statistical analysis. Options pricing models rely upon the "implied volatility" of options prices - that is, upon an extrapolation of price based on prior prices and the relationship between prior prices over time, etc. So, the fundamental thing that one trades - the price of an options contract - is itself a (derivative) statistical construct. Not surprising then that many options trading strategies are purely statistical approaches to the problem. And it is possible to make money trading options, of course. So, in the context of options trading, statistics does indeed work in trading.
"If The Fool persists in his Folly he will become wise." - William Blake
  • #96
  • May 15, 2014 3:19am May 15, 2014 3:19am
Quoting str8
What do you think of the statistics from this article? http://samuraitradingacademy.com/7-b...tion-patterns/ Anyone knows where I can find other statistics for traditional patterns to compare the results?
Quote
The head and shoulders patterns are statistically the most accurate of the price action patterns, reaching their projected target almost 85% of the time.

Really? well retirement here I come.....



Here is a study: http://cfr.statslab.cam.ac.uk/public.../H&S-part2.pdf
  • #97
  • May 15, 2014 3:26am May 15, 2014 3:26am
Imo, some methods of technical analysis really work but they just applied incorrectly, under wrong circumstances, etc. I think we have to interprete whole market operation with math, only in this case it will yield results.
  • #98
  • Edited 7:59am May 15, 2014 5:00am | Edited 7:59am
Quoting Quidsey
{quote} {quote} Really? well retirement here I come..... Here is a study: http://cfr.statslab.cam.ac.uk/public.../H&S-part2.pdf


Personally I have ~80% win rate with ~1:1 RR trading traditional patterns. Low-frequency trader.

But I don't look for the completion of patterns. I usually take full profit at 1:1 RR even though the pattern is far from being completed.

I also look for space. If I don't have space I don't take the trade usually.

But while I'm happy with my results I don't know how much profit I give up by only taking profit at 1:1 RR. Or by don't taking patterns without space.

In the encyclopedia of chart patterns 2nd edition it says that head and shoulders bottoms has 3% break even failure for bull markets and 4% break even failure for bear markets (74% meeting price target and 58% meeting price target).
Head and shoulders tops 4% break even failure and 1% break even failure (bull and bear markets), and they meet price target 55% and 56% (bull and bear markets).

The break even failure is so low that it's almost hard to believe to be true. What do others think about this book? I know the patterns don't have a fixed probability to win and the statistics always change but I can have an idea about the profitability of patterns. The statistics from that book are only from stocks but I trade only forex. I think a forex chart has other dynamics than a stock chart and stock statistics aren't so relevant if you trade forex.

Thanks for the link. It seems their trade management is very different from mine. I enter only after a nice close outside the breakout level. It's not enough for me to see price tries to break the neckline if I don't see a close below it. And I want a nice close, I don't want a big whick. If I watch a daily head and shoulders I won't go to 4hr or 1hr, I want a close on daily. The results from that study are from all time frames but many say small time frames are not so reliable as daily (don't know if this is true, I mainly trade daily and weekly). I prefer the neckline to be an horizontal level or a descending slope line or an ascending slope line with better angle because if it's an ascending trend line with small angle like those from the study it can have trouble to fall lower. As for exit strategies I never use MT4 trailing stops or don't take "time" into consideration. There are many factors that can make 2 studies completely different. The encyclopedia book used "close" for entry signals. I'm mainly interested in break even failure rates (because I like to take profit quick and a pattern with low failure rate can have a big win rate if you take profit quick) and how often the pattern is fully completed (to see if maybe it's better to let the trade run not to take profit quick).

To be honest I think what it really matters is the neckline from head and shoulders (and probably this is another reason I don't like to wait for the completion of the pattern, usually it has a lot of trouble areas to break in order to complete). It's just a coincidence that it's also head and shoulders but what is really important is the horizontal or sloping level. Usually the neckline is a solid level (I also trade breakouts of important levels even if it's not a pattern). Then space. This is if I use logic. I don't see any logic in why the head and shoulders would be important. Same thing with the other patterns. In triangles is really the triangle important or it's the horizontal or sloping level? How about channels, flags, wedges, rectangles? Of course I may be wrong, maybe a level that it's part of a pattern really works better than a level that isn't part of a pattern and this is where statistics are useful.
  • #99
  • May 15, 2014 5:32am May 15, 2014 5:32am
Quoting Aich
You guys are spending FAR too much time analysing the market and not enough time trading it!! Analysing is looking back, trading is looking forward. You make your money looking forward. Are you here for the mental exercise or to make money. You decide. Develop a system, develop a strategy to trade that system and then start looking forward. Your account will love you. Cheers, H
I think traders are spending far too little time analyzing the market and most trade based on faith. Faith will not be giving you any profit if the strategy isn't statistical profitable. How can you know a trader makes money? If he has a profit after a huge sample of trades, after a lot of years then he is making money. His statistics prove this. If he has a profit after a few trades with no big sample it can only be variance. In fact he could be a loser. Most of books and forums and training sites have no statistics. How can I trust a trading idea or strategy if I don't have proofs about the results on a big sample? An answer is with logic. But what if the strategy is profitable but the logic behind it is counter intuitive? Then I really need statistics with a good sample otherwise I would think the strategy doesn't win.
  • #100
  • Edited 5:53am May 15, 2014 5:35am | Edited 5:53am
Quoting str8
{quote} I think traders are spending far too little time analyzing the market and most trade based on faith. Faith will not be giving you any profit if the strategy isn't statistical profitable. How can you know a trader makes money? If he has a profit after a huge sample of trades, after a lot of years then he is making money. His statistics prove this. If he has a profit after a few trades with no big sample it can only be variance. In fact he could be a loser. Most of books and forums and training sites have no statistics. How can I trust a trading...
The assumption you're making is that the strategy is a static entity... what if the parameters of the strategy or perhaps even the strategy itself is constantly evolving as the market evolves?

No strategy is profitable, only the trader is.

Statistics are useful to analyse the trader's behaviour and outcomes of his behaviour - to help inform himself how he can make course corrections to improve his outcomes in an objective fashion.

Did you ever hear businesses talk about strategy as if the organisation's execution of the strategy doesn't matter or as if changing external conditions doesn't matter? It's weird that when it comes to trading we seem to think that everything can be predictable and made systematic when in fact a lot of it is dynamic and changing all the time... Trading is in the real world, despite how much it might look like a computer simulation/ game, and in the real world these things matter equally much as strategy.

The real challenge in trading is not in having a good plan - that's only the first step - it's how you execute it and how you make course corrections to your plan when things go wrong (without compromising the core values of your 'business' and your personal values). But many struggle with later steps and they blame it on the strategy - which in some cases is valid, but then again no strategy is perfect or will remain perfect.

Take for example the yen trade from 2012-2013. The context matters hugely - if you can assess the context well and apply the right strategy to solve the problem, then voila you will be profitable. Most traders are trying to fit a square peg into a round hole with their blindfolds on and they are complaining why it doesn't work!
Build good relationships with others.
0 traders viewing now