NexusFi: Find Your Edge


Home Menu

 





Coin Toss Experiment (Strategy)


Discussion in Psychology and Money Management

Updated
      Top Posters
    1. looks_one treydog999 with 21 posts (22 thanks)
    2. looks_two Fat Tails with 11 posts (45 thanks)
    3. looks_3 deaddog with 8 posts (0 thanks)
    4. looks_4 bryanzim with 7 posts (10 thanks)
      Best Posters
    1. looks_one Fat Tails with 4.1 thanks per post
    2. looks_two DavidHP with 2.5 thanks per post
    3. looks_3 trendisyourfriend with 2.4 thanks per post
    4. looks_4 treydog999 with 1 thanks per post
    1. trending_up 33,229 views
    2. thumb_up 127 thanks given
    3. group 27 followers
    1. forum 98 posts
    2. attach_file 8 attachments




 
Search this Thread

Coin Toss Experiment (Strategy)

  #91 (permalink)
 
Heiku's Avatar
 Heiku 
Frankfurt+Hessen/Germany
 
Experience: Beginner
Platform: NinjaTrader, Metastock
Broker: Interactive Brokers
Trading: Stocks, ETFs, ES
Posts: 5 since May 2013
Thanks Given: 24
Thanks Received: 2

Ah, thank you. This makes it clearer.

I read also the post by Fat Tails



I think my mistake is that I neglected that price has to move several ticks. If its 2 ticks target and 1 tick stop loss it has 50 % chance to move the first tick then 50 % chance to move the next. So it has a chance of 2 times 50 % = 25 % to reach the target whereas only 1 times 50 % to reach the stop - the probability to reach the target is only 25/75 =33 %.

The case in the Video is different. It is only one coinflip - the outcome stays 50:50.

uff, too much thinking

Reply With Quote
Thanked by:

Can you help answer these questions
from other members on NexusFi?
Online prop firm The Funded Trader (TFT) going under?
Traders Hideout
Are there any eval firms that allow you to sink to your …
Traders Hideout
Futures True Range Report
The Elite Circle
Ninja Mobile Trader VPS (ninjamobiletrader.com)
Trading Reviews and Vendors
Deepmoney LLM
Elite Quantitative GenAI/LLM
 
Best Threads (Most Thanked)
in the last 7 days on NexusFi
Get funded firms 2023/2024 - Any recommendations or word …
61 thanks
Funded Trader platforms
44 thanks
NexusFi site changelog and issues/problem reporting
24 thanks
GFIs1 1 DAX trade per day journal
22 thanks
The Program
19 thanks
  #92 (permalink)
sledmt
Kalispell montana
 
Posts: 158 since Jan 2015
Thanks Given: 157
Thanks Received: 23


treydog999 View Post
First this thread and result is regarding the coin toss experiment which is fundamentally a normal distribution. Second, if you add said filters to the coin toss, it doesn't matter because its an IID random variable. So no it would not shift this experiment at all.

Yes, everyone knows the market are not normally distributed. But in my experiments I was testing random entries vs trade management techniques to see if there was an inherent edge, there isn't. Which also confirms the academic literature. If you want to try it with filters on market data you can, but you can do that in a separate thread as it will no longer pertain to coin tosses. Or try to build it in a back tester to see the results.

Bit of a rabbit trail, but I was curious what you used for your back testing this theory? what application did you use?

Reply With Quote
  #93 (permalink)
 
Fat Tails's Avatar
 Fat Tails 
Berlin, Europe
Market Wizard
 
Experience: Advanced
Platform: NinjaTrader, MultiCharts
Broker: Interactive Brokers
Trading: Keyboard
Posts: 9,888 since Mar 2010
Thanks Given: 4,242
Thanks Received: 27,102



Heiku View Post
Ah, thank you. This makes it clearer.

I read also the post by Fat Tails



I think my mistake is that I neglected that price has to move several ticks. If its 2 ticks target and 1 tick stop loss it has 50 % chance to move the first tick then 50 % chance to move the next. So it has a chance of 2 times 50 % = 25 % to reach the target whereas only 1 times 50 % to reach the stop - the probability to reach the target is only 25/75 =33 %.

The case in the Video is different. It is only one coinflip - the outcome stays 50:50.

uff, too much thinking


Yes, thinking can be painful. This thread is full of thoughts.


Reply With Quote
Thanked by:
  #94 (permalink)
 
treydog999's Avatar
 treydog999 
seoul, Korea
 
Experience: Intermediate
Platform: Multicharts
Broker: CQG, DTN IQfeed
Trading: YM 6E
Posts: 897 since Jul 2012
Thanks Given: 291
Thanks Received: 1,039


sledmt View Post
Bit of a rabbit trail, but I was curious what you used for your back testing this theory? what application did you use?

I was using Multicharts with a random entry generator, I believe it was 6E as the underlying. I choose 6E because its relatively neutral drift wise, unlike any stock index futures or seasonal like commodities.

You could run the same experiment in R, but in multicharts it was just much faster and prettier reporting methods.

Reply With Quote
  #95 (permalink)
sledmt
Kalispell montana
 
Posts: 158 since Jan 2015
Thanks Given: 157
Thanks Received: 23


treydog999 View Post
I was using Multicharts with a random entry generator, I believe it was 6E as the underlying. I choose 6E because its relatively neutral drift wise, unlike any stock index futures or seasonal like commodities.

You could run the same experiment in R, but in multicharts it was just much faster and prettier reporting methods.

Thanks for the info. Never used Multicharts. Used R before. R can get the job done, but at a cost.

Reply With Quote
  #96 (permalink)
 
Heiku's Avatar
 Heiku 
Frankfurt+Hessen/Germany
 
Experience: Beginner
Platform: NinjaTrader, Metastock
Broker: Interactive Brokers
Trading: Stocks, ETFs, ES
Posts: 5 since May 2013
Thanks Given: 24
Thanks Received: 2


Fat Tails View Post
Yes, thinking can be painful. This thread is full of thoughts.


Haha, great so after some rest I will take more pain now :-)

I was listening to Futurestrader71 webinar on Risk again. If he had taken a bigger sample size in his statistical analysis he would finally have ended with an expectancy value of zero (or slightly negative because he took some slippage in his samples), right? No matter if 1:1, 2:1 or scaling whatever.

So what he claims that you could base a living on a toin coss is not true from a statistical point of view, correct? ( I do not intend to bash on FT, he is a really great guy and gives very valuable advice). I just want to get this point straight in the name of all newbies sitting at home tossing their coins ;-)

So when trade management does not provide an edge and it is commonly agreed that the entry is the least important thing so it does not provide an edge either - then where is the edge to be found?

Is it just the ability of the trader to take himself out of the game after some winners before his expectancy ultimately reverses to zero again? This sounds not like a very satisfying result. Is it the positive drift effect of some instruments like stock indizes combined with leverage? I think you can capitalize on that in the long term but there are successful traders in rotating sideways markets as well who constantly do well on that. Are these outliers?

Reply With Quote
  #97 (permalink)
 
trendisyourfriend's Avatar
 trendisyourfriend 
Quebec Canada
Market Wizard
 
Experience: Intermediate
Platform: NinjaTrader
Broker: AMP/CQG
Trading: ES, NQ, YM
Frequency: Daily
Duration: Minutes
Posts: 4,527 since Oct 2009
Thanks Given: 4,171
Thanks Received: 6,018


Heiku View Post
...

So when trade management does not provide an edge and it is commonly agreed that the entry is the least important thing so it does not provide an edge either - then where is the edge to be found?
...

You can find an edge by exploiting the mistakes of other traders and by keeping your risk under tight control.

Ask yourself constantly, Who just got screwed? Who just made a mistake? And realize that those people will likely provide fuel for the fire if the move reverses. Also, if the answer to who just made a mistake is "me", fix it before real damage is done. Finally, it is not true that the entry is the least important thing. Entry and exit goes hand in hand. You can't have an exit without an entry, there are both equally important. Why do you think people use tools like volume profile, vwap, footprint or support/resistance etc. it is to buy or sell contextually high or low to minimize the risk at the entry point.

Reply With Quote
Thanked by:
  #98 (permalink)
 
deaddog's Avatar
 deaddog 
Prince George BC Canada
Legendary Market Wizard
 
Experience: Advanced
Platform: National Bank Direct
Broker: NBD/BMO/Questrade
Trading: Stocks
Frequency: Every few days
Duration: Weeks
Posts: 1,283 since May 2013
Thanks Given: 183
Thanks Received: 3,962

I don’t have back testing capability.
This is strictly a trade management strategy;

Theory; We cannot predict what the market will do but if we cut losers short and let winner run we can make a profut.

One trade per day.
Pick a time frame; (5 min or15 min or 1 hour)
Pick a time with in the first couple of hours of the trading day.
Enter trade at the same time each day
Flip a coin: Heads = Long / Tails = Short.

After entry use the extremes of the entry bar as a stop.
As each bar forms move the stop to the extreme of the previous bar.
Flat at the end of the day.

I’d appreciate if anyone with back testing capabilities could test this theory.

"The days when I keep my gratitude higher than my expectations, I have really good days" RW Hubbard
Reply With Quote
  #99 (permalink)
 
eminijalapeno's Avatar
 eminijalapeno 
new york, new york
 
Experience: Advanced
Platform: thinkorswim(options PM) OEC, NT7, Jigsaw Tools (futures, Stocks, Options)
Broker: NT -CQG
Trading: Stocks, Options, Futures
Posts: 14 since Jan 2014
Thanks Given: 63
Thanks Received: 12

I have this believe that markets should be consider random because there is no way we can't read everyone's trading reasons, systems, etc.. Any trading entry/exit really is based on a 50/50 probability ( actually like 48.5 to 51.5 due to the free rate up drift), and because probabilities are a measure of trying to predict future success and failure based on either statistical data or some measure of volume or implied volatility..., again, number of occurrences in a big enough sample size can only provide a probability.. the problem for most of us... is that probability its all we got.... so then you have to measure what's the probability that probability theory works... loll.. moving on..

and at the same time we can (somehow) we decide to set a particular arrangement of rules that could help improving those probabilities, like tossing the coin at the beach instead of the basement, or tossing the coin with one hand while holding a drink in the other hand, or wearing a hat.. but we found that, that doesn't work..

One question I have entertained ( and started to figure as the years go by) is that when trying to do contract management (or life in general) is that nothing is free or that everything has a price or that the greater the risk the greater the probability of success...

Now, the idea is that when you have a long ES for 3 contracts per se.. and your stop is 3 points - your risk is 450 bucks.. now the moment you take the first contract out, something interesting happens.. you are no longer trading 3 contracts (genius), your risk in 3 point loss is now $300 (not including the $50 winner) and you have reduce your cost basis of the 2 remaining contracts which also can be stated as "you have reduced your average entry". So if you enter the trade with 2 contracts it would be as if you entered 50 bucks lower or 1/2 of point in those 2 contracts (So far so good?) Kind of like saying I had 3 contracts at 2000 but because I closed one at 2001 is like if I entered 2 at 1999.50

Now let's say you took the second contract at 3 points, now you have $50 of the first contract and $150 for the second contract, now you have 1 contract left that is open. So here is what we have... we have 1 contract left, that has $200 cushion and 1 open with an unrealized profit of $150 (total so far $200 realized +$150 unrealized.... meaning you need 2 more points in that 1 contract left to match your original 1:1 of $450 in 3 contracts).

So, If the trade was going to move all the way down to the original entry, (I guess we can still use 2000 as example) you still have those 200 bucks you collected in your 2 contracts.. (or 4 points of cushion) So before this 1 contract left can lose any money, it would have to drop $200 or 4 points for you to break even in the entire 3 contract trade.. which also means that your avg entry price for that 1 contract has a range of extra 4 points to the downside before you break even (hope so far I'm explaining this correctly-- I'm Hispanic so I write how I speak..

...if you are still with me...

So here is the first question I would ask.. are you managing 3 contracts in this trade? or managing 2 contracts? or are you managing one contract but risking 3 to be able to later manage that 1? The result will depend entirely in the probability of getting more that 2 points in the last contract, cause if you are making 350 but losing 450 in a 50/50 then we know how that ends.. that being said.. there would be a lot of "ifs" to consider.. Sometimes the trade goes in your direction to later turn around, so if it turns around and you need to get out, then you have ton of time, profits and space to make that decision, but if you have more % of losers than winners then... anyways.. I think that's a lot to consider... what do you guys think?

Follow me on Twitter Reply With Quote
Thanked by:




Last Updated on April 17, 2015


© 2024 NexusFi™, s.a., All Rights Reserved.
Av Ricardo J. Alfaro, Century Tower, Panama City, Panama, Ph: +507 833-9432 (Panama and Intl), +1 888-312-3001 (USA and Canada)
All information is for educational use only and is not investment advice. There is a substantial risk of loss in trading commodity futures, stocks, options and foreign exchange products. Past performance is not indicative of future results.
About Us - Contact Us - Site Rules, Acceptable Use, and Terms and Conditions - Privacy Policy - Downloads - Top
no new posts