NexusFi: Find Your Edge


Home Menu

 





Coin Flip - bias of an edge


Discussion in Traders Hideout

Updated
      Top Posters
    1. looks_one baruchs with 6 posts (2 thanks)
    2. looks_two Yukoner with 5 posts (5 thanks)
    3. looks_3 Cashish with 5 posts (7 thanks)
    4. looks_4 Itchymoku with 3 posts (5 thanks)
      Best Posters
    1. looks_one shodson with 4 thanks per post
    2. looks_two Itchymoku with 1.7 thanks per post
    3. looks_3 Cashish with 1.4 thanks per post
    4. looks_4 Yukoner with 1 thanks per post
    1. trending_up 16,349 views
    2. thumb_up 46 thanks given
    3. group 14 followers
    1. forum 33 posts
    2. attach_file 7 attachments




 
Search this Thread

Coin Flip - bias of an edge

  #11 (permalink)
 
Cashish's Avatar
 Cashish 
Miami FL USA
 
Experience: Advanced
Platform: Ensign 10, NT7 DOM
Broker: IB, IQ
Trading: Currency Futures
Posts: 802 since May 2011
Thanks Given: 811
Thanks Received: 2,102


SMCJB View Post
The timing may not be random, but it's a random entry because the defined entry criteria is dependent upon a random coin flip.


OK, I'll give you that (for now) but let's agree to disagree and honor each other's opinions, move this conversation forward and see where it goes. For those who don't know, I have a bad habit of writing lengthily and often disjointed posts, consider this your warning.




If I know the outcome of a coin toss is either going to be heads or tails I can predict the outcome 100% of the time with 100% accuracy, the coin is either going to land on heads or tails. This is information risk, and knowing all probable outcomes of the flip of a coin can be reduced to no more than two (2) outcomes, I have no information risk, I know for an absolute fact, the coin is going to be either heads or tails. So, by (us) agreeing there are only two possible outcomes of a coin toss, all randomness of this exercise is removed, (at this level).

You might be wondering what nonsense Cashish is going to pull out his ass next, well how about a pair of dice! Although the exact final outcome of the flip of a coin only has two (2) possibilities, the roll of a pair of dice has eleven (11) possibilities, this again is the information risk. I hope we can agree predicting the exact final outcome of 11 pieces of information is much harder than predicting the exact outcome of 2 pieces of information. But the fact is, in both examples (coin and dice) we are well aware ahead of time the exact final outcome will be random as you pointed out, but that random outcome will also fall within a predefined range.


Random


IMO, on it's own, the whole notion of entering a trade long or short based on a random entry, e.g. coin flip, makes for good conversation but until a specific purpose (for the exercise) is qualified, it could be viewed as next to meaningless. Let's say at 10am each morning for a week I call up my broker, flip a coin and enter a trade based on the outcome of the "flip." THEN I open my software and see exactly where my position is located in the market. Can we agree this entry would create a much different psychological response than the same position entered after analyzing the prior price movement for even a few minutes prior to placing the order, I think so. Historically, the random outcome of a coin toss since it only has two possibilities was used to settle a dispute, or simply force a decision between to parties. Heads, someone wins, tails, someone loses.

I believe the theory of a random entry fits well with analyzing our trading skill or trading performance for the sole reason as traders we have but two choices, we can either Buy or Sell. To remain clear, IMO, the OP is asking, can you manage yourself OUT of a series of coin flip entries and be profitable. Given the fact, the direction of entry is beyond our control, any participant in the exercise will be left with a single action to manage the outcome of each trade, the exit.

OK, let's assume we're in the trade based on a decision we had no control over, (random entry) and we're left with the single decision to exit the trade to achieve a profitable outcome. According to the guidelines posted in the OP we basically have the free will to use any and all trading tools we might have at our disposal, we also have Time. Yes we could "play the system" and only close out the trade if/when it becomes profitable, whether that's a minute, hour, day, week or month from now (you get the idea). Or we could take the exercise seriously and actually use the opportunity to sharpen our skills, and maybe, maybe, maybe learn something along the way.

Let's say we called the broker at the given time and placed the trade "blindly" without actually knowing where in the context of the market our order was being placed. THEN, we open our charting program and trading tool kit and begin analyzing the location of the position. Oh great!, I entered at 10am, I have a statistical analysis spreadsheet on over 1000 trades that reveals that a trade taken at 10am will be profitable on the close 87% of the time, this exercise is going to be a slam dunk, @Yukoner is going to shit a solid gold brick when he sees what I've brought to the party! No wait!, I also have a Fib Line running right through my entry, and this trade is proven to work 89% of the time and these trades on average "pay out" 25 ticks.

Do I need to go on? I hope most who are actually reading this already know where this post is going. The offsetting factor of a random entry is the random distribution of wins and losses of any trading system. No matter how you slice or dice either of the two fore mentioned "trade signals" even with win percentages in the high 80s, the outcome of the very next trade has a 50/50 chance of being a winner or loser. The point is, even if, the random entry couldn't have aligned itself more perfectly with the best signal in your trading tool kit, the random distribution of wins and losses still exists, therefore, the real exercise as defined by the OP is Controlling Risk.

Just as the range of possible outcomes of a coin toss is limited to the "sides of the coin" the same is true with a trade, the range of possible outcomes are limited by the random distribution of a win or a loss. Therefore, to achieve profitability on any consistent level, we (or at least our Trading Accouts) have to "survive" this Random Distribution, yes, it's personal, and yes it's relative, but I believe it is the absolute minimum prerequisite to becoming a consistently profitable trader. Risk first, anything else and everything else second.

Visit my NexusFi Trade Journal Reply With Quote
Thanked by:

Can you help answer these questions
from other members on NexusFi?
My NT8 Volume Profile Split by Asian/Euro/Open
NinjaTrader
Better Renko Gaps
The Elite Circle
NexusFi Journal Challenge - April 2024
Feedback and Announcements
The space time continuum and the dynamics of a financial …
Emini and Emicro Index
New Micros: Ultra 10-Year & Ultra T-Bond -- Live Now
Treasury Notes and Bonds
 
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
39 thanks
NexusFi site changelog and issues/problem reporting
26 thanks
GFIs1 1 DAX trade per day journal
18 thanks
The Program
18 thanks
  #12 (permalink)
 
DavidHP's Avatar
 DavidHP 
Isla Mujeres, MX
Legendary Market Wizard
 
Experience: Advanced
Platform: NinjaTrader
Broker: Ninjatrader / Optimus Futures / AmpFutures
Trading: ES / 6E / 6B / CL
Frequency: Every few days
Duration: Minutes
Posts: 1,609 since Aug 2009
Thanks Given: 11,327
Thanks Received: 2,743

You may find this useful...
Or not.


Rejoice in the Thunderstorms of Life . . .
Knowing it's not about Clouds or Wind. . .
But Learning to Dance in the Rain ! ! !
Follow me on Twitter Reply With Quote
Thanked by:
  #13 (permalink)
 Yukoner 
Vancouver Island, Canada
 
Experience: Advanced
Platform: T4, IB, TD Ameritrade
Broker: Stage5
Trading: ES CL RTY Stocks
Posts: 58 since Nov 2013
Thanks Given: 163
Thanks Received: 130



Cashish

Just as the range of possible outcomes of a coin toss is limited to the "sides of the coin" the same is true with a trade, the range of possible outcomes are limited by the random distribution of a win or a loss. Therefore, to achieve profitability on any consistent level, we (or at least our Trading Accouts) have to "survive" this Random Distribution, yes, it's personal, and yes it's relative, but I believe it is the absolute minimum prerequisite to becoming a consistently profitable trader. Risk first, anything else and everything else second.

Very well said. Those who might read this... If you are trading, and don't understand this post, reread it... study it... define it... and then trade.

Understanding random distribution helps to take away the emotional pain of a loss. Your focus becomes on the risk, the method, the plan... and not so much that particular trade. A book could be written on this topic alone.

This thread has me thinking more and more about the implications of random distribution on our trading psychology, and defining what exactly we can control as traders. Thanks Cashish for contributing.

Follow me on Twitter Started this thread Reply With Quote
Thanked by:
  #14 (permalink)
 baruchs 
Israel
 
Experience: Intermediate
Platform: NinjaTrader
Broker: pfg
Trading: eminis
Posts: 323 since Jun 2009


Quoting 
Do I need to go on? I hope most who are actually reading this already know where this post is going. The offsetting factor of a random entry is the random distribution of wins and losses of any trading system. No matter how you slice or dice either of the two fore mentioned "trade signals" even with win percentages in the high 80s, the outcome of the very next trade has a 50/50 chance of being a winner or loser. The point is, even if, the random entry couldn't have aligned itself more perfectly with the best signal in your trading tool kit, the random distribution of wins and losses still exists, therefore, the real exercise as defined by the OP is Controlling Risk.

Really?
How is it possible to have a system with 80% win rate and then believe that the chance of a next trade is 50/50?
If the next trade has a 50/50 chance and a trade after has also 50/50 and after it too, doesn't it be added to a strategy with 50% win rate?
Sure on each trade you have two outcomes: win or loose, but a probability for each outcome is based on your strategy probability.

Reply With Quote
  #15 (permalink)
 
Cashish's Avatar
 Cashish 
Miami FL USA
 
Experience: Advanced
Platform: Ensign 10, NT7 DOM
Broker: IB, IQ
Trading: Currency Futures
Posts: 802 since May 2011
Thanks Given: 811
Thanks Received: 2,102


baruchs View Post
Really?
How is it possible to have a system with 80% win rate and then believe that the chance of a next trade is 50/50?
If the next trade has a 50/50 chance and a trade after has also 50/50 and after it too, doesn't it be added to a strategy with 50% win rate?
Sure on each trade you have two outcomes: win or loose, but a probability for each outcome is based on your strategy probability.


Thanks @baruchs for continuing to participate in the conversation and Thanks @Yukoner for the opportunity. I'll take another shot at this, but first I hope we can all agree that an entry based on the flip of a coin is, without doubt, a Random Entry, and now, the focus of conversation has shifted to the randomness of profitable trades. I'm a discretionary trader and although my trading decisions are mostly based on a few studies/indicators, there are times when I intentionally initiate random entries, I call it sending in a solider. I believe a trader can usually make a case for an entry in either direction in the market at any given time. I also believe most traders don't do this or allow themselves to think along these lines because their "thinking" is restricted or restrained by the time frame they most generally trade, or they're most comfortable trading. So just to touch back on the notion of a random entry, I don't believe a trader using a 5m chart with an average holding time of less than 40 minutes is going to allow himself to be able to see the same trading opportunities of the same random entry that a 2 or 3 day position trader would see. Don't get me wrong, I'm not suggesting flip flopping between time frames here, I'm just pointing out there are countless available participants willing to participate in the market on countless available time frames. But the one thing they all have in common is, if they don't protect their trading capital, they won't be in the market for long.


baruchs View Post
How is it possible to have a system with 80% win rate and then believe that the chance of a next trade is 50/50?

The short answer is, the random distribution of wins and losses of any trading system. In order for any calculation to arrive at a percentage of winning trades verses losing trades you and I are going to have to agree we need a sample size. Using your example of 80% let's assume the sample size used to arrive at this percentage value (%) is 100 trades. Therefore, we must also agree that during the testing period of those 100 trades, eighty (80%) of those trades were in fact winning trades. Given the fact "numbers never lie," and since we are in agreement that eighty (80%) of the trades in our testing period were winning trades, we have to agree twenty (20%) were losing trades. Savvy traders know I'm getting ahead of myself here but I'm going to continue, only to make a point.

If, we had an actual data sample of these 100 trades that generated these actual win/loss percentages we would also know exactly (within our data set) how many consecutive winners we had and how many consecutive losers we had, based on the data sample. Yes, the probability of the system generating 80% winning trades is proven over the 100 trades in the sample, and we also know how many consecutive wins and loses were generated in the sample, but what we don't know is WHEN they will occur. This is the random distribution of wins and losses of any trading system. Since, all trades have only two possible outcomes, win or lose, the NEXT TRADE in any series of trades, generated by any trading system, regardless of win percentage, has a 50/50 chance of being a winning trade or a losing trade, PERIOD.



baruchs View Post
Sure on each trade you have two outcomes: win or loose, but a probability for each outcome is based on your strategy probability.

This is where we kind-of disagree (give me liberties here). You project the opinion it's Black, the outcome of every trade is based only on the "win percentage." And I believe you also project the opinion that I believe it's White, the outcome of every trade is based only on the "50/50 chance of being a winning trade or a losing trade." But I believe where our opinions on this issue differ is in these two words, "each outcome." I believe "each outcome," of any individual trade must first survive the random distribution of wins and loses, then it earns it's rightful place within the percentage of winning trades or losing trades, within the series of trades defined by the "system."

Funny thing is, I agree with what you wrote in your post. If I had a system with a 80% win ratio, I'd bang it every time. BUT, I also have to take full responsibility for the 20% of the time the trade is going to fail.

I wrote this earlier, "Savvy traders know I'm getting ahead of myself here but I'm going to continue, only to make a point." I hope I was able to truly set forth my views of why I believe the out come of the next trade is 50/50. The whole point of this post (thread) was to ask the question and get opinions on managing yourself OUT of a series of coin flip entries and be profitable. My opinion is Yes, but the only way to achieve profitability is to control the risk on every trade. I want to offer one more example of the notion of the random distribution of wins and loses. In the example above using the same win/loss percentages based on 100 trades, we are assuming there is a system that generated these ratios. That system (even a hypothetical system) has to be complete with entry, target and stop loss criteria. It also has to be completely mechanical to generate meaningful statistics. For the purpose of this example let's assume this system trades 24 hours a day, and on average over the 100 trades in our sample the average hold time for wins and loses is also 24 hours. Although the following is hypothetical and clearly improbable, we have to consider it as possible, for two reasons, anything can happen and past performance is not indicative of future results.

We take our system with 80% winning trades out of the box, polish it off, plug it in, sit back and start making a list of all the new "toys" we're going to buy when the money starts rolling in! The first day produces a losing trade, so does the second, the third and forth. We have untold amounts of money so we let the system run. In the first 20 days of running the system we have 20 loses. OK, over the period of 100 trades (1 trade per day) the system is actually preforming perfectly. So, since we have untold amounts of money and we know out of 100 trades we are going to see 80 winning trades we "stay the course." We stay the course for ANOTHER 20 days and every one of them is a loser! 40 trades in a row and ALL losers, we pull the plug, quit trading the system and give it to our Brother-in-law and tell him it has a 80% win rate. 160 days later we see our brother-in-law at his kids graduation party, he takes you out to his garage and shows you his new boat! He can't quit thanking you for the trading system, he says he hasn't had a losing day in 160 trading days. As you drive home you realize the system is preforming perfectly, still generating 80% winning trades over data sets of 100 trades. That, is random distribution.

Visit my NexusFi Trade Journal Reply With Quote
Thanked by:
  #16 (permalink)
 
Big Mike's Avatar
 Big Mike 
Manta, Ecuador
Site Administrator
Developer
Swing Trader
 
Experience: Advanced
Platform: Custom solution
Broker: IBKR
Trading: Stocks & Futures
Frequency: Every few days
Duration: Weeks
Posts: 50,396 since Jun 2009
Thanks Given: 33,172
Thanks Received: 101,536


baruchs View Post
Really?
How is it possible to have a system with 80% win rate and then believe that the chance of a next trade is 50/50?
If the next trade has a 50/50 chance and a trade after has also 50/50 and after it too, doesn't it be added to a strategy with 50% win rate?
Sure on each trade you have two outcomes: win or loose, but a probability for each outcome is based on your strategy probability.

A system win rate is over a series of trades greater than 1. For the very next trade, a series of 1, the odds are 50/50 if you include breakeven as a loss.

Mike

We're here to help: just ask the community or contact our Help Desk

Quick Links: Change your Username or Register as a Vendor
Searching for trading reviews? Review this list
Lifetime Elite Membership: Sign-up for only $149 USD
Exclusive money saving offers from our Site Sponsors: Browse Offers
Report problems with the site: Using the NexusFi changelog thread
Follow me on Twitter Visit my NexusFi Trade Journal Reply With Quote
Thanked by:
  #17 (permalink)
 
shodson's Avatar
 shodson 
OC, California, USA
Quantoholic
 
Experience: Advanced
Platform: IB/TWS, NinjaTrader, ToS
Broker: IB, ToS, Kinetick
Trading: stocks, options, futures, VIX
Posts: 1,976 since Jun 2009
Thanks Given: 533
Thanks Received: 3,709

I have found that studying coin flipping strategies can lead to interesting discoveries in the nature of the market you are studying. For example, I learned that on CL random BUY signals in the first hour of RTH is often winning strategy over time, using a 3*ATR(14) trailing stop as the only exit rule. Then you can look at that data and say "is this just a data-mined coincidence, or is there some potential explanation for this?" Then you can form a hypothesis and reasons which can be debated.

Hypothesis: CL tends to favor the long side in the first hour of RTH trading

Potential explanations:
  1. Commodities have intrinsic value and are subject to supply and demands of the physical world. It's not the same as paper-derived assets. This puts a floor on prices and buyers will always step in when prices get too low.
  2. CL tends to panic up better than panic down due to geopolitical surprises, hurricanes, etc. which cause demand spikes
  3. Since CL is priced in dollars, and the dollar, over time is constantly weakening due to dollar-devaluing federal reserve policies, it takes more dollars to buy oil, which gives CL a long-term upward inflationary drift.
  4. If you're an RTH position trader in CL and you need to get long at the "open", which is more often the case than not, then, as a day trader you're better off looking for long signals during 1st hour of RTH.

Add in some regime filters and other things to improve performance.

The point is that this all started studying flipping coins. I have found other gems from studying random strategies in certain markets during certain times of day.

Attached Thumbnails
Click image for larger version

Name:	chart.png
Views:	208
Size:	46.4 KB
ID:	131072   Click image for larger version

Name:	equity.png
Views:	181
Size:	43.0 KB
ID:	131073   Click image for larger version

Name:	hours.png
Views:	184
Size:	13.8 KB
ID:	131074   Click image for larger version

Name:	days.png
Views:	186
Size:	14.1 KB
ID:	131075   Click image for larger version

Name:	chart2.png
Views:	190
Size:	48.8 KB
ID:	131076  
Follow me on Twitter Visit my NexusFi Trade Journal Reply With Quote
  #18 (permalink)
 baruchs 
Israel
 
Experience: Intermediate
Platform: NinjaTrader
Broker: pfg
Trading: eminis
Posts: 323 since Jun 2009


Quoting 
A system win rate is over a series of trades greater than 1. For the very next trade, a series of 1, the odds are 50/50 if you include breakeven as a loss.

Mike


Quoting 
Since, all trades have only two possible outcomes, win or lose, the NEXT TRADE in any series of trades, generated by any trading system, regardless of win percentage, has a 50/50 chance of being a winning trade or a losing trade, PERIOD.

Randomness only says that in an experiment (system) with 80% win rate, you don't know which will be profitable and witch losing. So you need to take them all.

Let me try it one ,ore time.
I admit that I don't know anything about aviation safety statistics, but lets suppose that every 1M flights there is a crash.
You and I want to take the next flight. I will take it, because I know that there is a chance of a crash, but its a very small one 1:1M.
Will you take the next flight? lol

p.s.
Remember we want to take the same flight.

p.p.s.

A blond took a flight to Paris. She was sited next to a fat and smelly man. After a wile she couldn't stand it any more and demanded to be sited else ware. The flight attendant tried to explain that all the seats in economy are taken. The blond demanded to be sited in business then. The flight attendant tried to explain to her that its not possible, but the blond insisted and asked to speak to the captain.
The flight attended reluctantly went to the captain and explained the situation. The captain leaned to the blond, whispered few words and the blond stopped complaining. Later the flight attendant asked the captain, what did he say?
The captain replied that he said her that the front row don't fly to Paris.

I'm not blond. Are you?

Reply With Quote
  #19 (permalink)
 GFIs1 
who cares
Legendary Market Wizard
 
Experience: None
Platform: nobody interested
Broker: none
Trading: forget about it
Posts: 6,921 since Feb 2012
Thanks Given: 6,181
Thanks Received: 15,562

Statistical point of view:
The outcome is 50/50 over the long run: >10'000 rounds

Economic point of view: COSTS for
- investment of personal time for trading
- trading fees, platform, internet, broker, spreads etc.
- inflation
- risk of losing when investing in different currency
Alternative gains
- when investing the same amount elsewhere
- if putting the money under the bed (not getting any interests)

Result: To stay break even or on the winning side you need
to have a much better outcome than 50/50 to invest in trades.

GFIs1

Follow me on Twitter Visit my NexusFi Trade Journal Reply With Quote
Thanked by:
  #20 (permalink)
 
podski's Avatar
 podski 
Belgium
 
Experience: Intermediate
Platform: Ninjatrader
Broker: Ninjatrader Brokerage, CQG
Trading: FDAX, FDXM ...maybe
Posts: 388 since Sep 2012
Thanks Given: 499
Thanks Received: 470


For many discretionary traders there is a temptation and an understandable drive to be in a trade all the time. After all if you are not trading you can't MAKE money. At the end of the day of course you can look at your chart and say ... well there were only a few candles where it really looked good so I shouldn't have traded so much.

I do believe that there is an advantage of say - having a bias and stalking your trades from one side only. Seeing the market from either the long or the short side stops a trader from trying to see it from both sides all the time. In that way a random bias (any bias) is a potentially helpful lens.

There are most certainly better ways of looking at a market and developing a bias than flipping a coin. Examining the overall trend for example works well and if applied intelligently will allow a trader to better stalk the trade.

CL 2013:
Taking a look at the CL chart for 2013 and it is clear that the market moved in a less than random direction. Right ? When it went up it did so over a period of time and when it went down it was trending. A simple strategy of "trading in the direction of yesterday's dominant move" would have been a winner if managed correctly.



DAX 2013:
An even more pronounced trend in the DAX this year. Keep your coin in your pocket for this one.




Trends are apparent and continue longer than you expect. Although a coin flip bias is in my opinion better than no bias abdicating our decision making to randomness is far from optimal. We can do better.

p

Visit my NexusFi Trade Journal Reply With Quote
Thanked by:




Last Updated on March 4, 2014


© 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