Let's Try an Experiment

[BlueHorseshoe 10]

A brief article here investigating random entry portfolio trend-following:

 

Why random trade entry is as good as any technique

 

BlueHorseshoe

[BlueHorseshoe 10]

As I understand it,this thread is based on random entry based on a coin flip.Whereas the above quotes talk about random entry based on trend following,with the implication being that even trading randomly (with the trend) there is an edge in identifying a trend and trading with it.

 

Until the bit about "long term trend filters" towards the end, then it's not a random entry with the trend, it's a random entry that could be profitable if it is with the trend, or will be unprofitable if it is against the trend. No judgement regarding the direction of the trend is being made as part of the entry decision, as far as I can tell.

 

Of course, it's hard to know exactly what was tested.

 

And I personally believe that efficient entries (better than random) are possible and, if you accept that the data is largely random, then non-random entries make perfect sense.

 

And yes - this thread was just a bit of fun, I think.

 

BlueHorseshoe

[Mysticforex 13]

Conclusion

 

Of the above trades,how many were in the direction of the main trend?

It might be interesting to compare these results to a long only entry each time (if the trend is up) Otherwise,exactly what benchmark are we comparing random to?

Without a credible and tangible benchmark this exercise seems pretty pointless to me beyond light entertainment.

 

As has been pointed out a few times, It is entertainment. If you would like to join the fun, perhaps you could count the number of trades you thought were with the trend.

[Mysticforex 13]

Haven't had anyone come forward wi a flip for today

s trade, so here it is:

 

Tails/Sell at 09:30 EST.

[Mysticforex 13]

Sell @13252 09:30 EST

[Patuca 13]

The benchmark is luck. The definitive is fate. In other words, lo que será será. What is to be is to be and your lucky if you get ...what is to be. Whether you hit it via coin flips...pendulum..mathematical equations..fibs..or other technical contraptions....db's predictive boxes..patuca dipstick opinions...mightymouses mouse trap jailbreaks...bob's planetary rotations and revolutions...

 

Don't you believe in luck and fate mits?

Edited August 5, 2013 by Patuca

[zupcon 10]

Perhaps people could take 2 minutes to help out someone else with an interest in randomness, he's had over 40,000 responses so far, but a few more would help.

 

A Random Number

[Mysticforex 13]

Who was it that said " I'd rather be Lucky than Good"?

[SIUYA 11]

after Mitt - count my random coin flip - tails

[Patuca 13]

Does anyone believe in trying to predict fate? Predict opens a can of worms...fate is fixed...

[zdo 20]

:haha: got the runs yet?

[SIUYA 11]

Does anyone believe in trying to predict fate? Predict opens a can of worms...fate is fixed...

 

you can predict fate - you supposedly cant change it.

 

Mitts - as for a benchmark, its a theoretical benchmark based on what we would 'expect' (as in all probability) to occur.

 

and Zdo,,,,drum roll please...

[Mysticforex 13]

So we have Mitsu Tails/Sell for Tues, 8/6/13 And Siuya Tails/Sell For Weds, 8/7/13.

 

I could not get to my platform at 17:00 EST to manually close an open trade. So it was closed at 18:49 EST -5.7 pips.

[SIUYA 11]

Ok.Maybe it was mentioned and I missed it.What then,is the "theoretical benchmark"?

 

Not being overly concerned with the ideas of fate, belief and gods etc....so I will skip all that,

 

but in terms of a benchmark....

I called it a theoretical benchmark because in this unique case of flipping a coin there is no official benchmark. Whereas if comparing our results to some other index - the usual measure then there is an official benchmark (with its rules and composition) to index against.

When it comes to coin tossing, all we have is the mathematical expectation over the long run.

Otherwise, the benchmark would either be us....or anyone who flips a coin which would be rather pointless.

Even if we combined lots of different coin flippers to create a benchmark, then it should approximate to the theoretical benchmark over the long term.

 

This must surely be the idea that over the long term flipping a coin will provide no predictive edge, and hence why this is purely an experiment in tricking ourselves that it might, and hence why we can market this at a later date as a system with rules in the hope we can trick (ahem....convince) others that it works.

(Now if we introduced some money management and other forms of controls or filters we might actually have something.

[Mysticforex 13]

Here's one for the Gipper, uh, I mean Flipper ( sorry I couldn't help myself ).

 

Sell @13290 09:30 EST.

[Mysticforex 13]

S/L hit at 10:13 EDT -20 pips.

[Mysticforex 13]

Thank you Mysticforex for starting this experiment, this is very interesting!

 

The adjacent thread for the "Predict" part of the experiment is to be found here http://www.traderslaboratory.com/forums/forex-trading-laboratory/16994-predict-experiment.html

 

I have a few questions:

1. Are the trades opened at 09:30 EST disregarding ideal entry location?

2. I created a $100,000 account for this experiment - is this the size account you wish to use? I can change it before Sunday so the 2% of the accounts will be similar in size so one is not stopped out prematurely.

3. What size contract are we entering?

 

Additional Question:

4. I was thinking risk:reward=2%:2% - doesn't that mean over the long run our expectancy is breakeven? What about having a 2% risk, but let the trade run until end of the day and not take the 2% target?

 

Thank you!

 

 

 

Since my Real Day starts at London open, and the flip occurs at NY open, I am not always available to manually close the trade at end of day.

 

Then there is always the chance of letting a winner turn into a loser.

[Mysticforex 13]

sell @13305 09:30 EST

[Mysticforex 13]

Stopped out. At 10:14 EST

[Mysticforex 13]

Mitsu has just pointed out to me that I have made an error.

His flip, ( Tues. 8/6 of Buy was executed as a sell. With the resultant loss of 20 pips, and 2 %

 

The error was mine. We can decide how to deal with it at the end of the contest

 

We didn;t have a flip slated for today, so I just flipped my 5 CHF coin and got Heads/Buy.

 

Thank you Mits for catching that and bring it to our attention.

[Mysticforex 13]

Buy @13360 09:30 EST.

[Mysticforex 13]

TP hit at 10:04 EST +20 Pips.

[Mysticforex 13]

Flipped Heads today.

 

But @13354 09:30 EST

[Mysticforex 13]

Stop Loss hit 11:30 EST.

[onemove 10]

Mysticforex,

 

Found this site in case you need another helper RANDOM.ORG - Coin Flipper. Great pics of coins. You can also choose to flip multiple coins at once (any idea why one would want to flip multi coins at once?)

 

Enjoy!