Chart of 10,000 coin tosses

[damianoakley]

No I do not use random entry, because, like most people, I would feel a bit daft flipping a coin each time I wanted to enter the market.

However, and this is my real point, the entry strategy I use is VERY simple and probably has about the same reliability as the coin-flip, but my method of deciding on which direction to trade takes less time than flipping a coin.

As for the results, I average about 5% to 7% return on total capital per full week of trading. Average hit rate is about 35%-40%. So for example, if I started the week with £10,000 trading capital, I would expect to clear about £500 - £700 profit on the week or there-abouts.

On my coin-flip entry tests, the hit-rate was 36% on a 200 trade test. The profit was 7% return on capital.

I will now be leaving my computer till the morning, so I can continue this discussion from 8:30am tomorrow if anyone wants to.


Cheers

Damian

 

[bracke]

damianoakley

Being daft in the privacy of ones own home is one of lifes privileges!

If your method is quicker than flipping a coin are you prepared to divulge it?

Regards

bracke

 

[damianoakley]

My method of entry is super simple. It's so simple that I don't mind divulging it, although I suspect that it's just too simple for most people to follow. Most people would want to mess with it, and fine-tune it with additional indicators. Most people, I suspect, would advise that I should spend more time on deciding on my entry, but I prefer to focus the majority of my efforts on position sizing and exits.

I trade 5 min charts, making on average about 20 trades per day. At the start of the day, I simply determine the short-term trend of the market using the MACD indicator, and immediately enter the market in the direction of the trend. I place my stop immediately after entering the trade, and then trail the stop as the trade moves in my favour. If i'm stopped out, I re-enter the market in the opposite direction.

And that's it! The above simply describes how I enter the market. It mentions nothing about how much I stake on each trade or how I calculate my stop. It's these two things that ensure I profit over the long-term.

Instead of focusing on trying to be right every time I enter the market, I cut my losses straight away and try to stick with trades moving in my favour for as long as possible. My strategy is probably difficult for a lot of people to trade because they would want to spend a lot more time on entry than I do.

 

[bracke]

damianoakley

Thank you for your reply

You appear to be operating a simple straight forward system:

1 Follow the trend
2 Set stop and cut losses immediately
3 Let profits run and trail the stop

Each to their own, simple and straightforward gets my vote

Regards

bracke

 

[damianoakley]

That's pretty much it mate. Plain and simple.

I didn't really want to get into describing my strategy in any great detail to be honest. The point I was trying to make (in line with the original post) was that most indicators and entry techniques have little or no advantage over a coin-flip.

 

[bracke]

damianoakley

Can' t go with you on that, indeed your method gives lie to the coin flip approach.

You go in the direction of the trend, if you use a coin what parameter are you giving the coin - direction of trend, opposite direction of trend?

You set a stop - coin toss stop or no stop, what about level of stop.

You determine exit - coin toss exit or no exit

Your method is simple and straight forward, you determine the entry, exit, stop by experience and knowledge of the market - a coin has no knowledge or experience.

If you think that a coin toss can do the job as well as you, your description of the parameters that you would allocate to each coin toss would be of interest.

Regards

bracke

 

[damianoakley]

I think I need to re-clarify my original points.

THE COIN-FLIP ONLY DETERMINES THE DIRECTION OF THE TRADE - IE, LONG OR SHORT.

It does NOT determine stop-levels / exits - I determine those. It does NOT determine how much money I stake on each trade - I determine that.

This is the point I am making, and this has been proven: take a popular indicator, MACD, Stochastics, Moving Average Crossover. Over the long-term, these all work out around 50% accurate - no better than the random generation of a coin-flip.

 

[bracke]

So you might as well toss a coin to determine the direction of the trade instead of determining the trend.

Regards

bracke

 

[damianoakley]

That's right - it would work.

But because I trade Forex, which is a 24 hour market, when I start trading at 9am a trend may already be in place so it makes sense to go with it.

 

[donaldduke]

In a trending market a coin flip will make money in the long run but only if you cut your losers short and let your winners run.

In a non trending market using a coin will kill you.

But if you know the market is trending you obviously go with the trend and dont toss a coin and risk going against the trend.

Any usefull indicator that only works 50% (or less) of the time will give a signal when the magnitude of a win is likely to exceed the magnitude of a loss. Not the probability of a win exceeds the probality of a loss.

A coin toss just signals that the probabilies for a win or loss are the same, nothing about the
magnitude of your win.

So indicators that work only 50% of the time but signal big moves are obviously much better than flipping coins.

 

[damianoakley]

1) A coin flip entry will not kill you in a non-trending market if you are using good position sizing and a good exit strategy. ANY entry strategy would kill you without good risk and money management.

2) If the market is trending in your favour, then you're already in the market with a trailing stop and so
you wouldn't flip a coin to go against your current position.

3) Indicators as you rightly say do work 50% of the time and do signal big moves from time-to-time. A coin flip also works 50% of the time and also leads to big moves from time-to-time, so what's the difference?

 

[donaldduke]

It will kill you because you wont be able to beat your transaction costs (commisions and spreads), youll
get stopped out too often and you wont enjoy any big winners. At the end of the day youll probably break
even but your broker will clean out your account with his charges.

 

[damianoakley]

You've got a good point there about the transaction costs.

I still reckon though that over time, after cutting your losses and running your profits, your total winnings would outweigh your total losses even allowing for trading costs.

I think that this thread might be reaching a natural end for me!

One thing that bulleton boards like this prove, is that everyone has differing opinions.

And that's what makes a market in the first place.

 

[bracke]

damianoakley

Did you toss a coin to decide if the end was nigh!

Thanks for the conversation

Regards

bracke

 

[damianoakley]

Sorry - i just realised i might have tried to end the thread abruptly (and rudely) before it was ready!

Anyway, to summarise my standpoint (just so I don't get a repuation as some random trading coin-******!)

I don't use random-entry in my trading. I just believe that it can work when combined with a good money-management strategy, good position sizing, and a good exit strategy.

I trade the four major currency pairs on Forex intraday. My strategy is an always-in-the-market trend-following system.

If anyone has any other comments or questions, feel free to email me on [email protected]

Thanks everyone for their comments. It's been an interesting discussion.


Cheers

Damian

 

[TheBramble]

Anyone read "The Diceman"?

The power the central character had in relinquishing all his decisions to the roll of the dice was enormous.

No emotion. No thought process. No second guessing and no chickening out. He just set himself options and let the dice decide. Then he just did it.

Sounds like a fully automated, emotionless trading system to me!

I agree good money management and risk control would be essential (in this hypothetical coin-tossing trading system). But this is true for ANY trading system.

I also agree a trending market would favour a positive outcome.

I'm not sure a non-trending market would kill you, especially in FX where 'non-trending' is a very relative term.

However, over the long-run, with a de facto non-trending to trending market ratio of approx. 70% - it doesn't look like the best option to follow.

 

[bracke]

damianoakley

I han't thought you ended the thread abruptly or were rude, neither had I thought that you were an inveterate ****** - of coins - that is

Thank you for your time

The Bramble

Perhaps certain politicians have read The Diceman. Judging by some of their decisions one is inclined to consider that dice throwing is the limit of their thought processes.

Regards

bracke

 

[trendie]

Experiment in Coin-Flipping:

Why: although the thread is academically interesting, I think a real experiment would be more useful and entertaining.

When: over the next 8 weeks. ( 40 days )

What: The Dow.

How:
Rules; take a position end-of-day, preparing for the following day.
For sake of argument, take the closing price as starting point for tomorrows action.

Stop-Loss; if the Dow goes 60 points againt you.
Stop-Win; if the Dow goes 75 points in your favour.
Close the position at end of the Day if stops havent been hit.
( later variation may include moving stop to break-even - but lets get the basic experiment out of the way ).

I have not considered trading costs, or slippage, or spreads.
Why; if the experiment fails, it doesnt matter !!
If the experiment results in some profits, it may be worth looking at refining it.

I have decided to flip the coin now, so that there is no suspicion that I am guessing the direction of the market the night prior !!
( actually, you can have your suspicions about the first day, but not the 2nd onwards !! )

Out of interest, the coin is a Spanish 500 peseta, minted in 1989.
Heads is BUY, Tails is SELL.

NB: The flipping of this coin does not constitute investment advice, and is for educational purposes only.
If in doubt, please consult your financial advisor, or flip your own coin.

Week 01:
Jun 07: BUY
Jun 08: SELL
Jun 09: SELL
Jun 10: SELL
Jun 11: SELL

Week 02:
Jun 14: SELL
Jun 15: BUY
Jun 16: SELL
Jun 17: BUY
Jun 18: BUY

Week 03:
Jun 21: SELL
Jun 22: SELL
Jun 23: BUY
Jun 24: BUY
Jun 25: SELL

Week 04:
Jun 28: BUY
Jun 29: SELL
Jun 30: SELL
Jul 01: SELL
Jul 02: BUY

Week 05:
Jul 05: BUY
Jul 06: SELL
Jul 07: SELL
Jul 08: SELL
Jul 09: SELL

Week 06:
Jul 12: BUY
Jul 13: SELL
Jul 14: BUY
Jul 15: BUY
Jul 16: SELL

Week 07:
Jul 19: BUY
Jul 20: SELL
Jul 21: SELL
Jul 22: BUY
Jul 23: SELL

Week 8:
Jul 26: SELL
Jul 27: SELL
Jul 28: SELL
Jul 29: BUY
Jul 30: BUY

Let the games begin ......

 

[pkfryer]

Trendie, if this new coin flipping method turns out profitable you could write a book about it and sell it on the web for a few grand! There is a theory amongst the more mystic traders, that nothing is by chance... infact we could start a whole new trading approach based on I Ching!

I'm also trying to devise a TA indicator based on tea leave patterns that I think might have promising results... I'll keep you posted on that one

 

[trendie]

Yes !! its called synchronicity.

As well as "coining" it selling books, we could give seminars on the secret art of coin-flipping!.

You could sell proprietary tea-cups without which the tea-leaves wont work.