Tuesday, March 13, 2007

continued ...

Peter said...

Thanks for your reply, you have given me plenty of information to digest.

No worries :)

An issue I'm struggling with at the moment is finding a viable way of determining if my system does actually put the odds in my favour. My strategy is nothing flash, 34,55&200 MA for trend, S&R and reading candlesticks on 30min and 1hour. I either trade the breakout or trade the bounce. I use metatrader and have an account with IG index (who i hate but have yet to find better - recommendations of brokers would be greatly appreciated).

Ok, first of all IG is not a broker - it's a bookmaker - just to make sure you understand that. Although it's virtually the same as trading on a broker they do have different motives.

Because my system is largely based on identifying S&R and candlestick reversals etc it is hard to back test the data due to not having a clue how to actually transfer all this info into excel. As you say how can i repeat a gut feeling of where i deem a Support or Resistance line is identically correctly placed in a future scenario?

However you have presented me with an interesting theory, 1 that could definitely make my life easier and completely remove the discretion from my decision making process....

Theoretically, i should just scrap everything I've learnt this past year with S&R and candles etc, find a dumbass MA crossover strategy (which i always hated the thought of because of the sheer no brains simplicity of it), play with the MA's, stops and profit limits and backtest until i find the best probability and just trade away blindly to probability success.

Is it really that simple? Is that the holy grail? Am I talking gold dust or just crap?

Your time and thoughts into this are greatly appreciated.


thanks again

Ok, the last bit then.

Firstly let me say that some people (albeit a small number) can trade like that with s&r pivots etc and consistently make money, lots of it in fact.

But as it is gut feeling based and given that you can only develop gut feelings with experience - this method takes years and years to be good at. I can do it and it took me about 2 solid years of 12 hour days watching charts to get good enough. even when you are good enough your results are affected by outside influences like your health, mood, women etc.

There aren't many people who can do that and that's why they make the big bucks.

It's not how I personally choose to secure my future earnings. You have to understand something fundamental about me personally.

The way that i discovered trading in the first place was by asking myself a single question. and I still ask the same question every day.

"How can I make the most amount of money ... with the least effort"

To me trading s&r, Pivots, MA's etc etc is hard work and draining. what's more it's very hit and miss. This is why I do not choose to trade in this manner.

As for your question - is it really as simple as just testing ma's and money management - is that the holy grail?

My answer is yes it is. finding one that works is the hard bit, and believe me there is no moving average cross that works on any pair in any combination - how do i know that? - because i've tested them ...all.

But that does not mean you shouldn't test them - trading methods are like the notes in a song - despite the fact that every note has been used billions of times - every day someone comes up with a new song from the same seven notes. The same is true with methods of trading and you may just have the next winning system. But unless you have a way of measuring it then you won't know if you have a winning system or a bag of crap.

What I can tell you though is that systems that make little money are not super rare - there are many ways of making a little money in this game - however I can tell you that winning systems that consistently make money time after time after time with enough profit and consistency on which to place some serious money - are as rare as hens teeth.

My trading at the moment takes roughly 30 minutes a day - but i spend the other 10 hours doing nothing but research and testing. Testing a single system properly takes about 2 man months

Ive worked that way now day in day out for around 2 years , and it's not just me, there are several others in my team- 10 hours a day research and development for 2 years - lot of hours i think you will agree. How many systems do you think I've developed in that time that deserve real money placing on them - and I'm not talking a few thousand here ....

Well I'll tell you - two systems, that's all. and just one of them is plenty enough to make me millions and millions. see, - rare as hens teeth.

My team have tested hundreds of systems - but there are hundreds more that we have not even heard of yet. that's why we keep on testing and testing, day after day. When the systems we have have made us the money we want then we will probably stop but for now ... back to the grindstone I'm afraid :)

My advice to you for your next step? - learn to use Excel :)


Soren said...

Just wanted to drop a quick comment in. It is possible to make consistent profits using a simple moving average cross over strategy in an automated fashion. The key is that no one set of averages will work forever. Because the underlying dynamics of the market place are in constant state of change as is, the price itself, you have to design a way to constantly adjust the parameters of your crossover to be the most profitable combination for the current market dynamics.

Just some food for thought. Soul - I like what you've done with the blog, keep up the good work!

Soultrader said...

that's very interesting Soren - meaning i guess you have found a way to make a dynamic moving average.

very very interesting, ill have to have a think about that one. Based on something like a multiple of atr maybe or daily range ... hmmm

Soren said...

Think a bit more outside the box. What I am describing is not a new type of moving average calculation or formula...

Soultrader said...

no, that's not what i meant - I meant something like getting an automated strat to decide on the ma lengths by using something like daily range or some volatility indicator.

I wasn't thinking you had invented a new indicator - is there any more to invent?? :) - just a way of determining the correct values for a standard sma but based on volatility or momentum.

am i still in the box?

Soren said...

Ah yes, you're thinking down the right path now. Now let me through something else at you. If the underlying Market dynamics are in constant flux, then the optimal ATR settings with which to determine the optimal SMA settings would also be changing with time. In order to break out of that loop, one needs to find a way to select optimal parameters which is not dependent on the market data itself.

Soultrader said...

Define 'Market Data Itself' - are you talking about nothing whatsoever to do with price action?

that only leaves fundamentals - or is the box i have to think outside rather bigger?

Soren said...

Well, there as two approaches I can think of. You are essentially describing one of them. You could attempt to determine the fundamental factors the drive the underlying market dynamics themselves as use those factors to determine optimal parameters for an sma crossover or something similar. IMO that would be an extraordinarily complex task.

There is an easier method which doesn't require one to model the market dynamics explicitly :)

Anyway, just wanted to share somethings to think about. I try to remember personally to constatly challenge my own assumptions when it comes to trading and I encourage other to do the same.

cheer mate!

Soultrader said...

sure, thanks for the discussion, makes a change to speak to someone positive and i know you of old :)

To be honest at the mo im all full up with testing stuff for the next couple of months but i'll add this to my ideas list.

To be honest i'm not sure I need much more - my two main strategies now have offspring and trade on many pairs but you just can't get out of the habit of looking :).