My Automated Trading Strategy (3Years+ of Historical Data) Please Critique!

Also, keep in mind it isn't hard to create something that looks extremely profitable, but it doesn't work out at all during real-time trading. In the attached pictures, these profit and loss values are per share.
 

Attachments

there should be a program, a method, a statistical algorithm
that analyzes these performance curves and decides
how likely it is that such performance and curve-smoothness
would happen with a random strategy !

Presumably the strategy is built in retrospective choosing
from millions of indicators, but even then the peformance
is still amazing.
What could be the explanation, what psychology of
other traders is being anticipated and successfully
outperformed here ?

I haven't read the whole thread.
Has it been confirmed ?
I mean, that it's a simple strategy, one out of less
than a million possible ones or such.
 
Alright, I'll show an example of a random trading day with the entries and exits. As you can clearly see, every candle that initiated a signal, at least went ONE tick against the trade, therefore meaning the limit order would be filled.

<a href="http://content.screencast.com/users/UATrading/folders/Automated%20Strategies/media/2770b75a-c493-48d6-9a0f-7201d2aca3cf/SP500StrategyTradeExamples.png"><img class="embeddedObject" src="http://content.screencast.com/users/UATrading/folders/Automated%20Strategies/media/2770b75a-c493-48d6-9a0f-7201d2aca3cf/SP500StrategyTradeExamples.png" width="1686" height="981" border="0" /></a>

I hope this helps visualize what type of trading my strategy does.

BTW CORRECTION: It is not ALWAYS long or short. Occasionally the strategy will recognize a changing market condition, and it will halt on the trade.
 
Quote from forsalenyc:

You mean to say you started looking into all this when you were only 10 years old? that's disgusting.........take it as a compliment if you could. I think you'll come out on top even before you turn 30 years old. having said that, what do your friends do for fun?
More...

lol Im still an average kid, I just have a different interest. I play sports, hang out etc. etc.

My friends do the same, sports, hang out etc. though some of them have started asking questions about my trading :cool:
 
Nice to see a thread that has not turned into a hate fest.

My only 2c is keep it up kid, you are light years ahead of many by starting now and already into Ninja doing backtesting.
 
Quote from brownsfan019:



My only 2c is keep it up kid, you are light years ahead of many by starting now and already into Ninja doing backtesting.
More...


+ 1
 
Quote from julian0625:

Because it depends on your broker. My commission on the s&p emini is about $4.50 a RT so it will cut into a little less than half of the profits as the strategy makes about $10 a trade.

$10-$4.50= $5.50

SO with commision the emini strategy made $55k or so.

The euro strategy would be $100k or so net as well.

So this obviously slipped my mind, but this strategy would have returned 1500% not 2500%.

Still pretty impressive IMO. Is this what I should expect though?
More...

Great Job on your simulation and following your interest.

I would say though your average profit is so low that even the slightest inaccuracy in your back-testing software or data will eliminate the expected profit. For example:

-It is not realistic to think you can consistently enter at the price that your bar chart says is the close. Just a bit of variation here from your simulation will take away your average profit.

-What simulation software are you using? Some programs use such a liberal algorithm for Fills that almost no strategy is needed to show a profit over short holding periods. For such a small expected profit the algorithm specification is critical.

-A strategy that has commissions = to 50% the expected profit while in backtesting/simulation is extremely vulnerable.
It often means you are trying to get an edge that is not really open to you as a retail trader (and that is assuming the backtest does not have other issues). All else equal, the higher the expected profit to transaction costs, the greater the likelihood the strategy will work.

I would suggest:

Look to develop strategies with a significantly higher expected profit, particularly relative to ALL transaction costs.
 
You must log in or register to reply here.
Back
Top