Designing and Building a Profitable Automated Trading System

We all remember the days before PC's and software.

Then the indicator inventors were working to have the ability to measure the market.

Market's moved much more slowly then. it took 6 to 8 days to make 10 or 20% on an instrument. There were no e mini's or instruments like that.

for stocks, the postal service worked for clearing trades. A stack of envelopes appeared daily. I used the telephone to call in trades.

On the DJXX each call involved 11 steps. I kept a sheet that I filled in vertically as the IB spoke the trade ID's etc. C and R was going on all the time as the market changed.

So indicators were invented and then they were tweaked over time.

We will look at the math of each one and see why each aspect is used to come up with the signals used in the demo scott and I did.
 
"Technical analysis from A to Z" is a good reference for indicators. Achelis is in the computerized platform business and he is a nice guy who cares about his clients.

Equis International has a lot of clients and stays on the cutting edge. He is the CEO.
 
Quote from jack hershey:

Here is the general picture in terms of purposes:

attachment.php
More...

This text appeared before regarding the stack. 1 is volume 2 is MACD, etc.


Were the four components arranged in a top to bottom order their respective use would be for the following respectively:

1. patterns, end effects, differentiating retraces from reversals and dominance/non dominance differentiating (knowing which internal price pattern sub set was in play.)

2. money velocity of profit segments and its corresponding acceleration and deceleration.

3. Exit signals and handling trading in coarse chop.

4. Sentiment and Entry signals and handling fine chop trading.

As you see "stacking" this set of considerations from level to level is very handy and instructive.
 
If anyone would benefit from a chart of todays action. I have incuded an image of the ES 5 minute fractal with the requisite indicators (I think), although I am not aware of a way to re-order volume for Ensign to match the visual stack prescribed.
 

Attachments

Quote from ehorn:

If anyone would benefit from a chart for todays action. I have incuded an image of the ES 5 minute fractal with the requisite indicators (I think), although I am not aware of a way to re-order volume for Ensign to match the visual stack prescribed.
More...

Thanks so much.

We can do the day in several ways as time passes.

There are the issues of hindsight, flappers, etc. Knocking those off as we get the coding done with market slices will be fun.

Again I am an advocate of P, V in modern times. Here we are exploring getting coding and integration of signals in hand.
 
Quote from jack hershey:

...Again I am an advocate of P, V in modern times. Here we are exploring getting coding and integration of signals in hand.
More...

YVW,

LOL! I have used only P/V for so long now (relatively speaking) that I am not even certain how to properly add other indicators to my charting package. :D

As always, it should be fun and educational to follow along and Thanks for your contributions Jack.
 
Hi,
What's the objective here? To codify and validate the concept by backtest? I hope so.

There seems to be an assumption here that this works when it has yet to be shown to be likely to do so.


Thx
D
 
I will type for 30 more mins and call it a day.

we need to have the math for each indicator in hand.

this sets the scene for relating the indicators to the bars and their vlaues (OHLC). The inventors focuses on these four values.

STOCH emphasizes the close relative to the volatility and MACD is just an advance on MA's and their families.

I liked the fast STOCH in the stock market because it gave me the advantage of cycle compression to make money.

As is news, the news squeezes out between the fingers like wet clay when you make a fist with wet clay in your hand. A little comes out here and there.

Doing price gaps in stocks is very profitable and it takes little time.

Lets look at the closes of bars before an opening price gap. There are tells on these bars. So it was natural to go pick up an indicator that hypes the tells.

This is the essence of indicator trading. the inventors were those resourceful guys who tailored their designs to "tells".

Today high school kids and even 7th graders catch on to this kind of thing quite easily. It is advanced fish, the card game.

By using STOCH which deals with an emphasis on closes (this is why it goes from 0 to 100), we get to see when a set of bars is tending in a direction even though the volatility and the range of price may not be changing.

I picked a small set of bars, purpoesfully for looking at this tendency under low volume conditions which was the case BEFORE as stock made a price move. Sleepy time.

By honing my sensitivity to what was coming down the pike (an expression for oldest roads), I got the benefit of being pushed by the herd when they woke up.

In optimizing making money to the unbelievable state, I decided to buy gapping stocks in the pm before the am opening gap.

I did get in trouble for this at times because the SEC doesn't like "insider" trading.

Naturally this thread is an illegal thing since if and when it is deleted it represents hiding evidence of my illegal coding efforts. I promised to not get into humor and off topic stuff. Sorry.

Liz has finsihed using the paper shredder early so we are going out to eat. Sorry

Be back tomorrow.
 
Quote from fundjunkie:

Hi,
What's the objective here? To codify and validate the concept by backtest? I hope so.

There seems to be an assumption here that this works when it has yet to be shown to be likely to do so.


Thx
D
More...

No we are going to go through the process of building and ATS that can be used on 25 or more exchanges as a cash cow.

Forward testing is how building ATS's work out.

Whether it works or not has been done using cash and trading accounts.

By forwarding account information directly from a broker to a neutral thrid party, it was determined that about 17 points net occurred on one day in 31 partial fills for one stock where 100,000 shares were traded to exit those shares where the purchase evidence was also provided. So that little part is done.

You may have missed the series of posts that were involved.
 
You must log in or register to reply here.
Back
Top