Thread replies: 6
Thread images: 1
Thread images: 1
Anonymous
Custom Stock Screener 2016-08-05 09:54:47 Post No. 1425454
[Report] Image search: [Google]
Custom Stock Screener 2016-08-05 09:54:47 Post No. 1425454
[Report] Image search: [Google]
File: screener.png (419KB, 653x403px) Image search:
[Google]
419KB, 653x403px
Hi /biz/bro's,
With the ASX market game up and running, I've decided to test my trading process/system.
I was hoping to get some of your feedback regarding pros/cons of the system that I'm currently developing/testing. I am a newbie so please be kind and helpful.
So what I've built so far is a simple stock screener. and my trading process is as follows:
1. It downloads Yahoo Finance records into an SQL Server DB for every single ASX stock (OHLC, Value Metrics like PBR,PER, PERGRWOTH, Earnings, etc).
2. Then I run SQL script that essentially screens these stocks based on criteria that revolves around the following:
- does it follow a bullish candlestick pattern (is it a Morning Star, is it an inverted hammer?)
- what are the fundamental value metrics of the stock ? (is the Price to Book ratio less than the industry average?, is the PER less than the industry average etc.)
- is the stock in mean-reversion with respect to the 250-day average?
3. From the list of companies that the script generates, I read the financial statements, and determine which ones to invest on.
4. Once I have made my choices, I determine the TP and StopLoss levels for each company, based on 250-day Standard Deviation.
5. I trade it, monitor my portfolio exposure + free cash (i dont buy more than 5 companies at a time).
Yesterday morning (4/8/2016) it gave out the following stocks among others:
Harami Pattern: STO (Santos)
Engulging Pattern: TPM (TPG Internet)
This morning (5/8/2016) it gave this:
Morning Star: GXL (Green Cross)
Harami: SPO (Spotless Group)
Does /biz/anon have any criticisms of my trading process?
>>
Have you backtested it? What were your results? If not, why the fuck not?
>>
>>1425462
Good point anon, i havent. Mainly because when i was writing that part of the code i found it hard to implement each candlestick pattern. But yes i would love to add that part in the process. Preferrably as part of step 4.
>>
>>1425469
Well I'd make it my top priority if I were you. You have to watch out for accounting for fees, latency etc. and also you risk over-fitting to your test data. But it's probably the quickest way to test if your candlestick patterns even make sense, and to tweak some of your parameters.
Other than that, consistently sticking to a system sounds like the proper way to trade.
>>
>>1425473
Alright, a sensible advice.
Ill make a profitability back test by detecting all the times a stocks time series follows a pattern and then check the closing price after 2 bars.
It will be very simplistic but i guess its a start? Ill start workingon it this weekend i guess haha
Thanks mang
>>
So I just download this thingy and I get free money?
Thread posts: 6
Thread images: 1
Thread images: 1