I would suggest not to “marry” to one broker API. Use platform like NinjaTrader that can connect to verious brokers. They already have plumbing done and you just need to concentrate effort on your strategy.
How many of you need to learn coding to work with data ?
- Thread starter Spikeet
- Start date
I would suggest not to “marry” to one broker API. Use platform like NinjaTrader that can connect to verious brokers. They already have plumbing done and you just need to concentrate effort on your strategy.
This business model makes no sense to me.
If you can't code then how do you even use these data scans? You need code to backtest them and other code to trade them.
And if you can code, then just create the scans yourself for much cheaper.
For a SaaS business to make sense it needs to save me money compared to doing it myself, this seems to do the opposite. I can build your $499 product (for example) using polygon data for $200 and save myself $299/mo, not to mention having more flexibility to add any features I need that you don't support.
no, research is when you have an entire universe of stock data, and you can query it with words instead of writing code. those are just templates, you can change and improve on top of them by adding filters/more data points.
I don't see how you can "reproduce" any of what i mentioned, since it's a huge DB linked to a complex SQL filtering system.
Building a trading system is far too complex for the solution he proposes.
And what if you wish to test multiple timeframes simultaneously?
dump the data into https://questdb.io/ and do queries. what is complex to one is trivial to another.
But I think most data platforms focus on sales and graphics (shiny things). Add a grid of numbers that blink different colors
I'm manly talking about fetching the market data to build the strategies for now. What do you think about that ?
Yes ?
if I give you bulk of data can you pull all stocks that had earnings on the day, gapped 30 %, spiked 10% in the first 30 min of mkt open had a low price in the first 30 min which is lower than the low price in first 60 mins, broke out through the 200 MA, below VWAP, etc etc, all in a few seconds ?
I'm drawing lines on Tradingview as honestly as possible representing a trend following idea I have, and related variations. I've not yet taken the time to fully replicate the thought process in Pinecode which is new to me. If I can do that then I can test more variations of what I'm intending. So for now I use Excel to help gather the data points, and then VBA to chart the data, and then to apply money management. I am most interested in drawdowns, and, what is the shortest time frame I can use and still overcome trading costs like fees and spreads. Also if there are better hours during the 24 that are better than others. Of course I want to know all the stats like average win, average loss, profit factor, win%. These could possibly be used for other studies involving synthetic trades for Monte Carlos. I'm testing an always in method and comparing that with in and out in accord with my baseline always in. And then I want to see how things go with additional markets and how I could increase or decrease leverage according to rated profit factors and drawdowns for the various performance stats. Basically looking at all angles before starting to reduce unhappy surprises. The manual drawing of lines is good for refining the thought process for the rules of trading.
So anyway, coding helps with all of this and will help to execute. I had to develop a nifty way to speed up data entry and to produce charts with granularity of three minutes per data point so as to compare side by side the different time frames, different markets, I am proposing to work with, and how money management might improve performance.
Speeding up data entry I needed some custom code that starts whenever double click on a cell in a column to plot enter/exit prices, along with the position direction (long or short) on a date/minute template. If price moved from 28678 to 28589, for example, I only need type in, or voice in, the last three digits and it takes care the rest. Things like that are good applications of an ability to code.
This is all because you can't yet draw a line on Tradingview, and get Pinecode to read it into a CSV file the way it can if you have a robot set up. And even if I were able to robotisize this now, it's still valuable to manually go through every situation just to be able to formulate Pinecode using bar data.
I'll be using VBA to test position sizing. Ild like to be able to designate a drawdown tolerance, say, 20%, or 5% (if trying to qualify for a prop firm) and be told, for each situation, what my leverage should be, or how much exposure I should have relative to a turnaround price on the chart (or stop loss price).
The idea is to use coding ability to almost guarantee success before committing.
Couldn't agree more. You really need a profitable strategy to begin with. No amount of coding experience will somehow magically give you a profitable strategy (unless you're overfitting).
A profitable strategy can be overfitted too. Most strategies are based on mathematics (at least for successful funds) so you need coding. Why do you think that Renaissance/Medallion has a lot of mathematicians and lots of computer power?