Algo trading, the use of computer programs to execute trades based on pre-determined rules and conditions, has taken the investment world by storm. It's fast, efficient, and can help maximize returns while minimizing risk. But did you know that the success of algo trading depends heavily on the quality of the data used to build the algorithms?
Inaccurate or incomplete data can lead to wrong assumptions and suboptimal trading decisions, resulting in reduced profits or even losses. Therefore, it's crucial for algo traders to ensure that their data is accurate, reliable, and up-to-date.
After going through tons of websites and sources, we have curated a list of top 5 credible sources to help you build your model. If you just want to see the sources, skip to the bottom. Otherwise read on to understand WHY it's important to choose the right sources of data.
The issues that may arise in your data strategy
Building an algo trading strategy involves several steps, including data collection, data cleaning, data analysis, back testing, and refinement. The quality of the data used at each stage of the process can significantly impact the success of the strategy.
Data cleaning is a crucial step in the process of building an algo trading strategy. The format in which your model will read data will inevitably be completely different from your source data.
Secondly, with the market there are a lot of inconsistencies that may pop up. Some sources (not naming them here) have given us different results of day close, day high for multiple stocks.
We later cross verified them with the current day on Zerodha, and weeded out the bad sites!
What data do you need apart from Market Price?
As you, an algo trader, back-test and check your algorithm performance, you will start to look for ways to improve the performance. You need to refine your algorithm by changing the weighting of existing data points, or modifying the rules that govern the algorithm. Testing in different scenarios is key to prevent over fitting while staying highly optimal.
If you decide to change your rules, a new type of variable may help you, like if you are only using price points, maybe volume data would help. Or maybe instead of using daily data, you would prefer an hourly structure. Depends on your trade execution strategy and the kind of results you are getting.
In conclusion, the success of your algo trading strategy depends heavily on the quality of the data used to build it. Algo traders must ensure that their data is accurate, reliable, and up-to-date. They must also have a deep understanding of the data they are using and continuously refine their algorithms based on the testing and results. Don't be afraid to experiment and try out different strategies. You never know what might work!
Here are the top 5 credible sources that can be used to build an algo trading model:
1. Screener - https://www.screener.in/
Screener is a website that gives great insights about stocks, their qualitative anyalisis and quantitative analysis. It's definitely one of my favourite websites to refer to. You need a premium subscription to run your testing here.
2. The OG - National Stock Exchange (NSE) - https://www.nseindia.com/
The NSE website provides real-time market data, historical data, and reports on stock performance. You will get data in Excel format, including daily and historical stock prices, index values, corporate actions, and other data points.
3. Yahoo Finance India - https://in.finance.yahoo.com/
Yahoo Finance India used to be one of the most popular places to download financial data, and can be useful if you want historical data for training your models. Google Finance has now replaced it as one of the top choices for financial news and analysis. Most of the data is free to download.
4. Intrinio - https://intrinio.com/
Intrinio is a data platform that provides access to financial and market data. The platform supports Excel add-ins for downloading data in various formats, including Excel. It's only based for the US market, but its one of the few platforms that give Options pricing data, which is generally very tricky. And yes, its a paid subscription.
5. Alpha Vantage - https://www.alphavantage.co/
Alpha Vantage is a popular data provider that offers real-time and historical data on stocks, forex, and cryptocurrencies. This platform also provides APIs for accessing the data, making it easy to integrate with your algo trading model. I personally have not used this, but they offer global coverage, which is the biggest advantage if you are building for a global strategy.
