How to Source Market Data for Algo Trading?

Algo trading involves the use of computer programs or ‘algorithms’ to execute orders in the financial markets (stocks, currencies, commodities, derivatives, etc.) based on pre-defined conditions. These conditions include price, timing, and quantity instructions. It’s become extremely popular among many traders globally due to its high speed and efficiency. And one could argue that the first step in algo trading is collecting market data! This data has to be accurate and precise to form an effective algo trading system.

The amount and type of data required depends on your approach to algo trading, i.e. whether you code your strategy or use pre-built strategies. Traders using third-party platforms need not worry about collecting data. In this article, we will provide a complete overview of various market data sources, the associated costs, and a detailed analysis to help you make informed decisions

What is Market Data and Why Do We Need it?

Just as good-quality ingredients are crucial for cooking up a delicious meal, high-quality market data is essential for backtesting and executing successful trades. Market data refers to trading-related information on the prices and volume reported by exchanges (like NSE or BSE in India). Without this data, a trader won’t be able to form any strategy or place orders in the market. 

A trader needs to collect three types of market data for algo trading; real-time data, delayed data, and historical data. Real-time data is used while a trader executes an order, it is taken into account at the time of the trade. Delayed data refers to market data made available after a short period, usually ranging from 1 minute to 15 minutes. Historical data (collected from past events) is used for analysis and backtesting strategies. It checks whether the strategy would have worked well in previous markets and gives a green signal to the trader. 

Why Good Quality Market Data is Important for Trading:

• Forming your Strategy and Backtesting: By collecting past market data, traders can notice patterns and form strategies. These strategies can be backtested to gain confidence and make sure there are no major flaws. A trader must always backtest their strategy before deploying it in the current market.
  
  Read: Ultimate Guide to Backtesting Algo Trading Strategies
  
• Monitor and Adapt: Regularly analysing and gathering market information allows traders to make timely adjustments as the market fluctuates. This is where real-time market data becomes essential. On the other hand, delayed data is typically used for research, analysis, and educational purposes. Examples of institutions that offer delayed data are Yahoo Finance, Bloomberg, and Reuters.

• On-spot Execution: Real-time market data allows traders to execute trades on time, reducing slippages and increasing accuracy. If there is latency (lag), orders will be placed based on delayed information which can reduce profits. 

Market data is very crucial to the entire trading process. You can make informed decisions and execute trades with the right sources to collect this data. Let us see what these sources are!

Sources of Market Data for Algo Trading

Now that we have understood the importance of market data, let us break down the sources and costs involved in collecting market data:

1. Stock Exchanges 

In India, traders can access market information from stock exchanges like the National Stock Exchange (NSE) or Bombay Stock Exchange (BSE). NSE Data & Analytics Ltd provides market data from the NSE, including quotes and data across the capital market (equities), currency derivative market (CDS), futures and options (F&O), and much more. The data is available in different levels: Level 1 shows the best buy and sell prices, Level 2 displays up to five top prices, Level 3 covers up to twenty, and tick-by-tick data captures every single market movement.

The cost of directly collecting information is quite high. However, brokers access live data provided through subscriptions to the NSE. Apart from that, for real-time stock-wise data, NSE charges a tariff of ₹10,000 per year for a single stock and ₹25,000 per year for up to 5 stocks (with an add-on of ₹2,500 per stock for up to 10 stocks). 15-minute delayed data is provided at ₹1,00,000 per year for the Capital Market and F&O segment. Historical trade data for researchers’ use is provided at ₹18,000 annually in the capital market and F&O segment. [These costs are as of August 2024].

Traders who wish to enhance the speed and efficiency of trade execution and obtain an edge over other traders will have to pay a fee for co-location services and set up advanced infrastructure to access the data as well.

[Co-location refers to the service of setting up servers closer to the exchange. This improves connectivity and reduces latency or delays in collecting data.]

BSE has its own set of tariffs that are different from NSE. Collecting data officially from the stock exchange is quite costly and complex. As a result, this source is mostly used by big institutions rather than retail traders. 

2. Authorised Data Vendors

Data vendors are companies that specialise in providing essential market data, which includes real-time and historical information about financial instruments such as stocks, commodities, and currencies. Authorised by the stock exchange, these companies distribute accurate real-time and historical data without much latency or delays. They provide Open, High, Low and Close (OHCL) price, End of Day (EOD) data, and volume data. Data vendors also help integrate various software platforms needed for charting and technical analysis. 

Platforms like TrueData, Global Datafeeds, and Accelpix offer market data services through monthly subscriptions, with costs varying based on the segment and features you need. These platforms not only provide raw data but also decode it to reveal patterns that traders can analyse.

Data vendors provide an all-in-one solution for acquiring market data, ensuring traders receive comprehensive information from start to finish. This option is ideal for institutional traders who require large volumes of data for efficient analysis.

Broker Platforms 

Several brokers in India provide trading APIs that allow users to build custom trading platforms and automate their strategies. [An API is a set of protocols and tools that enables software to interact with data vendors and place orders on different trading platforms, exchanges, or brokers.] These platforms usually allow users to access data free of cost after signing up. They provide all the necessary information like ask/bid price, volumes, OHLC, last traded prices, etc, in a user-friendly and easy-to-read interface. 

Zerodha Kite, Upstox, Fyers, and 5paisa Capital are some examples of such data service providers. They also provide historical data in the capital market ranging from 1-day to 5-year trends, along with the execution services to carry out your trade. These companies collect their data from official sources and restructure it for the retail market, making it the optimal choice for individual traders. 

Conclusion

Market data is an indispensable part of algo trading. Whether you code your own strategy or use existing strategies, regular market data is extremely essential to execute successful trades. In trading, one must move quickly and make dynamic decisions, which is possible by keeping a check on data trends. 

Stock exchanges and data vendors are great sources for institutions. Retain traders can use broker APIs as it's more economical. As a trader, you must be quick and analytical, and good-quality data is the way to go!