Use Case | Backtesting

TickSmith’s GOLD Platform harmonizes and organizes historical data for efficient backtesting 

Backtesting– A Method For Ensuring Success

Before investing any capital, financial institutions use the method of backtesting to analyze the risk and profitability of their trade strategy. Backtesting confirms the viability of a trading model, strategy, or algorithm by simulating it’s performance on historical data to generate results. Normally, every time a trading strategy is developed, the first rule is to backtest it.

In the past, backtesting was performed mostly by large institutions and professional money managers because of the high cost of managing the necessary datasets. Thanks to advances in technology, this is no longer the case. Although companies must have internal resources to use the data for backtesting, accessing the mishmash of datasets in multiple formats still poses a problem. How is it possible to organize, clean and consolidate historical data from various sources? 

TickSmith’s GOLD Platform Gathers, Organizes and Enables Users to Leverage Data

Our GOLD Platform is engineered to gather raw data, which includes alternative data, into the software stack and standardize it into a single format. Using a powerful ETL, it identifies the outliers– data that doesn’t fit into the set format– within the data and is able to change its behavior. 

Quality data access for backtesting will:

  • Improve existing algorithms
  • Calibrate algorithms
  • Build out new trading algorithms
  • Get an edge over competitor’s algoes
  • Simplify creating AI and machine learning based algorithms

The GOLD Platform provides data to various groups for backtesting:

  • Trading Desks
  • Quantitative Analysts
  • By-side Firms
  • Sell-side Firms
  • Market Makers
  • Account Managers

Once stored within the platform, the data is organized through categorizing, is easily accessible, and readable in a human-friendly format. Then users can leverage the data for use in a data-driven approach, either for machine learning, analytics, backtesting and more. Leveraging data requires quality data that is clean, because polluted data will taint the results.

The highly proprietary coded programs used for leveraging the data is added and maintained on the client’s side. The GOLD Platform can easily be connected to the data that is in the cloud and to the programs, algorithms, and strategic systems produced by clients in-house. Therefore, the secret sauce of their competitive edge remains private.

The Benefits of Being on the Cloud

When it comes to data ingestion, cloud technology has several advantages over an in-house IT environment.  Thanks to the cloud, the platform can inexpensively ingest every type data imaginable to backtest. As a result, quants can run wild with their imaginations in terms of what they wish to backtest, because the data is readily available. The cost-implications of backtesting and experimentation that existed in the past are no longer an issue.

The National Bank of Canada has benefitted in many ways from moving to the cloud. For example, the bank mainly uses the GOLD Platform for Book Replay. However, the same datasets are used for backtesting because the platform is able to distribute data for use to multiple groups within the bank.

Before the cloud, simply accessing data would take days with the bank’s in-house legacy system. Data queries are now executed within seconds. Their old and fragile IT infrastructure could not scale, however, cloud technology allows for unlimited scalability. Both the benefits of quick data access and scalability are core to the National Bank’s trading and market making activities.

The Future is AI

The GOLD Platform is the future-proof backend plumbing for companies who need to manage, control, and access quality data. Data is what algoes feed on. Those who use the platform can customize which data points are pertinent, therefore optimizing their strategy. The platform can ingest a variety of raw data, including trade data, alternative datasets, and consumer sentiment data.

The current market trading environment moves at such a fast pace today. The advancement in quality data access has allowed organizations in financial and Capital Markets to fine-tune their models and be highly performant. The ultimate measure of success is how optimal a trade execution strategy is. The activity of backtesting is the necessary process of due-diligence to perfect the performance of the strategy.

Share this information:


Learn more about our Big Data Solutions

Scroll to Top