Our Financial Data Providers and Approach to Historical Data Accuracy

This article provides an overview of Portfolio123 data sources and our approach to historical data for backtesting, how it is managed within our platform, and how it affects the quality and reliability of the financial data presented. Our current data providers include FactSet and ICE, which we used for many years.

FactSet

FactSet is our primary source for financial data, providing valuable insight into companies' financial conditions and prospects worldwide. Founded in 1978, FactSet is a leading provider of integrated financial information and analytical applications for investment professionals around the globe. It is our primary data provider for Fundamental and Estimate data.

https://go.factset.com/marketplace/catalog/product/factset-fundamentals

https://go.factset.com/marketplace/catalog/product/factset-estimates-consensus

ICE

The Intercontinental Exchange, commonly known as ICE, primarily supplies the price and volume data we use. ICE is an American network of exchanges and clearing houses for financial and commodity markets. Founded in 2000, it operates regulated exchanges, including the New York Stock Exchange, for a broad array of derivatives and securities contracts.

https://www.theice.com/solutions/data

Survivorship Bias

Survivorship bias is a common pitfall in financial data analysis that can lead to distorted outcomes and inflated performance projections. It occurs when the failure instances—such as delisted stocks or bankrupt companies—are removed from the data set, which can lead to an overly optimistic view of investment potential.

Portfolio123 provides a more accurate picture of market dynamics and investment strategy performance by maintaining a comprehensive data set that includes current and delisted stocks. It's an approach that sets us apart and helps our users make informed decisions based on comprehensive and reliable data.

FactSet Data Handling

Though FactSet is a reliable data provider, it is important to note that it does not strictly adhere to a point-in-time approach. When companies announce earnings, certain data points, such as earnings, share count, and sales figures, are always expected. However, what other data a company provides can vary.

FactSet typically backfills their database with later-released figures following an announcement. The implications of this practice may impact the backtest and result in performance differences on the same trading system run on different dates.

Our Testing and approach to PIT

In our own testing, we found that the backfilling practice did not significantly impact the accuracy of our data. Primarily, this is because we use end-of-day prices, making the need for momentary, just-released data less critical. Moreover, when released, most information is known or can be calculated from other known data in the financial statements.

Since we began using Factset as our primary data provider, we strived to improve how close the data is to true PIT by tracking changes and backfills by Factset.

Preliminary Data Settings

If the handling of real-time historical data is a concern, users have the ability to manipulate the use of preliminary data in their tests via the "PIT Method - Prelims" setting. Users can include preliminary data for a more comprehensive view or exclude it for stricter testing conditions.