Bloomberg has introduced its new Open-High-Low-Close (OHLC) Bar product, a customisable intraday pricing solution for investment research. Designed to streamline quantitative workflows, the product allows clients to quickly generate tailored intraday pricing datasets using either Bloomberg’s predefined templates or their own customised configurations of trade condition codes, aimed at accelerating insights and alpha generation.
The OHLC Bar product joins Bloomberg’s existing quant pricing offerings, such as its cross-asset Tick History Data, as part of a growing suite of Quant Research Data solutions. By enabling easier management of traditionally large Tick History datasets, the OHLC Bar product helps reduce data wrangling efforts, and leverages Bloomberg’s premium Tick History Data, which provides access to both tick-by-tick intraday quotes and trades, as well as newly introduced OHLC bars. The data covers over 350 global exchanges and includes Bloomberg’s recognised fixed income pricing, spanning more than 70 million instruments for application across back, middle, and front office operations.
Enhancing Quant Workflows
Angana Jacob, Head of Enterprise Research Data at Bloomberg, explains to TradingTech Insight how the ability to generate bespoke OHLC bars can influence quant workflows, particularly in terms of data storage and processing efficiency.
“To benefit from market fluctuations during the trading session and capture additional sources of alpha, quantitative, quantamental and systematic workflows are increasingly expanding into intraday trading strategies,” she says. “In order to backtest and simulate their strategies effectively, quant and systematic customers need access to deep intraday price history across all the asset classes they trade in. Bloomberg’s OHLC Bar product covers over 70 million instruments with deep history and is fully customisable, allowing customers to tailor their historical bars for their specific use case and application, with transparency into the metadata behind each bar.”
Previously, being able to utilise valuable information in historical tick-by-tick format for backtesting and quant research was out of bounds for many firms due to high barriers to entry and the need for sophisticated infrastructure and quant development, says Jacob. “For context, Bloomberg’s Tick History is our largest data product as it is offered in tick-by-tick format and while we significantly simplify data delivery and operations for our clients, it still requires large-scale storage and a specialised time series database on the client side for ongoing maintenance and deployment of their various use cases. Typically, clients of tick history are on the more sophisticated end of the spectrum with specialised use cases such as algo trading, execution and TCA.”
Bloomberg’s OHLC Bar product significantly decreases the size of tick data and the complexity of using it, while still retaining valuable intraday patterns for quant research, suggests Jacob. “It aggregates all the trades in a customer-defined time interval (eg. 1 second, 1 min, 5 min, 20 min, etc.) and filters for customer or Bloomberg-defined trade condition codes to represent the trading activity for that particular bar. As a result, the solution simplifies the data engineering, quant development and operations needed for managing tick history data, allowing our quant clients to focus on generating alpha. Bloomberg’s OHLC Bar product also helps bridge the gap from strategy research to execution by allowing clients to backtest their intraday strategies more accurately and check their assumptions on slippage and trading costs.”
Integrating AI
Accessible via REST API and natively integrated with AWS, Azure, and GCP, the OHLC Bar product is available for enterprise-wide use through Bloomberg’s Data License service. The product’s flexibility allows the generation of bars for global equities, futures, FX, swaps, fixed income, and options, and aligns with the growing trend of integrating AI into quant investment strategies, says Jacob.
“With ever-increasing volumes of data as well as new sources of alternative data, quant analysts and systematic teams are facing significant complexity in evaluating, acquiring and productionising data in their strategies,” she says. “AI techniques, be it traditional ML/deep learning or Generative AI, which has grown in adoption over the last two years, are extremely dependent on the underlying data. If the data is not accurate, clean and point-in-time with granular metadata/annotations, models will struggle to provide accurate backtest results… or even worse, will show a fantastic-looking backtest but the strategy could perform poorly when out-of-sample and live. Thus, the data foundations for algos and AI/ML workflows are extremely important and getting that right for our clients is a big focus for Bloomberg. The OHLC Bar product, with its deep history and condition codes, accessibility, and comprehensive and consistent coverage across asset classes, enables quant clients to improve their models and additionally reduces their time to alpha as they can run backtests out-of-the-box without significant quant development, testing and hardware.”
Key features
The product offers several key features aimed at enhancing flexibility and transparency for users. It enables customisation, allowing firms to tailor their OHLC bars according to specific requirements. This includes selecting input data to generate the bars, defining custom time intervals, and receiving the output in either Parquet or CSV format. Users also benefit from filtering options, with access to Bloomberg’s templates designed to exclude auction indications, duplicate trades, and erroneous transactions, all of which can be modified to suit individual needs.
Transparency is another feature, as Bloomberg provides detailed metadata for the OHLC bars, such as timestamps, counts, and the condition codes used in their creation. Additionally, the product allows for bespoke solutions by enabling firms to generate OHLC bars that cater to particular use cases, incorporating only specific trade conditions as required.
The new product augments Bloomberg’s tailored research data solutions for quantitative and systematic investors seeking to enhance their investment processes. These research data solutions are interoperable with Bloomberg’s Real-Time solutions, Data License (DL), and Data License Plus (DL+), as well as third-party platforms and major cloud providers, facilitating seamless integration into various trading strategies.
Subscribe to our newsletter
