In a recent Nasdaq survey, more than half (60%) of dissatisfied quantitative portfolio managers complained about an inability to quickly test new data sets. Survey respondents cited a number of issues with data, including an inability to quickly onboard new data, difficulty accessing data in their organization, inadequate resources, and outdated technology. In pre-Covid financial investing, low-volatility investing was the safe bet. After watching the fickleness of the stock market during the pandemic, mitigating risk through quant research has never been more important.
While quant teams are spending more time dealing with these data issues or limitations on compute resources, they’re spending less time testing new academic research, conducting value-added analyses, and performing complex backtests. Their inability to access and onboard new data, such as ESG, sentiment, or cryptocurrency data, lengthens data pipelines and time to market. The data movement required with legacy systems leads to high costs of ETL and rising data management TCO. Teams are also struggling to implement new approaches to research, including using AI and machine learning (ML) capabilities. It’s clear that traditional asset managers and hedge funds need to drive greater efficiencies in portfolio construction, trade implementation, and risk mitigation.
Snowflake allows organizations to seamlessly bring together internal data, portfolio warehouse data, and external vendor data into a single platform. Data access via Snowflake Marketplace or private share enables business teams, quants, and data scientists to take advantage of new and differentiated data, including market, identity, geospatial, ESG, and cryptocurrency data. Data co-location enables quant teams to access, join, query, and analyze internal and external vendor data with minimal to no ETL. To support AI and ML endeavors, Snowflake connects to SageMaker Autopilot and Azure ML. The Snowflake Data Cloud enables a single data repository and native support for structured, semi-structured, and unstructured data. With Snowpark, all ML-driven data science use cases can be facilitated directly in Snowflake.
This means that quant researchers are spending less time managing data and more time gathering insight from data, further lowering overall TCO. Quants need to rebalance and run optimization in ever-narrowing time frames. With global portfolios, it’s that much harder to find the time to run rebalances. To manage the market openings, these queries need to be as fast as possible, requiring even more compute resources. Snowflake has the ability to scale up to the demand, then back down for day-to-day functions. Teams can leverage scalable, on-demand compute when required to construct portfolios, run multi-factor models, backtest, and perform attribution and optimization processes—all without contention.
Here’s how one of our financial services customers is using Snowflake to optimize its research and trading efforts:
At PIMCO, multiple decisions and teams are responsible for the planning, construction, and performance of the company’s portfolios. PIMCO develops proprietary, quant-based performance attribution models to bring greater clarity and flexibility around the sources of risk and return. These models are compute- and data-intensive, as they analyze multiple factors across decades of historical data. With Snowflake, PIMCO was able to scale and speed up its quant-based attribution model by leveraging the Data Cloud to meet its data storage and compute requirements. Snowflake’s architecture separates data from compute, allowing PIMCO to ingest 70 terabytes of input data required for its analyses. Snowflake’s elastic performance engine enables different teams across PIMCO to run isolated and independent workloads without database contention. By architecting on Snowflake, PIMCO was able to speed up time to insight by 3x and scale the number of reports and scenarios.
Discover how your organization can use Snowflake to optimize quantitative research and trading:
- Learn more about the Snowflake Financial Services Data Cloud.
- Read our Financial Services Success Guide: 8 Ways Financial Services Organizations Deliver Innovation and Security with the Data Cloud.
- Read our blog post: Banking on Data, the New Currency of Financial Services.
- Watch our video: Create a Multi-Currency P&L Stock Trading Portfolio View with Snowflake and dbt.
- Read our industry brief: Snowflake’s Financial Services Data Cloud.