Financial Services
Canada
Founded in 1817, the Bank of Montreal (BMO) provides commercial and retail banking services to more than 13 million customers globally and manages CA$1.37 trillion in assets. With over 50,000 employees, BMO is among the largest diversified financial services providers in North America.
Scaling a proof-of-concept application
In a highly regulated industry such as financial services, accurate and timely compliance reporting is essential to operations. Forecasting long-term risk is a crucial element of regulatory reporting for IFRS9 and CECL compliance in the US and Canada. So, BMO needed an application to make credit loan risk calculations, which rely on comprehensive data analysis of past economic events, current conditions and the ability to model future economic conditions.
Work began in 2019 to develop a proof-of-concept application to power BMO's credit risk applications. The goal is to replace an existing end-to-end solution provided by another advanced analytics and data warehousing vendor. The existing vendor solution was functional but very slow and expensive.
BMO wanted to escape vendor lock-in and expand capabilities to support the credit risk forecasting use case through the use of modern, open-source software. These tools would allow them to scale their platform to support the growing needs of the business.
In 2021, BMO decided to move the platform in-house and off the incumbent provider’s architecture. Cloudera was selected as the new data and analytics platform to give BMO the needed control and flexibility.
“Success with our initial proof-of-concept proved the potential of using a new data platform, as it made clear that we were being held back by the limitations of our existing provider. For our new application to reach its full potential, we needed greater flexibility and the ability to scale, which Cloudera enabled,” said Chad Mukherjee, Managing Director of Credit Model and Climate Risk Technology at BMO.
Collaboratively delivering a flexible data platform
The Cloudera Professional Services team worked closely with BMO on deploying Cloudera across all environments. The team’s knowledge and expertise ensured that the migration was a quick and seamless experience.
BMO's newly named Cheetah platform was deployed in production on Cloudera in 2021. Forecasting for CECL regulatory reporting was the first use case, followed by IFRS9 and then Credit Risk Stress Testing.
“From a technology standpoint, Cloudera delivered everything we needed. From efficient deployment across environments to offering advice on optimal cluster configurations, Cloudera has been with us every step of the way,” Mukherjee said.
Faster reporting, further into the future
BMO used Cloudera to develop an open data architecture for quantitative analytics applications, utilizing Apache Spark for batch processing and model execution. By consolidating its tech stack onto a single platform, BMO has more than covered the cost of the Cloudera license. Since deploying Cheetah on Cloudera, runtime for batch Credit Risk Applications has been reduced by 90%, giving BMO the ability to generate regulatory reporting results much faster and more frequently.
Cheetah can also handle large data volumes through Cloudera, with up to 18 million records analyzed in monthly intervals. Working with such volumes enables BMO to forecast as far as 30 years into the future, making well over a billion calculations per forecast.
