Two Canadian financial institutions are using SAS analytics software to take calculated risks based on their big data flows.
Scotiabank, a longtime SAS customer, has been using Base SAS to program risk models for their lending portfolio. Jonathan Audino, director of information services at Scotiabank, says the bank is now automating its internal reporting functions within SAS, allowing it to deliver summaries faster and dig down into transaction-level data where needed.
“A lot of times our business partners, when looking at summary data, they may have questions about the atomic-level data,” says Audino. “They may recognize patterns and they want to understand more. SAS allowed us to do that quickly, provide that feedback back to the client.”
The analytics software also integrates nicely with the bank’s backend infrastructure, he says. “We’re heavily invested in DB2 and SQL Server, and it also helped us pull data on the mainframe side, so we’re able to bring that data together into a SAS dataset.
“And then we can actually write back…we haven’t seen that in some of the other vendors.”
Meanwhile, Aviva Canada Inc.
is using SAS analytics products to estimate risk in insurance coverage, identify fraud and run its Web services.
Michel Sabourin senior Manager of data analytics for the insurance company, says in one week, SAS BI Web Services replaced a system that had taken Aviva took four months, $70,000, and plenty of Java programming to create.
“Now we can send requests to SAS Web Services and it’s as simple as the other solution we designed. It’s very, very nice for us,” says Sabourin.
The new Web interface, he adds, is giving Aviva “a lot of flexibility that may not be found in other companies right now.”
“Now we’re starting to play with Google, overlaying our information,” says Sabourin.
Maxime Lafleur-Forcier, manager of actuarial research and development at Aviva, says the company harvests textual data to model insurance risk. For example, the company can analyze insurance claims to see how often water leaks originate in a pipe, or a window. People “think that what they enter is not going to be looked at. It’s not true anymore,” says Lafleur-Forcier.
“It’s a good source of information [that] can be used at different levels,” he says, “one of them being fraud.”
Aviva can also compare the data it collects against warranties on particular products; if a roof comes with a 20-year warranty but is shown over time to suffer defects after 15 years, insurance rates can be adjusted accordingly.
This will translate into a more accurate pricing structure and eventually, lower costs to the consumer, the company says.
Sabourin says the possibilities offered in big data analysis are transforming the company. “We’re living in an exciting world,” he says. “This year is going to be a big year for Aviva.”