The Bank of England had a busy end to September. Last Wednesday it announced it would buy £65bn (about $72bn) worth of government bonds after the pound fell to historic lows and pension funds collapsed, all seemingly the result of the government’s mini-budget days before.
The move sent shockwaves through global financial markets. Regardless of politics, the Bank of England’s job is to ensure monetary and financial stability – and it relies heavily on data to do so.
To improve the data side of its operations, the bank has embarked on a far-reaching data quality improvement program across the financial sector, while at the same time being in the early stages of migrating its on-premises data analytics platform to the cloud.
Speak with The registryPeter Eckley, head of data and analytics strategy at the Bank of England, said that in times of crisis or instability, as Britain saw last week, the bank relies on a multitude of data and forecasts.
Gathering unconventional data sources…means economists don’t have to wait for month-end or quarter-end stats. You are fighting a crisis and you can see what is happening day by day
“When events like the pandemic or the war in Ukraine trigger changes in macroeconomics that trigger changes in forecasts… it affects that way,” he said.
For example, at the time of the pandemic, the bank’s strategy led to gathering unconventional data sources to understand economic activity rather than waiting for quarterly updates.
“That could be the number of trucks parked on the Dover motorway, or hotel occupancy, flight data and so on, to try and build a higher frequency picture in real-time of what is happening with macroeconomic activity. That means economists don’t have to wait for month-end or quarter-end statistics. You are fighting a crisis and you can see what is happening day by day,” he said.
The Bank of England is currently using an on-premises data stack based on a Hadoop data lake and a SQL data warehouse. It uses a range of analysis tools including Microsoft Excel, R, Python, Git for version control and Tableau for data visualization. It also uses more specialized econometrics tools like Stata and EViews.
“It’s a pretty mixed picture right now, and the mid-term vision is to migrate the data analytics workloads to the cloud,” he said.
It’s perhaps coincidental that Eckley spoke at a Teradata event and met with executives from the vendor, eager to show they’d successfully revamped their data warehouse and analytics platform for the cloud. Teradata counts global banks among its customers, including NatWest and HSBC.
The Bank of England is the central bank of the United Kingdom. It was given responsibility for setting interest rates in 1997 by the then Labor government. In 2013, the Financial Policy Committee (FPC) was set up as part of the new regulatory regime put in place to improve financial stability in the wake of the financial crisis.
Eckley said at the Teradata event in London this week: “As regulators, we have a responsibility for monetary and financial stability: we need to stabilize an incredibly complex global system. We need to be able to see the risks so we can mitigate them We need a lot of data and companies are spending hundreds of millions if not billions of pounds a year in the UK to compile that data. And yet the data is of variable quality and not always fit for purpose.”
To meet this challenge, the Bank of England has been working with the Financial Conduct Authority on a program to improve data quality for a year. The project is expected to last a decade.
But new data technologies aren’t just helping the Bank of England fulfill its mandate for monetary and financial stability. They also create greater challenges on the regulatory side of their remit, as many of the institutions they regulate use AI and machine learning to make decisions. Ensuring they do so without introducing undue risk into the system is a challenge for the bank.
“A lot of the companies we regulate are taking over [AI and ML], and so it is this kind of boundary of policy that addresses operational resilience. We need to look at the use of machine learning and artificial intelligence and make sure they are applied in a way that leverages innovation but also avoids threats to business security and soundness or financial stability,” Eckley said.
In its report released in February, the bank said that AI can benefit consumers, businesses and the broader economy, but it can also amplify risks and create new challenges. “The AI models used in the financial system are becoming more sophisticated. Their speed, scale and complexity, as well as their ability to make autonomous decisions, have already sparked considerable debate.
With the global economy and financial markets poised for a stormy ride in the near term, despite this morning’s reversal in the top tax rate providing some immediate relief, the old lady of Threadneedle Street is at least struggling to get a handle on the data that it needs to improve his vision of the future. ®
https://www.theregister.com/2022/10/03/bank_of_england_dta_analytics/ The Bank of England had a tough week. Now for analytics in the cloud • The Register