Learning from the past: how technology can help avoid the Housing Bubble 2.0

By Dave Ferguson, SAS Ireland

A recent article in the Irish Times about the Swedish housing market gave me an eerie sense of déjà vu. With soaring house prices and increasing debt, some Swedish economists and politicians are concerned about a Housing Bubble 2.0 – while others argue “the risks are manageable”. Sound familiar?

It takes me back to 2008, and the finger pointing that followed the first property crash. As a technologist working in the area of risk, I clearly remember the cries of “the risk models made me do it”. And, sure, weak risk models would have played a part in bad lending decisions – up to a point. But I also believe that technology has a huge role to play in supporting better decision-making moving forward.

As talk of the Housing Bubble 2.0 seeps into Irish public consciousness, the key lies in learning from the past and avoiding the same pitfalls a second time around. Many areas of the IT industry have made progress in this area: achieving shorter development and deployment cycles, and becoming more responsive and robust. But risk model development and deployment typically hasn’t followed suit – why is that?

In my experience, developing and applying risk models can be a fraught business. There are a multitude of reasons why but, for me, the main handicap is that risk model development is still incorrectly seen as an IT-owned process. This means lenders are hampered by:

  • People: the competing agendas and requirements of IT, risk analysts and business users are not communicated or prioritised in the building of models.
  • Processes: are often loosely defined, with poor communication and understanding of what needs to be done, when, and by whom. Operating models are based on silos of responsibility, which leads to ineffective handovers and poor insight.
  • Technology: systems that force IT and the business to create ‘black box’ mini processes are expensive, risky, and slow to develop and deploy models to the business.

The result? Suboptimal models are deployed into situations where they are destined to underperform. And time delays can severely limit the bank’s ability to react to changes in the market or accurately reflect the true risk the bank is holding.

But the mistakes of the past are the opportunities of the future, and getting the risk models right is a huge step towards making better lending decisions. So how can this be done? Some thoughts:

  • People: the process of managing risk should be owned by all stakeholders, elevating it from poor second cousin to a priority for collaboration – like any other ‘sexier’ initiative in the bank.
  • Processes: collective understanding (though not necessarily responsibility) of the processes for deploying risk models will ensure it is not a case of ‘throwing it over the fence’ for others to deal with.
  • Technology: it is essential to choose service-oriented technologies that can cost-effectively bridge systems and provide ways to minimise re-coding of logic from one technology to another.

Capabilities such as in-database solutions can massively reduce the time to develop, update, and deploy risk models. In addition, technologies that enable a ‘model factory’ approach – where the process of data acquisition to model deployment is seamless, controlled and, where appropriate, automated – will help to significantly reduce cost and risk.

While I am under no illusion that technology is the panacea to ensuring Housing Bubble 2.0 does not happen, I am certain it will go a long way to giving banks the platform to enable better risk management practices.

Get in touch to find out more about SAS® Risk Management for Banks, or follow SAS Ireland on Twitter or LinkedIn to keep up to date on our latest news.

 

About SAS

SAS helps 80,000 organisations around the world take their data… And do amazing things… We help organisations turn large amounts of data into knowledge they can use, and we do it better than anyone. It’s no wonder an overwhelming majority of customers continue to use SAS year after year. Find out more, Why SAS?

 

Property For Sale
Capabilities such as in-database solutions can massively reduce the time to develop, update, and deploy risk models. In addition, technologies that enable a ‘model factory’ approach – where the process of data acquisition to model deployment is seamless, controlled and, where appropriate, automated – will help to significantly reduce cost and risk.

Back to Top