'I get a real kick out of building models that are used worldwide'

“Our bank has been active in the food and agricultural sector since the beginning. For more than a century, we’ve supported large and small entrepreneurs in the agricultural and horticultural sectors in the Netherlands and around the world. In order to provide loans to farmers in countries like Brazil, Australia and America, we have to be able to accurately estimate whether the entrepreneurs are credit-worthy. To do that, we need different credit risk models than the ones we use in the Netherlands. My team develops those new models. That gives me a kick, because it’s the bank’s core business. Our new models will eventually be used to estimate some very important risks. How cool is that?!?”

Collecting data around the world

“We collect huge amounts of data from our customers all over the world to use in the new models. We can use their historical data to predict whether they’re credit-worthy, or if they might have a bankruptcy waiting in the future. Has this agrarian entrepreneur always paid his bills on time? Does he have any outstanding debts? Has he always kept his books in order? We request those kinds of data from the international Rabobank offices, taking the applicable privacy legislation into consideration. As a result, we’re in constant contact with the entire world: a conference call with Brazil, a work visit to America. And our data analysts play a crucial role in all of it.”

Creating useful data sets

“The data analysts collect and interpret all of the data, so they need to have a good understanding of the bank’s processes. For example: we ask our colleagues at a foreign office how much money a customer owed the bank per month. Sometimes there’s some confusion about the definitions. Because what should be included in what the customer ‘owes’? It takes good communication skills and interpretation by our data analysts to get the data we need. They then organize the data from a variety of sources to create useful data sets. It’s their thorough work that allows the modellers to develop the algorithms.”

Avoiding unpleasant situations

“In this international project, we take the socio-economic differences between other countries into consideration. From a quantitative perspective, the losses in countries like Brazil are higher than in the Netherlands. When a Brazilian entrepreneur goes bankrupt and still has debts to the bank, then the court decides whether the entrepreneur has to pay back the debts. In the Netherlands, that gets taken care of quickly. But in Brazil it can take up to 10 years, so the bank’s losses can continue to accumulate. We need extremely good credit risk models to prevent those kinds of unpleasant situations.”

International collaboration

“Our team is very international. Around 30 percent come from the Netherlands, but the others are well-educated expats from Europe, India, China, Africa, you name it. You talk to people from all over the world every day – both our international offices and our own co-workers. They bring a different perspective of the world. In the end, you and all of those people with different perspectives work towards the same goal, and I think that’s really inspiring.”