Rabobank Real Estate Finance switches to data-driven PDCA

About Rabo Real Estate

Rabo Real Estate Finance is part of Rabobank. Rabobank is a Dutch international financial services provider operating on a cooperative basis. The bank provides services to a total of 7.3 million private and corporate clients in the areas of retail banking, wholesale banking, private banking, leasing, and real estate.

As a cooperative bank, it claims to put its clients’ interests first in its services. In doing so, Rabobank wants to make a substantial contribution to welfare and prosperity in the Netherlands and to the sustainable feeding of the world. Through Banking for the Netherlands (national) and Banking for food (global), Rabobank wants to make a difference as a cooperative bank in the Netherlands and in the world, which is expressed concisely in the mission statement: ‘Growing a better world together’.

Opting for a high-tech, high-touch approach

In addition to its strong focus on the customer, Rabobank is going to strengthen its balance sheet and improve performance, partly under the influence of regulations, so that it can face the future as a solid bank. The environment for banks is changing very rapidly. Rabobank, like other banks, is in a transition phase in which extra efforts will be made on digitization and innovation in order to continue to serve customers as well as possible on the basis of a so-called high-tech, high-touch approach.

Focus on sustainability, transformation, and innovation

Commercial real estate refers to financing provided by Rabo Real Estate Finance to clients who invest in, for instance, stores, residential properties, offices, business premises, and hotels. These real estate investments are held by the bank’s clients for rental income and/or capital appreciation. As a cooperative, Rabobank has a focus on sustainability, transformation, and innovation, with a focus on local communities. Not only in major cities but also in viable city and village centers. In this way, they link financial returns to social gains.

Portfolio management in essence

Portfolio Management is fundamentally about:

• Portfolio Analytics: analysis of a sector or portfolio with combined insights into risks, returns, customers, loans, and collateral. In this case, it is about “bricks and mortar”.
• Management: supporting strategic, tactical, and operational decision-making about portfolio direction and adjustment.

The goal of component one is to deliver “Actionable Data-Driven Insights” (see Figure 1). The goal of component two is to support ‘Data-driven decision-making’ with, in time, more automated signaling and decision-making through the use of technology.

Figure 1: Actionable Data-Driven Insights and then what?

Some considerations

• Why couldn’t the technology used to determine the price (and thus the risk) of reserving an airline seat eventually be applied, in part through the use of big data, to financing an investment property, for example?
• In addition, the insights can, within the confines of the law, be shared with customers. Enabling customers to make better decisions that reduce the likelihood of bankruptcy is a win-win for three parties: the bank, the entrepreneur, and society. By putting the insight back in the hands of entrepreneurs, Rabobank helps and facilitates them to do better business.
• The business value the team aims for is the step-by-step delivery of data-driven portfolio insights that lead to action. We call these “actionable insights.” Without limit, objective, or benchmark, the insights are nice to know, but do not spur action and the desirable and necessary PDCA improvement cycle.
• Subsequently, these insights should lead to step-by-step more data-driven decision-making at all strategic, tactical, and derived operational levels. In combination with the deployment of insight, visualization, and analysis technology, Rabobank seeks to steer the portfolio within the bank’s strategic and social objectives.

We elaborate on two ambitions of Rabo Real Estate Finance in the context of ‘Growing a better world together’.

100 percent green portfolio within ten years

Rabobank is aiming for a fully sustainable commercial real estate portfolio within the next few years. It wants to achieve this by 2028 at the latest. The focus is not only on energy, but also on circularity, vitality for the local environment, and operation. It will start by making the commercial real estate portfolio more sustainable. The basic question is: how sustainable is the portfolio and what should be understood by sustainability? The time frame within which it must be achieved is determined.

60 percent of financing outside the four major Dutch cities

Rabobank aims to do at least 60 percent of its corporate real estate financing outside the four major Dutch cities. They do not exclude in advance any regions in the Netherlands. The basic question is: what is Rabobank’s commercial real estate exposure outside the four largest cities? The objective is fixed.

Besides the social objectives, every bank has the objective of improving returns (partly due to regulatory influence) and reducing the risk of lending. After all, a healthy bank, like a normal business, must make a profit to survive in today’s disruptive innovation environment.

“IS” situation versus “SHOULD BE” situation

The case study here is: how can data-driven portfolio management improve loan portfolio composition in commercial real estate? To improve composition, it is important to know well what the composition is (IS situation) and what the composition should be in part, in terms of risk and return (SHOULD BE situation). This explicitly defines the goals to work toward and the decisions to be made in terms of:

• What customers do I want to serve (creditworthiness, professionalism) and what collateral (offices, stores et cetera) do I want to finance where? (= Acceptance)
• What amount do I want to provide, what is the amount for financing? (= Amount)
• What price (interest rate) will I charge for what period? (= Price).

And of course combinations of the above on Acceptance-Amount-Price in combination with new loans, existing loans (refinances) and a constantly changing playing field in terms of market, innovation, and regulation.

4 implications of the datacratic PDCA

In the above process, you can recognize a number of implications.

• There is implicit steering rather than explicit steering of banks. Through a clear focus, strategic objectives, and adequate monitoring of KPIs, adjustments can be made based on the PDCA. In appropriate cases, this can still be done by persons and committees (explicit), but in the long run, it will be done by technological possibilities based on automated business rules within a set bandwidth (implicit). Overruns will then arrive ‘spot on’ to the right person via automatic alerts.
• The need exists to be able to compare many parameters simultaneously. Data on separate lists no longer says enough and certainly does not show the connection with the rest of the data. If you want to maintain insight into all the movements and dynamics in a complex matter, you need a system in which the algorithms automatically weigh all the parameters and are able to link them to each other. Whereas you can still cross two lists of criteria and produce a matrix, the complexity increases enormously with a multitude of simultaneous criteria. However, there are so many criteria on which so much depends that you are looking for ways to cross many lists simultaneously and meaningfully with each other in parallel. This desire increases the need for data-driven work. The need for more sophisticated machine learning is also increasing.
• There is a need for more and more data, software, and statistics. The combination of (historical) data, statistics, software, and industry knowledge are the basic ingredients to optimize your decision-making. By applying predictive methods, you can determine whether a limit will be broken or a target will be met.
• Anticipate instead of being surprised. In the “old world,” you faced a breach of a limit after the fact. In the “new world”, based on predictive data, you can anticipate it and avoid it.

The monitoring process in 6 steps

As with any trip, determining the destination is the first step. Often it turns out that the journey is actually more important than the destination. Gradually, the journey also teaches you much more about your goal than previously thought. Below we give an impetus for monitoring in six steps.

1. Step 1 is the most important. It starts with the intrinsic motivation to want to know where you stand. Not because someone else asks, but because insights will help you make better decisions and increase customer satisfaction.
2. Step 2 is determining your position in relation to your goal: are you above or below it and within what timeframe do you want to achieve your goal?
3. Step 3 is determining what path you have taken, by quarter, by month, by day, and by minute. This depends on your business or industry. Trading on the stock market versus financing commercial real estate. Where are you coming from and where are you going
4. Step 4 is predicting whether you are on the right path. Are we on track? If not, adjustment is necessary.
5. Step 5 is to observe, based on continuous monitoring, whether the adjustment actions have had the desired effect.
6. Step 6 involves gathering even more insight into the dynamics of the data. In a broader context, did the adjustment action actually have an effect? In an up economy, variables will change. This is not due to an action, but to the positive economic development that causes prices to rise. Sector, product, and market knowledge remain prerequisites to recognize developments and patterns.

Figure 2: Monitoring in six steps.

The insight (Check) in relation to the objective (Plan) will, if all goes well, lead to an action or adjustment. Thus, based on predictive insights, you can make earlier adjustments toward the ultimate goal, or avoiding a limit.

What does the future look like?

• There is an urgent need for dashboards, linking data, and building infrastructure. Because the department is as data-driven as possible, it needs dashboards and a solid data infrastructure, and good data quality is required. In this, the department has already invested a lot of time because it depends on connecting all kinds of departments, each with their data flow. This is not an independent data warehouse or data lake, but a derived combined intelligent agile data mart. To keep things scalable, the team chose to invest in both a good technical and human infrastructure where reliability is to be found in the structure and business rules themselves.
• Whereas in the past auditors looked at the annual report and output, in the future auditing will focus more on input and transformation. If the data is good and complete and the automated business rules and derived calculation rules are correct, then it is likely that the output will also be correct.
• Making the determination and monitoring of the target portfolio data-rich will greatly increase the speed of steering and adjustment. Portfolio management is done from the intrinsic motivation of digitally understanding and steering the commercial real estate loan portfolio. This must be supported by a dedicated team with different data specialties and the space and opportunity for development and innovation.
• The step-by-step approach is a prerequisite for this. Every day, step by step, we build actionable insights that result in support for data-driven decision-making at every level of the organization. Software and tooling are supportive – but the will and motivation to improve are not. Through a smart and layered data infrastructure, insights can be delivered at every level. Small individual decisions then contribute to the strategic direction and goals.

Conclusion

You can think of this customer case as a hybrid of developmental learning and design learning. Developmental learning can be seen in the extent to which small decisions contribute to adjustments toward a predetermined goal. This allows increasingly accurate predictions (developing outcomes) that have the desired impact on the target portfolio. Design learning is reflected in how the outcomes from the algorithm (the designed outcomes) guide all member banks.