edited by DC12 Heli Song
30 years of MSCA and coming together in Antwerp
In May, the QuiVal consortium met again at the University of Antwerp for the Co-Creation Workshop. The timing was especially meaningful, marking 30 years of the Marie Skłodowska-Curie Actions, a programme supporting interdisciplinary research, mobility, and collaboration across Europe. After several months apart, it was the first full in-person meeting. The focus was on sharing progress, discussing ongoing challenges, and strengthening collaboration. Alongside presentations, structured listening sessions allowed for deeper exchange across different research perspectives. A huge thank you and appreciation goes to supervisors Daan Schraven, Lara Schrijver, Theo Dounas, Wu Chen, Rodrigo Martinez-Mazza, Giuliano Poli, and Hilde Remøy for their time, tips, and careful listening to each PhD’s input. In a project like QuiVal, where disciplines and methods differ significantly, these conversations often turn out to be the most valuable part. Across the discussions, one idea kept resurfacing: entanglement (a concept already introduced in the first blog in this series). It refers to the way systems, actors, and outcomes are connected rather than operating in isolation.
Spillovers in housing markets and why entanglement feels relevant
In urban economics, policy changes, neighbourhood developments, or building-level shocks often spill over into surrounding areas and related market segments. Entanglement is a useful way of thinking about this, not as a replacement for existing theory, but as a way to describe how interconnected cities and housing systems are in practice. One mechanism in my research is the repricing of risk. When new information enters the market, perceptions of safety, financing conditions, and future costs adjust, with effects that extend beyond directly affected areas through expectations and relative pricing. A similar pattern is visible between sales and rental markets. Once considered jointly, they are difficult to treat as separate systems, as changes in one segment transmit into the other through tenure choices, investor behaviour, affordability constraints and demand shifts.
A simple experiment with real estate data
A co-creation outcome was the shared familiarity with machine learning methods, which naturally led me to think about practical ways these tools can complement traditional commercial real estate valuation. One simple exercise I find useful, and would share here as a small food for thought, is to compare standard econometric models with machine learning methods in this context. It is also something that encourages experimenting with new tools and approaches rather than sticking to a single way of doing things. Using office market data such as rents from past leases, property characteristics, population growth, inflation, interest rates, cap rates, and employment trends, office rents can be estimated in two ways: a traditional hedonic regression and a machine learning model such as random forests or gradient boosting. The aim is not to replace econometrics, but to reflect on where standard models may miss nonlinear relationships and how predictions change when models are allowed to learn from richer patterns in the data.
Looking ahead
The Antwerp workshop reinforced that quantum-inspired valuation is less a single method and more a shared way of engaging with complexity across disciplines. Co-creation remains central, as ideas become clearer when discussed across perspectives. Looking forward to seeing everyone again in July in Vienna and in Naples!
