This year’s edition takes place on 12–13 September 2026 at Pembroke College, Cambridge.

The format is deliberately unhurried: each paper gets 45–50 minutes, with time for a formal discussant and open Q&A. Topics range across real estate finance and economics broadly — risk management, capital structure, sustainability, international investment, ML and AI in real estate, and more.

Submissions are due 1 June 2026. Around eight papers will be selected, with decisions by 15 June. If you have work in progress that fits, send it my way: htl24@cam.ac.uk.

The full call for papers has all the details.

A small writing change makes the necessary quality checks a bit easier for me:

In academic writing, it is uncommon to say “I”. The convention is “we”, even in sole-authored papers. In the age of AI-enabled research, I think that should change to “I”. When the text says, I did this. I tested that. I find this, it becomes clear who is responsible. The author, alone. No “we”. No hiding behind “Claude wrote this”. It is my text. When I claim authorship, I also own the errors and that’s why I cannot be complacent.

Yes, the text above sounds like AI slob. The short sentences. One line paragraphs. A good example of the indirect AI effect on our communication, as documented by Yakura et al. (2025).

The way I “do stuff” has changed tremendously. The long nights of getting the data ducks in a neat row are disappearing. Visualisations and interactive data explorations are so much easier now. I love those interactive HTML slides that let me quickly zoom into spatial estimation output, much faster than importing everything into GIS.

But the longer I work this way, the more I realise that the way I “think about stuff” has not fundamentally changed. A lot of the real work still happens away from the keyboard: while cycling into work, doing the dishes, sitting in a meeting, or mentally turning over a research step. Running empirical tests requires as many critical checks as before. We still need to think carefully about data characteristics and limitations, the data generation process, measurement, context, theory, and links to other sources.

Just taking the output from Claude Code, however smooth it looks, is a recipe for disaster. And that will not change, even with better models. Quality control is as important as ever.

To ensure quality, a researcher needs to know their empirics, the literature, the theory, and how real-world data were generated in the real world. AI promises the automation of many steps in the research process. But this does not mean an erosion of true research skill. On the contrary. The better the tools become, the more important it is to identify mistakes and to earn another “Good catch!” response from Claude Code. Maybe we are not redundant (yet).

For a writer, some AI help with ideas or outlines may be acceptable. For a researcher, I would draw the line more narrowly. The core contribution is not just the prose; it is the research question, framing, hypotheses, interpretation and judgement.

If you want AI to help write or polish an abstract, fine in my book. That is a summary of work already done. But use it to suggest the ideas? Not really.

In research, the ideas are the thing. The question, the angle, the theoretical move and the interpretation of evidence are what make the work yours.

I might change my view on this in the future.

AI is everywhere in academic research. Kobak et al. (2025, Science Advances) tracked words that language models overuse — “delve,” “nuanced,” “meticulous” — across 14 million biomedical abstracts and found at least 13.5% of 2024 papers were processed by an LLM. The same pattern shows up in the real estate literature, as a quick replication on 100K real estate papers indexed by OpenAlex shows.

That is the writing layer in the research process. The more consequential shift is deeper. A growing number of papers rely on AI not for drafting but for execution — work that could not exist without machine learning carrying out the core analysis. Bartik, Gupta and Milo (2025) read thousands of municipal zoning codes and built regulation measures that no research team could produce by hand. Calainho, van de Minne and Francke (2024) replaced linear hedonic models with ML on 30,000 New York transactions and showed systematic gains in out-of-sample accuracy. Shen and Ross (2021) extracted a description-quality measure from MLS listing text that captures soft information about property quality invisible to structured data. Leow and Lindenthal (2025) applied the Gu-Kelly-Xiu ML asset-pricing framework to REIT factor returns and showed substantial forecast improvements over OLS.

In each case, AI enables a measurement or prediction the research requires. Remove it and the paper disappears. But the role is still that of a skilled research assistant (RA): executing tasks specified by a human. The principal investigator (PI) — the person deciding what to study and why — remains human.

AI outperforms human researcher in many dimension (speaking for myself, obviously). The question is whether it can shine higher up the value chain. Can LLMs suggest research topics that are genuinely innovative and plausibly doable — functioning more as a PI than as an RA? Do humans still have a competitive edge?

The new working paper tests this. I mapped the full published corpus of Real Estate Economics (1,676 articles, 1973–2026) and real-estate-relevant subsets of JREFE, JUE, AER, JF, and RFS into a shared semantic embedding space. The result is a coordinate system for the field — not a literature review, but a map. Against that map, I generated 1,499 research ideas under eight conditions, varying what context the model received: nothing, the full corpus, individual cluster seeds, methods borrowed from economics and finance, methods from psychology. Each idea was scored on atypicality (a measure of unusual knowledge combination that retroactively predicts citations) and mapped back into the research space.

The figure below shows where generated ideas land. Grey dots are the full corpus; blue dots are REE papers; red dots are AI-generated ideas. Condition A is naïve generation from training data alone. Condition F draws on methods and paradigms from economics and finance journals.

Methodological scaffolding moves ideas outward into less-explored territory. Topical scaffolding alone does not. The best ideas — particularly those generated through method transfer from economics and finance — score comparably to the median published paper on the citation-predictive criterion. Some land squarely on papers published twenty years ago, having rediscovered questions the field already answered. But that is also true of human research proposals.

There is an uncomfortable regularity in how AI gets adopted: if a system offers a plausible-looking shortcut for a task it was never designed for, people will happily use it anyway, and it takes a lot of effort to convince them of the limits. Researchers will use LLMs to generate research ideas. They already do. The useful question is not whether this is a misguided idea but what the machines actually serve them when they try — and under what conditions the output is worth anything. That is what this paper is about.

Working paper (PDF) — Interactive idea explorer

At yesterday’s ECREDA conference in London, I tried to triangulate exactly this. Using real estate research as a testing ground, I explored how LLMs perform on research idea generation — varying domain knowledge and constraints, then scoring ideas for novelty and predicted citation impact. AI can expand the frontier of what gets considered. But the gym still matters.

Slides are available here.

Then, at some point, the author touches on something I do understand.

It is rarely a dramatic mistake; more often it is a small one, a claim that is just a little too neat, or a generalisation that overlooks something obvious, yet it is enough to unsettle the whole structure, because it makes me question not only that specific point but also the parts I had previously taken on trust.

A recent example is the idea that desk-based labour will stop being scarce (as argued here: https://sahajgarg.github.io/blog/cognitive-labor/), which strikes me as broadly plausible, since if cognitive work can be replicated or scaled, its scarcity, and therefore its value, should diminish.

The contrast offered is property, presented as something that remains scarce and therefore insulated from this shift.

But that does not quite hold, because while certain locations are indeed scarce, such as a house on Lake Zurich or a flat in central London, that scarcity depends heavily on where people need to be, rather than on any absolute shortage of habitable or even desirable places.

If work becomes less tied to location, that constraint begins to dissolve, and with it the concentration of demand, since there are many lakes, many cities, and many landscapes that are currently considered “out of reach”, not because they lack value, but because they sit outside existing commuting patterns. In that sense, property is not nearly as insulated as it first appears.

What unsettles me is not the specific oversimplification itself, but what it reveals about the argument as a whole, because if the part I understand does not hold up particularly well, it becomes difficult to assume that the parts I do not understand are any more robust.

We study housing affordability in seven European cities from 1500-2024, using nearly half a million rent observations linked to wages, quality, and inequality. Constant-quality real rents rose slowly, while housing quality improved substantially. The 1910s-1970s saw a “housing affordability revolution,” with rapid wage growth relative to rents, declining inequality, and large-scale housing policies. Yet expenditure shares increased, particularly among low-income households. We use a Stone–Geary framework to reconcile these facts: rising minimum housing standards steepen Engel curves, raising budget shares at the bottom even when rent-to-wage affordability improves. Prices are an incomplete guide to affordability when housing standards evolve.

This figure plots the real rent index, the real wage index, and the implied “affordability” index (wages relative to rents) for a weighted average of Amsterdam, London, Paris, and the four Belgian cities in our sample, 1500–present.

Until the early twentieth century, real wages and rents moved broadly together, implying no sustained trend improvement in conventional rent-to-wage affordability. A marked divergence emerges only in the twentieth century: from the 1910s through the postwar decades, wages rose much faster than rents, generating large and persistent gains in rent-to-wage affordability. These gains largely plateaued in the late 20th century, as wage growth slowed and public investments in housing were rolled back, while housing consumption per capita continued to rise. A natural interpretation of this figure is that affordability improved dramatically in the last century. However, that reading clashes with the dominant contemporary narrative in cities around the world. Our paper shows how both can be the case.

The first risk is the AI-avoiding “Last Samurai”. This is the stance that treats AI use as inferior, even as a form of cheating, and therefore refuses to engage with it at all. For centuries, samurai culture revolved around mastery of swordsmanship, discipline, and personal honour. When firearms arrived in Japan, they were dismissed as inelegant and unworthy of a true warrior. Yet mass-produced guns quickly proved decisive. Samurai who clung to traditional mastery found themselves outpaced by armies that required less individual skill but delivered far greater collective power. Their fundamentals were impeccable; the race had simply changed. The analogy extends to Amish communities in the US who, broadly speaking, chose to avoid technologies beyond the nineteenth century. In higher education, students who never learn to work with AI tools will be slower to achieve their goals, less competitive in roles where productivity is amplified by new tools, and will miss the chance to focus on areas where humans still have a clear edge over machines.

The second risk is the opposite: Skill atrophy (or, a bit more graphic, AI-induced brain rot). It is real, and it is already visible. Students can breeze through their education by outsourcing thinking, writing, and problem-solving to large language models. In doing so, they risk failing to develop core skills, resilience, creativity, and intellectual grit because the heavy lifting has been done for them. In this world, students struggle to draft and structure text. When fewer and fewer people can read complex texts, deep reading may become a new superpower in an age of shallow, AI-mediated summaries. Convenience slowly erodes competence, and fluency is mistaken for understanding.

The challenge for higher education is not to choose between rejection and surrender, but to navigate a narrow path between them. We need to teach students how to use AI deliberately and critically, while still demanding genuine thinking, effort, and originality. The goal is neither the honourable but obsolete samurai, nor the complacent passenger, but graduates who can wield new tools without letting those tools replace their minds. But how exactly can we achieve this?

Commercial Real Estate Analysis for Investment, Finance, and Development

By David M. Geltner, Norman G. Miller, Alex Van De Minne, Piet Eichholtz, Thies Lindenthal, Lily Shen.

Commercial Real Estate Analysis for Investment, Finance and Development, a fully revised fourth edition of the authors’ leading textbook, presents the foundations of real estate investment analysis with the rigor of general finance and economics. This book introduces the essential building blocks of the field: market assumptions, valuation, financial analysis, and development. Drawing from extensive academic and industry experience, the authors approach the investment analysis process using a combination of theory and practical tools in a discussion tailored to advanced students.

Topics include value concepts, mortgage analysis, financing alternatives, option value, leverage and risk analysis, as well as institutional and capital market trends. Additionally, the new edition addresses climate risks, alternative property types, and the impact of technology on real estate as an asset class. New supplemental online resources complement the book’s conceptual and quantitative study questions, chapter summaries, and other useful pedagogical features.

Combining a practical grounding in economics and finance with updated tools and resources, this edition of Commercial Real Estate Analysis for Investment, Finance and Development provides a new generation of professionals the foundation and tools they need to excel as investment managers, advisers, and analysts. Ideal for graduate studies in real estate, finance, and business, this textbook prepares students for the real-world complexities and challenges of commercial real estate.

For access to additional, online chapters and other Instructor and Student Resources, please visit: www.routledge.com/cw/geltner-miller

We updated and streamlined the book: The first part (printed) gives a foundation in real estate finance that should nicely fit into a semester/term. The second part (online) allows for customisation with more specialised topics. This structure will hopefully be closer to the reality of how instructors have used the book in their courses – at least this is how I teach here in Cambridge. Importantly, the new format also brought the price down: With current discounts, it will cost £59 in the UK, which is less than what I paid 20 years ago (even in nominal terms!).