Skip to main content
To KTH's start page

When Risk Hits Home

Natural Hazards, Transition Risks and Housing Prices in Sweden

Time: Thu 2026-12-03 09.00

Location: Flodis (F3), Lindstedtsvägen 26, Campus, missing video conference [MISSING]

Language: English

Subject area: Real Estate and Construction Management

Doctoral student: Tommaso Piseddu , Fastighetsekonomi och finans, Stockholm Environment Institute, Stockholm, Sweden

Opponent: Universitetslektor Lovisa Högberg, Fastighetsekonomi och finans, Mittuniversitetet, Östersund

Supervisor: Professor Mats Wilhelmsson, Fastighetsekonomi och finans

Export to calendar

QC 20261118

Abstract

The goal of this doctoral thesis is to investigate whether climate risks, both physical (natural hazards) and transition-related (policy and energy shifts), affect the Swedish housing market. This kappa provides the framing text to four papers.

Introduction The thesis begins by collecting evidence that rising averagetemperatures are increasing the frequency and the costs of natural hazards.2024 has been the warmest year on record both globally and in Europe. Swedishinsurance data display growing damage from floods and storms. Real estate isargued to be especially relevant to study the impact of climate risks oneconomies as real estate assets cannot be relocated, mortgages run for decades,markets are often illiquid and information is asymmetric. Most existing researchpresents evidence of significant impacts on the US markets, with little attentionpaid to the topics in Europe and, in particular, in Sweden.

Theoretical and methodological framework The thesis adopts adiscounted cash flow approach (DCF) to justify why significant impacts to realestate prices should be reasonably expected as a consequence of exposure toclimate risks. Although originally designed for commercial property valuations,DCF explains why exposure to climate risk should lower prices, through higherinsurance premiums, higher expected damages, increased operational andmaintenance costs, tighter credit conditions and higher vacancy rates. Themethods chapter reviews the models that are adopted in the four papers that areincluded in this thesis: OLS and hedonic models, difference-in-differneces(DiD), repeat-sales and the synthetic control method (SCM).

Paper 1 is a systematic map of the existing evidence on climate risks and realestate valuations. It finds a dominance of US case studies, with a specific hotspotin the Miami-Dade county and a focus on hydroclimatic hazards. Little attentionis paid to compounding and cascading hazards, while wildfires outside of the USare very rare. Overall, a strong bias towards single-family housing is noted.

Paper 2 investigates the impact of the 2014 Västmanland wildfire, the larges inSweden’s history, relying on a DiD approach that builds on a dataset of 100,000housing transactions. Final selling prices are found to have fallen by about 2%within 20km from the perimeter of the event and about 10% within 5km fromthat. The effect fades at a distance of about 27km. Transactions in the directionof the smoke showed a more substantial impact.

Paper 3 examines the impact of the 2021 Gävle flood using SCM. Monthlymunicipal price indices are constructed by the authors, given that these are notpublicly available. The flood significantly reduced final selling prices, with signsof rebound after a year. Place tests and a DiD alternative approach confirm therobustness of the findings. This represents, to the best of the authors’knowledge, a first application of SCM to the Swedish housing market.

Paper 4 analyzes the effects of the news of the 2011 Fukushima accident onfinal selling prices of assets located in the proximity of Swedish nuclear powerplants using a DiD approach. Prices fell up to 4% and time-on-market increasedby 5%, with effects diminishing with distance. A test on the decommissionedBarsebäck plant showed no significant impact, supporting the interpretationand the robustness of the results.Conclusions. Across the four papers, exposure to climate risks is found toresult in significant discounts in selling prices, consistent with the predictions ofthe DCF model. The author suggests future research on why prices recover, howrisk maps interact with actual events, the effects of adaptation measures on finalselling prices and the impact on credit risks and probabilities of default.

Link to DiVA