A Statistical Approach to Real Estate Scenario Analysis : Exploring Application of Forecast Intervals ; En statistisk procedur för scenarioanalys inom fastigheter : Tillämpning av prognosintervall

Investing in real estate carries inherent risks due to fluctuations in economic activity, changes in population dynamics, and shifts in market demand. While traditional approaches to scenario analysis, grounded in market expertise and keen intuition, have stood the test of time, they are also subjective and prone to human error and external influences. Therefore, an objective approach based on statistical inference was sought to serve as a supplementary instrument for real estate industry professionals. With efficacy and practical functionality in consideration, this thesis explores various solutions and determines autoregressive processes as a prime candidate for such an instrument. An instructive procedure is developed and applied to two data sets of historical Stockholm office rents and yields respectively. Starting with data typically available to real estate investors and advisors, this procedure implements locally weighted scatterplot smoothing, polynomial regression, autoregressive integrated moving average processes and matrix transformations to derive forecast intervals which may be applied to prescribe probability to precise ranges or points of the users variable of choice, several quarters into the future. The results demonstrate limitations in the distance of forecasting using this procedure but display satisfactory performance in the short to medium term. Additionally, the practical applicability of the procedure is reflected upon. ; Investering i fastigheter medför inneboende risker på grund av fluktuationer i ekonomisk aktivitet, förändringar i befolkningsdynamik och efterfrågan på marknaden. Medan traditionella tillvägagångssätt för scenarioanalys, grundade på marknadsexpertis och skarp intuition, har bestått tidens tand, är de också subjektiva och medför risk för mänskliga fel och externa faktorer. Därav eftertraktades en objektiv metod baserad på statistiska processer för att fungera som ett kompletterande verktyg i fastighetsbranschen. Med hänsyn till effektivitet och ...