Quantifying investment risks

Quantifying investment risks: Forecasting delivery time of new-build projects of Dutch housing associations

Seda, Edwin (TU Delft Architecture and the Built Environment)

Geurts, E.H.M. (mentor)
Gruis, V.H. (graduation committee)
van 't Hek, Maarten (graduation committee)

Delft University of Technology

Architecture, Urbanism and Building Sciences | Management in the Built Environment

2022-06-17

Dutch Housing associations (HA’s) are responsible for producing, maintaining, and managing about 30% of all Dutch housing stock. HA’s draw up their investment forecasts yearly for the next 5 years to construct, improve or maintain homes and other real estate investments. Since 2013, the realization rate of new construction plans by HA’s, which is the comparison of forecasts (dPi) against realized plans (dVi) decreased due to HA’s not realizing new build homes within the time they propose to realize them in their forecast plans. HA’s currently use valuation methods which assist them to mitigate emerging risks that affect new build plans of HA’s. However, valuation methods have been found to focus on indexable risks and capture financial loss while excluding time effect of risks. This means that new build investment forecast as currently conducted yields inaccurate results and are considered too optimistic. Forecasts that are too optimistic lead to disappointments from tenant organizations and municipalities, reduced financial guarantees from lenders, long waiting times for tenants and affects financial feasibilities which rely on accurate prediction of time to completion of projects.
The aim of the research is to explore how new build plans can be made more realistic by accurately predicting the delivery time of investment forecasts. The study results in the identification of risks that lead to delay of new build investment plans and their subsequent indicators. The risks include long permit procedures, long land acquisition processes or lack of land positions to build, long tendering procedures, contractor related delays, rise in construction costs and lack of capacity at municipal level in dealing with development projects. The indicators of risks which statistically significantly predicted project time are construction costs, change in input price index of material and labour costs as of date when decision was made to tender, municipal location, and type of construction i.e., on empty ground or existing site that needs demolition. The project indicators can be used by HA's to accurately predict project time via stochastic decision tree models (SDTA) that rely on multiple linear regression (MLR) and Monte Carlo simulations (MCS). Supervisory bodies can also use these to gauge realism of new build investment forecast.

Housing associations
new build
Stochastic Decision Tree Analysis (SDTA)
Multiple linear regression (MLR)
Monte Carlo Simulation (MCS)
systematic and unsystematic risks
investment forecast
delivery time

http://resolver.tudelft.nl/uuid:e379f532-51a0-4d3b-9b2e-c015b5404955

52.132600, 5.291300

Student theses

master thesis

© 2022 Edwin Seda