A Stress Testing Scenario Analysis for House Prices

A Stress Testing Scenario Analysis for House Prices: An Application to the Greek Market Michael Doumpos Technical University of Crete Dimitrios Papastamos Eurobank Property Services S. A. Constantin Zopounidis Technical University of Crete Dimitrios Andritsos Eurobank Property Services S. A. June 28 - July 1, 2017, Delft, The Netherlands

2 Introduction

Stress tests 3

Objectives of the study 4

Automated valuation models (AVMs) 5

6 Data Description: Sample • The data were provided by the Eurobank Property Services S. A. • Hedonic characteristics of real estate properties • The sample consists of 75, 991 residential properties that have been professionally evaluated in the period 2007 – 2015 • 240 different administrative sectors covering all areas in Greece • 32 aggregated administrative areas (Index Areas)

7 Data Description: Geographical Distribution

8 Data Description: Distribution of Properties • 80% of the properties are flats • 11% are houses • 6% maisonettes • 3% of type duplex

9 Data Description: Initial set of Hedonic Variables V 01 V 02 V 03 V 04 V 05 V 06 V 07 V 08 V 09 V 10 V 11 V 12 V 13 V 14 V 15 V 16 V 17 V 18 V 19 V 20 V 21 V 22 Record code Year of valuation Month of valuation Administrative sector Urban classification Survey value Type of residence Usable residence area Land area Year of construction Distance from CBD Floor Total number of floors Existence of parking space Type of parking Type of heating Quality of construction Number of bedrooms Touristic hotspot Elevator View Number of bathrooms Value Year Month no. Code value Euro Code value Sq. m. Year km Number Yes/no (1/0) Code value (0 -3) Number Yes/no (1/0) Code value

The structure of the AVM under consideration (database) The procedure applied by EPS during the Commercial Index production, is comprised of 4 phases: 10

The structure of the AVM under consideration (valuation process) 11

Econometric model 12

13 Estimation results

14 Monte Carlo Simulation The methodology implemented for constructing commercial property price indices is based on “repeat valuation” of commercial properties approach:

15 Simulation parameters

16 Profit / loss distribution (baseline scenario)

17 Profit / loss distribution (stressed scenario)

18 Results by region (baseline scenario)

19 Conclusions and future perspectives