Forecasting Commercial Property Market Performance: Beyond the Primary Reliance on Econometric Models A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy Treshani Perera BSc (Hons) in Quantity Surveying – University of Moratuwa BSc (Hons) in Applied Accounting – Oxford Brookes University MAIQS, ACCA Affiliate School of Property Construction and Project Management College of Design and Social Context RMIT University August 2018 DECLARATION I certify that except where due acknowledgement has been made, the work is that of the author alone; the work has not been submitted previously, in whole or in part, to qualify for any other academic award; the content of the thesis is the result of work which has been carried out since the official commencement date of the approved research program; any editorial work, paid or unpaid, carried out by a third party is acknowledged; and, ethics procedures and guidelines have been followed. Treshani Perera August 2018 Forecasting Commercial Property Market Performance: Beyond the Primary Reliance on Econometric Models i ACKNOWLEDGEMENT First and foremost, I wish to thank my principal supervisor, Dr. Wejendra Reddy, for his invaluable guidance, advice and assistance with the many challenges faced during my Ph. This thesis has benefited greatly from many stimulating discussions we had, and his insightful comments and feedback.
I wish to give an equal acknowledgement to Professor David Higgins who helped to get this research started and continuously supported as an external advisor. David was instrumental with industry liaison for data collection. His supervision was very motivational, always providing the encouragement to reach the research goals. I also owe a special gratitude to my co-supervisors, Dr.
Woon-Weng Wong who provided useful input into the quantitative modelling and Professor Ron Wakefield, for their encouragement and constructive advice throughout the research. The nature and enormity of research meant this thesis would not have been possible without the support of industry personnel. I would also like to acknowledge the support of the many property and financial market experts who gave their time and contributed their knowledge in the semi-structured interview research phase. Their expert advice and recommendations were valuable in shaping this research and were significant in result validation.
Special thanks go to Mark Wist (property consultant) who provided constructive comments and suggestions to improve my research. I have been very fortunate to receive the RMIT International Ph. I am grateful to RMIT University for the sponsorships and technical support that made this doctoral study achievable. I acknowledge the assistance of Mr.
Robert Sheehan from Sharp Words Consultancy for his editorial comments. I also appreciate the support provided by the staff and fellow Ph. colleagues at the School of Property, Construction and Project Management, RMIT University. I thank Professor Kerry London for her meaningful introduction to the world of research philosophy.
I also acknowledge Professor Chris Eves, Dr. Eric Too, Associate Professor Ashton De Silva, Dr. Mehrdad Arashpour and Dr. Michelle Turner, for their constructive feedback during my Ph.
Finally, I acknowledge the support, patience and understanding of all my family members. My late father’s blessings and treasured love have always motivated me to achieve the fruit of my academic endeavour. I owe my deepest gratitude to my husband Dimuthu. This thesis would not have been possible without his unwavering support, love, encouragement and tolerance.
Also, I feel blessed to conceive my little sweetheart at the end of my Ph. I would also like to thank my mother and brother Shehan for their motivational support, love and encouragement. This thesis is dedicated to my mother who inspires me and to whom I owe forever for everything I have achieved. To all, I thank you for your support, guidance and encouragement.
It is highly appreciated. Treshani Perera Forecasting Commercial Property Market Performance: Beyond the Primary Reliance on Econometric Models ii TABLE OF CONTENTS Table of Contents. iii List of Figures. v List of Tables.
viii List of Equations. x List of Abbreviations .1 Background to the Research .2 Statement of the Problem .3 Research Aim and Objectives .5 Contribution to Knowledge .6 Scope and Limitations .7 Thesis Layout and Structure .8 Publications and Presentations. LITERATURE REVIEW – COMMERCIAL PROPERTY MARKET FORECASTING AND THE OUTLOOK ON DOWNSIDE RISK EXPOSURE. 13 Commercial Property Market Forecasting .2 Commercial Property Market Structure and Features .3 Principles of Forecasting .4 Commercial Property Market Forecasting.
50 Downside Risks in the Commercial Real Estate Environment .5 Downside Risks in Commercial Property Market .6 Decision Maker’s Imperatives for Downside Risks. METHODOLOGY – PRAGMATIC RESEARCH DESIGN .2 Overview of Methodology .4 Strategies of Inquiry .5 Methods of Data Collection and Analysis .6 Validity and Reliability. 143 Forecasting Commercial Property Market Performance: Beyond the Primary Reliance on Econometric Models iii CHAPTER 4. QUANTITATIVE ANALYSIS – EVALUATING COMMERCIAL PROPERTY MARKET FORECAST ACCURACY .2 Secondary Data Arrangement.3 Testing for Forecast Accuracy.4 Analysing Outliers of Forecast Errors .5 Testing for Relationships between Variables.
QUALITATIVE ANALYSIS – DETERMINING THE CURRENT STATUS OF COMMERCIAL PROPERTY MARKET FORECASTING AND THE LEVEL OF DOWNSIDE RISK EXPOSURE .2 Semi-Structured Interview Plan .3 Australian Commercial Property Market Forecasting Practice .4 Downside Risk Exposure in the Australian Commercial Property Market. FORECAST DECISION MAKING MODEL DEVELOPMENT: INDUSTRY EVALUATION .2 Best Practices for Improving Forecast Accuracy .3 ADSV Decision Making Model for Integrating Downside Risks to Improve Forecast Accuracy 253 6.4 Expert Panel Comments, Feedback and Recommendations. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS .3 Achievement of Research Objectives .4 Contribution to Knowledge .6 Further Research Directions. 318 Appendix 1: Journal Publications.
318 Appendix 2: Supplementary Appendix to Literature Review. 356 Appendix 3: Semi-Structured Interview Guideline. 360 Appendix 4: Participant Information and Consent Form. 362 Appendix 5: Test Results of Augmented Dickey Fuller Test.
366 Appendix 6: Test Results of Vector Autoregression in between Property and Economic Forecast Errors 376 Appendix 7: Semi-Structured Interview Respondents Profile. 380 Forecasting Commercial Property Market Performance: Beyond the Primary Reliance on Econometric Models iv LIST OF FIGURES Figure 1.1: Modelling the Economic Environment.3: Thesis Layout and Structure .1: Property Investment Options for Investors .2: Links between Space and Capital Markets .3: The Relationship of Space, Capital and Property Markets .4: The Real Estate System .6: Asset Class Return for the Year Ended 31st December 2017 .7: Property Sector Weights as at 31st March 2017 .8: Australian Commercial Property Market Total Returns 1987-2017.9: Commercial Property Sector 12-Month Total Returns .10: All ADIs’ Commercial Property Exposures .11: Property Yield compression over the Timeline .12: Ex Post and Ex Ante Forecasting Periods .13: Scientific Methods of Forecasting .14: Forecasting Methods and Their Relationships .15: Steps in Formulating an Econometric Model .16: The Tangent Illustration for MAPE and MAAPE .17: A Theoretical Structure for the Determination of Office Rents .18: A Cobweb Market Adjustment Process.19: Judgemental Intervention in the Property Market Forecasting Process .20: Comparison between Normal Distribution and Power Law Distribution .21: Modelling the Economic Environment.22: Illusions of Certainty .23: Knowledge Transition in Cynefin Model .24: The Black Swan’s Surprising Aspect: Micro Perspective .25: Distinguishing the Knowns and Unknowns: Black Swan Event Framework .26: The Dispersion of Worldwide Natural Catastrophes in 2017 .27: Natural Catastrophes and Manmade Disasters: Number of Events 1970-2016 .28: Total Reported Natural Disasters around Australia between 1970-2017 .29: Estimated Risk Appetite .30: Ranking of Demographic Focussed Megatrends .31: Major Driving Forces in Real Estate .32: Trends for Sustainable Development in Property and Construction.33: Modelling Uncertainty in Real Asset Development Projects .34: KuU Risk Assessment with Associated Probabilities.35: Nonlinearity of Fragility and Antifragility .36: Simple-Complex Model Considerations .37: Disaster Risk Index: Global Map .38: Summary of Structural Changes in the Property Market .1: The Research Process .2: The Research Onion .3: Sequential Exploratory Design. 130 Forecasting Commercial Property Market Performance: Beyond the Primary Reliance on Econometric Models v Figure 3.4: The Ways of Mixing Methods .5: Methods of Data Collection.6: Types of Sampling.1: The Sequential Approach for the Quantitative Analysis .2: The Australian GDP Growth Rate and Employment to Population Ratio .3: Cash Rate – Forecasts vs.4: Bond Rate – Forecasts vs.5: AUD/USD – Forecasts vs.7: Rental Movement-Prime – Forecasts vs.8: Yield-Prime – Forecasts vs.9: Total Vacancy – Forecasts vs.10: Net Absorption – Forecasts vs.11: Economic Forecast Accuracy Based on Scaled-independent Metrics –3M Vs 6M.12: Rental Movement-Prime Forecast Accuracy Based on Scaled-independent Metrics .13: Yield-Prime Forecast Accuracy Based on Scaled-independent Metrics .14: Total Vacancy Forecast Accuracy Based on Scaled-independent Metrics .15: Net Absorption Forecast Accuracy Based on Scaled-independent Metrics .16: The Comparison of Economic and Property Forecast Accuracy .17: A Conceptual Structure of Property Market Forecast Determinants .18: Box-and-Whisker Plot for Rental Movement – Prime Forecast Errors .19: Box-and-Whisker Plot for Yield – Prime Forecast Errors .20: Box-and-Whisker Plot for Total Vacancy Forecast Errors .21: Box-and-Whisker Plot for Net Absorption Forecast Errors .22: Box-and-Whisker Plot for Economic and Property Forecast Errors .23: Stacked Area Diagram for Economic Forecast Percentage Errors .24: Stacked Area Diagram for Property Forecasts Errors, in Percentages .25: Line Diagram for Property Forecasts Errors in Percentages .26: Granger Causality of Property and Economic Forecast Errors.1: Semi-structured Interview Respondents Structures .2: Experience Level of the Respondents .3: Driving Factors for Setting-up Commercial Property Forecast Objectives .4: Respondents’ Sources of Forecast Data Collection .5: Types of Input Data for Forecasting .6: Methodological Orientation of Commercial Property Market Forecasting .7: Forecast Output Validation.8: Level of Confidence of Forecast Models .9: Forecast Model Accuracy Measurement in Current Practice .10: Greed vs Fear in Forecasting .11: Structural Changes in the Property Market .12: Example of Commercial Property Conversions by Location .13: Structural Changes Related to Demographics .14: Physical and Digital Presence in Business Lifecycle .15: The ‘Give and Take’ Effect in the Commercial Property Market .1: The Onion Model for Improving Forecast Accuracy .2: Stress Testing Methodology in Property Market Practice. 250 Forecasting Commercial Property Market Performance: Beyond the Primary Reliance on Econometric Models vi Figure 6.3: Investment Mangers’ Risk Response Strategies .4: ADSV Decision making Model for Integrating Downside Risks to Improve Forecast Accuracy.5: Middle Line of Forecasting within Limits .1: Modelling the Economic Environment.2: Conceptual Framework of Structural Changes in Property Market.3: The Comparison of Economic and Property Forecast Accuracy .4: Granger Causality of Property and Economic Forecast Errors.5: The Onion Model of Improving Forecast Accuracy.6: ADSV Decision making Model to Improve Forecast Accuracy.
281 Forecasting Commercial Property Market Performance: Beyond the Primary Reliance on Econometric Models vii LIST OF TABLES Table 2.1: Four Quadrant Investment Market and Property Investment Products .2: Global Real Estate Transparency 2016 .3: Contingency Table for Directional Accuracy Test .4: What Do We Model and Forecast in Real Estate? .5: Principal Determinants for Net Effective Rent Forecasts .6: Principal Determinants for Equivalent Yields Forecasts .7: Commercial Property Market Rent and Rental Income Models .8: Commercial Property Market Yield, Capital Return Models .9: The Evaluation of Accuracy of Competing Models .10: Model/ Data Dichotomy of Four Classes of Uncertainties .11: The Transition of Levels of Uncertainty from Determinism to Total Ignorance .12: Knowledge as Measurement and Theory .13: Probability of Exceeding Multiples of Sigma .14: A Fractal Law with a Tail Exponent (α) of 2 .15: Types of Disasters.16: Comparison of World Natural Catastrophes in the First Half of 2017 .17: Sigma Event Selection Criteria for 2016 .18: Place and Space Risks of Black Swan Events .19: Four Quadrants of Decision making .1: Methodological Framework for Achieving Research Objectives .2: Alternative Knowledge Claim Positions .3: Major Differences between Deductive and Inductive Approaches to Research .4: Alternative Strategies of Inquiry.5: Quantitative, Qualitative and Mixed Methods Procedures .6: The Differences between Sampling in Quantitative and Qualitative Research .7: Procedures in Quantitative and Qualitative Data Analysis .1: The List of Economists in the AFR’s Quarterly Survey of Economists (2001-2011) .2: Market Share of Australian CBD Office Markets - July 2011.3: Unit Root in Level .4: Unit Root in 1st Difference .5: Economic Forecast Accuracy Based on Scaled-dependent Metrics – 3M Vs 6M .6: Rental Movement-Prime Forecast Accuracy Based on Scaled-dependent Metrics .