Applied Quantitative Analysis for Real Estate

Sotiris Tsolacos, Mark Andrew

• Editorial: Routledge
• Año: 2020
• ISBN:
  9781351359009
• Edición: 1
• Categoría: Business
• Categoría: Contabilità finanziaria

To fully function in today's global real estate industry, students and professionals increasingly need to understand how to implement essential and cutting-edge quantitative techniques.

This book presents an easy-to-read guide to applying quantitative analysis in real estate aimed at non-cognate undergraduate and masters students, and meets the requirements of modern professional practice. Through case studies and examples illustrating applications using data sourced from dedicated real estate information providers and major firms in the industry, the book provides an introduction to the foundations underlying statistical data analysis, common data manipulations and understanding descriptive statistics, before gradually building up to more advanced quantitative analysis, modelling and forecasting of real estate markets.

Our examples and case studies within the chapters have been specifically compiled for this book and explicitly designed to help the reader acquire a better understanding of the quantitative methods addressed in each chapter. Our objective is to equip readers with the skills needed to confidently carry out their own quantitative analysis and be able to interpret empirical results from academic work and practitioner studies in the field of real estate and in other asset classes.

Both undergraduate and masters level students, as well as real estate analysts in the professions, will find this book to be essential reading.

1
Introduction

The focus of this book is the application of quantitative techniques to real estate. The material presented takes the reader from basic statistical analysis to more advanced topics addressing a range of quantitative methods employed both in real estate education and the workplace. The themes in this book are presented in an applied manner. Throughout the book we provide examples to illustrate the statistical concepts in the context of the real estate market. The online resource for this book contains further applications. The aim of the book is to make the reader confident with the application of quantitative techniques to real estate.

1.1 Motivation and rationale for this book

We highlight four inter-related trends in the real estate field that necessitate the existence of a textbook on the applied quantitative analysis of real estate markets.

• (i) The application of statistical tools to analyse data is a much-sought skill in the real estate business. It is increasingly becoming an integral part of real estate education. Students in real estate degree programmes are expected to have at least a fair knowledge of basic statistical techniques. A good grasp of statistical analysis is helpful for the study of other subjects such as real estate investment, appraisals and portfolio management.
• (ii) It is recognised that the real estate market interacts with the economic and broader investment environments. Quantitative analysis will assist us to quantify relationships and test them empirically. Such analysis varies from simple descriptions of features of the data, examining correlations to constructing a variety of econometric models. It opens up a greater range of options in the empirical investigation of real estate markets.
• (iii) The recognition of real estate as a mainstream asset class poses challenges to how analysis in real estate markets is conducted. Investors in other asset classes are accustomed to the application of quantitative analysis and would expect similar practices in the real estate market.
• (iv) The availability of data is growing. Databases are getting larger and becoming more readily available. Universities are increasingly gaining access to more databases, including proprietary data from firms and organisations. A good background of quantitative analysis enables students, researchers, analysts and others to utilise the growing availability of data.

Universities have incorporated quantitative research techniques into their real estate programmes. Modules covering simple data analysis and statistical modelling are part of the curricula, depending of course on the nature of the programme. Real estate investment and finance programmes will incorporate more advanced analysis whereas the more traditional real estate programmes will contain essential quantitative techniques. This book is motivated by the needs of both groups of students either at the undergraduate or postgraduate levels. A number of these programmes are conversion courses, usually taken by non-cognate students who may come from a subject area with little statistical background. This book is intended to bring these students up to speed with the application of quantitative techniques to real estate data analysis. The book aims to facilitate the development of these skills. Students will become familiar with the most commonly used techniques in practice and will be well equipped to directly carry out empirical work.

The work in this book is also inspired by the quantitative needs of real estate analysts in the industry. Analysts have access to large datasets. Some of them may have to brush up their quantitative knowledge. A quantitative background gives them flexibility to undertake their own work and investigate relationships using the wealth of data available. Ability to make sense of the data and carry out empirical analyses is a valuable skill for a professional in the real estate field.

1.2 Broad themes covered in the book

Quantitative analysis comprises a large set of mathematical and statistical procedures and tools for factual analysis. This book is not intended to present any branch of quantitative analysis that is potentially relevant to real estate. This would be an onerous task. The book focuses on themes most prevalent in real estate and presents them in detail and in a practical way. The book helps the reader build the necessary background for further quantitative analysis. In this section we give a short outline of four themes covered in this book. They are data manipulation, probability and inferential statistics in real estate, applied econometric modelling and forecasting.

1.2.1 Data manipulation and summarising information

Understanding and summarising data to reveal systematic patterns is informative. Information in its raw form may not be appropriate to conduct analysis or make comparisons and can be difficult to comprehend. Consider, for example, rents recorded in different units (e.g. local currency per square metre per year vis-à-vis in local currency per square foot per month). Or for that matter, rents obtained from countries with low and high inflation. Data transformations help us to understand the information they contain about the market. Rents should be expressed in similar units, or one should produce an index or adjust them for inflation. It is valuable to understand the distribution of the data, obtain statistics such as the mean and standard deviation and use techniques such as correlation analysis to study relationships between data series. Think of a situation in which we find that rents in market A are more strongly associated with the economy than in market B. What are the reasons behind this?

1.2.2 Probability analysis, inferential statistics and applications

Analysing the market entails dealing with uncertainty that arises from the complex nature of the real world. Real estate is subject to uncertainty of the same nature and intensity, as for instance in the stock market. Following data transformations and rudimentary statistical analysis, we introduce the concepts of probability, expected values and hypothesis testing. They enable the analyst to draw conclusions reflecting the most probable outcomes. Probability distributions are the basic statistical framework for undertaking such analyses. Suppose when we study house price growth we observe that the most common growth rate values lie between 4% and 6% per annum. There are occasions, however, when house price growth can be negative, say −10%, or strongly positive, +20%. These are rare occasions, but they still occur reflecting major economic events in the market. By studying the distributional properties of the data, we are able to make useful inferences about expected values and probabilities of them occurring.

Related to this topic is sampling. Usually we work with samples from a larger and often unknown population or have a preconceived idea about the data generating process (DGP). This concept is not easy to grasp at first, a fact we acknowledge. Results reported from calculations are sample-based results with which we infer conclusions for the unknown or unmeasurable population. For example, when we work with 40 observations of rent data, we assume that this is either a subset of the unknown population of rents or the unknown DGP. If our objective is to estimate, for example, the mean value of house price growth out of sample estimates, we can utilise the concepts of a sampling distribution to determine the likelihood of the sample estimates reflecting the population or DGP value. Again, any inferences will involve the language of probability. These concepts are additionally used in hypothesis testing and modelling relationships.

1.2.3 Model building and applied econometric analysis

The book proceeds to employ and present formal techniques to study relationships within the real estate market and between the real estate market and the economy and investment environment. The complexity of the real estate market, its linkages to the economy and the importance of real estate in credit and investment markets have necessitated closer study of the dynamics of the real estate market. This market has taken a major part in causing economic and financial crises; an example is the global financial crisis in 2008–2009. As a result, we have seen an explosion in the use of quantitative analysis to explore how adjustments take place within the real estate market and measure its linkages to the external environment.

The relationships can be complex. We try to establish the factor or factors that influence a real estate variable. Housing construction can be influenced not just by economic growth but also by other influences including interest rates, construction costs and other. The empirical investigation of relationships will give evidence about prevailing theories and a priori arguments. It will be based on what happened in the past, will identify systematic influences and will use it to explain the relationship and forecast as we will discuss. These systematic influences are best uncovered by statistical techniques, as the human brain cannot work out exact systematic relationships. We will address a range of questions, for example, if interest rates rise by a percentage point, what is the expected impact on mortgage rates and eventually house prices? Such questions can be answered by regression analysis. We discuss the diagnostics tests to run to confirm the validity of the model. This background is used to extend the analysis to study relationships with more advanced tools.

1.2.4 Forecasting

Finally, we recognise that a definitive goal in the quantitative analysis of real estate is to make predictions and forecasts about the market. Explicitly addressing real estate forecasting is necessitated by the growing need for forecasts in the industry. The book exposes the reader to a variety of econometric models, forecasting techniques and assessment procedures that can help the researcher carry out forecasting work effectively. In the forecast section of the book, we highlight the limitations both of quantitative (or model-based) and qualitative (or judge-mental forecasting) and discuss how they can be combined. We also discuss how forecasting takes place in practice and offer our views about the process.

1.3 Book online resource

The online resource for this book contains accompanying notes to chapters with further examples. It also contains chapters on additional topics. Most of the statistical procedures in this book can be carried out in Excel. In the online resource we illustrate the procedures in the econometric package EViews, a software package common in the real estate industry. Datafiles are posted so that the reader can replicate the analysis and practice EViews and Excel.

2
Real estate data

2.1 Introduction

This chapter provides an overview of real estate data. Before dipping into the statistical analysis of data, it is appropriate to describe the main data series available to study and analyse the real estate market. It contains definitions of data, brief explanations of how key measures are calculated, sources and a number of graphical illustrations. We present the raw data. Data transformation is the subject of subsequent chapters. In data analysis and the model building process, the particular features of real estate data should be well understood so that quantitative analysis is not conducted in a vacuum or with ignorance of the meaning of real estate data. Appropriate interpretations can then be made.

The chapter is intended to be a primer for real estate data. It introduces data and data sources in the real estate market to both undergraduate and postgraduate students previously unfamiliar with real estate. An explicit coverage of real estate data is valuable for their studies, pursuit of research topics and triggering the use of additional data not reported in this chapter. A further objective is to illustrate the plurality of data to a broader audience from other fields (e.g. other real asset classes) who would like to monitor and study the real estate market or perform analytics with data from this market. This chapter is a short version of a fuller data document available in the online resource of this book.

In order to provide structure to this chapter, we break down the real estate market into its main components: user or occupier, development (building construction) and investment sub-markets or segments. Distinctive data describe trends in each of these submarkets. Data may be produced by different sources and the method of obtaining/compiling data and constructing metrics is not identical in each segment. To complicate things further, different series exist describing the same aspect of the real estate market. For example, there are different data series for rents.

We briefly discuss key data in the direct or private market (also known as the underlying market) and the unlisted market. An introduction to real estate investment trusts (REITs) and debt market data is contained in the book’s online resource for this chapter.

2.2 Segments of the real estate market

A simple framework of the real estate market proposed by Keogh (1994) is useful for our purposes (also see Tsolacos et al., 1998). It breaks down the market into three major parts: the user or occupier market, the investment market and the development market (Figure 2.1). The segments of the markets are also analysed in DiPasquale and Wheaton (1992), Ball et al. (1998), Jowsey (2011, ch. 17) and Pirounaki...