Fundamentals for automating due diligence processes in property transactions
| Pages | 97-124 |
| DOI | https://doi.org/10.1108/JPIF-09-2019-0130 |
| Date | 29 April 2020 |
| Published date | 29 April 2020 |
| Subject Matter | Property management & built environment,Real estate & property,Property valuation & finance |
| Author | Philipp Maximilian Müller,Philipp Päuser,Björn-Martin Kurzrock |
Fundamentals for automating due
diligence processes in
property transactions
Philipp Maximilian M€
uller
Department of Civil Engineering,
Unit of Real Estate Studies, Technische Universit€
at Kaiserslautern,
Kaiserslautern, Germany
Philipp P€
auser
Architrave GmbH, Berlin, Germany, and
Bj€
orn-Martin Kurzrock
Department of Civil Engineering,
Unit of Real Estate Studies, Technische Universit€
at Kaiserslautern,
Kaiserslautern, Germany
Abstract
Purpose –This research provides fundamentals for generating (partially) automated standardized due
diligence reports. Based on original digital building documents from (institutional) investors, the potential for
automated information extraction through machine learning algorithms is demonstrated. Preferred sources for
key information of technical due diligence reports are presented. The paper concludes with challenges towards
an automated information extraction in due diligence processes.
Design/methodology/approach –The comprehensive building documentation including n58,339 digital
documents of 14 properties and 21 technical due diligence reports serve as a basis for identifying key
information. To structure documents for due diligence, 410 document classes are derived and documents
principally checked for machine readability. General rules are developed for prioritized document classes
according to relevance and machine readability of documents.
Findings –The analysis reveals that a substantial part of all relevant digital building documents is poorly
suited for automated information extraction. The availability and content of documents vary greatly from
owner to owner and between document classes. The prioritization of document classes according to machine
readability reveals potentials for using artificial intelligence in due diligence processes.
Practical implications –The paper includes recommendations for improving the machine readability of
documents and indicates the potential for (partially) automating due diligence processes. Therefore, document
classes are derived, reviewed and prioritized. Transaction risks can be countered by an automated check for
completeness of relevant documents.
Originality/value –This paper is the first published(empirical) research to specifically assess the automated
digital processing of due diligence reports. The findings are helpful for improving due diligence processes and,
more generally, promoting the use of machine learning in the property sector.
Keywords Artificial intelligence, Machine learning, Property transactions, Due diligence, Digital building
documentation, Document classes
Paper type Research paper
1. Research background and previous research
Information is key for making right decisions and achieving corporate goals (Bhatt, 2001,
p. 70). Property is changing from an asset-intensive sector to an information-intensive sector.
The majority of property data are collected in daily operations, e.g. regarding maintenance,
letting, services charges or energy consumption. However, there is a fundamental lack of
Due diligence
in property
transactions
97
This paper forms part of a special section “PropTech and Entrepreneurship - Innovation in Real Estate
II”, guest edited by Dr Larry Wofford, Dr David Wyman, Dr Elaine Worzala.
The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/1463-578X.htm
Received 30 September 2019
Revised 21 January 2020
3 March 2020
Accepted 6 March 2020
Journal of Property Investment &
Finance
Vol. 39 No. 2, 2021
pp. 97-124
© Emerald Publishing Limited
1463-578X
DOI 10.1108/JPIF-09-2019-0130
instruments for leveraging internal data and a risk of overlooking the essentials (Bawden and
Robinson, 2009, pp. 182-184).
Traditionally, many property investors ground their decisions on a combination of
intuition and retrospective data (Asaftei et al., 2018). Although property investments are
generally assumed to follow a rational process many cognitive limitations may and do apply
(Salzman and Zwinkels, 2017, pp. 81-83). Reports may be subject to overconfidence or
confirmation bias, especially if the underlying data are difficult to access (
€
Ohman et al., 2013).
Using available digital documents in the best possible way may help to obtain an unbiased
view of a property investment. A global commercial property transaction volume of more
than US$800 billion in 2019 (Jones Lang LaSalle, 2020) underpins the relevance of receiving
building information at the right time, in the right place, in the right format, in the right level
of detail and above all, the right information (RICS, Royal Institution of Chartered Surveyors,
2017). It is still particularly difficult to assess the physical structure of a building due to
shortcomings in data structure and quality (Wouda and Opdenakker, 2019, pp. 573-576).
Different interests, systems and data requirements lead to regular building documentation
losses over time (Bodenbender et al., 2019, p. 177).
Due to the growing amount of data and the use of new technologies, it is now possible to
make investment decisions more rational. For transactions, documents are usually made
available in data rooms that ensure availability, timeliness and consistency. Conventional
data rooms are used just for the purpose of a transaction. To use information further for
building operation, there is an ongoing development from pure transaction data rooms to
permanent data rooms. The aim is to provide information and documents without media
breaks as well as to identify information unambiguously over all life cycle phases of a
property (Bodenbender et al., 2019).
Due diligence reports in property transactions usually relate to tax, technical,
environmental, legal and commercial issues. Technical due diligence especially focuses
on the assessment and evaluation of the structural condition and maintenance
requirements (RICS, Royal Institution of Chartered Surveyors, 2009). Since construction
and maintenance are associated with substantial costs, identified risks or doubts may
trigger purchase price discounts, guarantees or even termination of negotiations.
Generally, in just a few weeks a wide range of documents must be elaborately prepared
and made available. Gaps in the documentation can only be identified with great effort.The
quality and completeness of digital building documentation is increasingly becoming a
factor as “deal maker”and “deal breaker”in transactions. On e of the greatest potentials of
artificial intelligence (AI) in property relates to services that support reporting (EY Real
Estate, Vonovia, 2019).
Artificial intelligence can well assist in the classification of documents as key for the
structuring of a building documentation. This requires a distinct set of document classes.
Classification refers to the assignment of a document to a (predefined) class. Ideally, every
document is assigned to one single document class. Previous achievements in automated
classification already help to recognize, classify, name and sort digital documents in digital
data rooms. Algorithms used are mostly based on machine learning (ML) and natural
language processing (NLP) techniques. With different classification algorithms (Naı€ve Bayes,
Support Vector Machine, Deep Learning) the error rate in recognizing and interpreting text in
documents could continually be reduced. With methods of supervised learning, hit rates
above 80% are already achieved (Bodenbender et al., 2019;Russell and Norvig, 2016). The
next step is information extraction. The goal is to recognize and extract predefined values and
information such as entities or relationships (Jurafsky and James, 2019). In the long run,
artificial intelligence could extract relevant information from documents to assist investment
decision making.
JPIF
39,2
98
To continue reading
Request your trial