Property Booms and Busts: Retail Property and Office Yield Trends Compared

1
Property Booms and Busts Retail Property and
Office Yield Trends Compared

• Colin Jones and Michael White

2
Introduction

• Downturn following credit crunch saw investment
  funds dramatically dry up.
• Impact saw first yields rise then fall in
  development as it became unviable and unfundable,
  then rents fell
• Aftermath credit crunch highlights role of
  external influences
• Increase attention in the concept of asset
  bubbles
• This paper examines the contribution of bubbles
  by an analysis of urban office and retail yields
• It makes a comparison between these two property
  sectors

3
Office and Retail Yields London
4
Office and Retail Yields Edinburgh
5
Bubbles or Cycles

• Term bubble is inconsistent with the concept of
  efficient markets
• Notion of a bubble creates another debate about
  its relationship with the cyclical behaviour of
  real estate markets
• Can speculative bubbles occur independently of
  property cycles or do they amplify cycles?
• Potential for bubbles in real estate markets
  given inefficiencies

6
Existing Research

• Extended period of house price inflation across
  the world in the 1980s and especially from the
  mid 1990s stimulated academic studies
• Tested whether these trends can be justified by
  reference to fundamentals
• Only one by Hendershott (2000) of Sydney,
  Australia, has examined commercial property
• 11 years study period over one cycle
• Approach deviations of a fundamental value to
  replacement cost ratio and requiring an
  independent forecast of vacancy rates.
• Conclusions suggest the existence of a bubble

7
Property and Business Cycle

• Capital investment as the key cyclical driver of
  economic growth
• New development key influence on the business
  cycle
• Occupation demand is linked to business cycle
• Property market is inevitably cyclical
• But there is a further factor - development time
  lags

8
Stylised Commercial Cycles

• Development lags amplify economic cycle
• Studies conceptualised the property cycle by
  reference to the 'cobweb model' and the role of
  development lags
• Different types of real estate have different
  cycles reflecting different development lags

9
Barras Model
10
Internal Dynamics of Cycles

• Cycles amplified by internal mechanics
• Valuers or surveyors tend to over-price in booms
  (under-price in downturns)
• Pressures by developers to support their
  decisions based on current values or expected
  values derived from extrapolating the past
• Investment demand/ development via adaptive
  expectations gt prices beyond fundamentals
• Supporting element in this process is akin to a
  Ponzi scheme in that new investors/developers are
  attracted by the profits made by first investors

11
continued

• Processes can operate in reverse direction
• Simplification
• Banks financing development also support these
  processes and display lack of caution in upturns
• Arguably a collective under-pricing of risk
• Consequences are that construction /development
  activity at peak of the cycle becomes too high
  and unsustainable
• Downturn or slump is also greater

12
Bubble Processes

• Mechanisms overshoot price fundamentals -gt a
  bubble
• Parallels with bubble processes identified in
  financial asset markets
• Shiller -gt bubbles driven by feedback loop theory
  -gt exogenous shocks instigate price increases
  encouraging further investment demand through
  adaptive expectations.
• Property studies centre on housing - bubble as
  abnormal
• Shiller in the latter sees the bubble as a normal
  phenomenon/cycle - an exaggerated cycle

13
Bubbles

• Smith and Smith (2006) define a bubble as a
  situation in which the market prices of certain
  assets (such as stocks or real estate) rise far
  above the present value of the anticipated cash
  flow from the asset
• The rapid price rise brings speculative activity
  on expectations of future price increases rather
  than focusing on the assets cash flow or
  fundamental factors driving the market.
• In practice it may prove difficult to identify
  bubbles.
• Protagonists of bubbles implicitly assume that
  prices were equal to fundamental values
  historically before a period of rapid asset price
  inflation.
• Opponents argue that observed prices may have
  been too low and that increasing prices may have
  (at least in part) reflected an adjustment
  process towards long run equilibrium values.

14
Argument

• Unlike asset bubbles instigated by exogenous
  shocks property cycles driven by internal lags
• Shillers model of a bubble incorporating
  endogenous feedback driven by adaptive
  expectations generic cyclical model
• Barras cycle is special case that adds time
  lags
• Bublical model is assymetrical
• Bubbles are not one off occurrences but are a
  natural part of property cycles but that the
  adaptive expectations processes of amplification
  are greater in the upturn than the downturn

15
Empirical Research

• Continuous dynamic price system can be decomposed
  into a fundamental and bubble component with
  former present value of future income
• Himmelberg et al (2005) approach - examine
  housing bubbles via a user cost approach
• Assess whether property is over or under valued
• Essentially user cost if a function of the risk
  free rate of borrowing, expected capital value
  change (and income growth for commercial
  property) plus a risk premium
• Property taxes etc assumed zero

16
Some detail

• Within this framework in long run equilibrium
  capital value to income ratio should equal the
  inverse of the user cost, ie initial yield
• Dunse et al (2007) model extended - link to real
  property market via inclusion of key demand
  variable, finance and business services output
• Property markets are taken to be essentially
  urban with different rent cycles and have
  different associated risk.
• Based on the office markets of British cities
  over the period 1981- 2006.
• City level property data from IPD SWIP

17
Model

• Overall yield model can therefore be rewritten as
  
• IY f (?RRVI, GY, LEP, EA)
• GY return on long dated gilts and LEP average
  yield on the FTSE 100 index, ?RRVI current real
  rental growth
• Equation models Barras model although development
  lags are not explicit, implicit in expected
  rental growth
• Model is extended to include market fundamentals
  capturing the economic activity (EA) by value of
  finance and business services output (FBO) for
  offices and GDP or consumers expenditure for
  retail
• Estimated by panel time series multiple
  regression in logarithmic form.
• Long run equilibrium models for the cities

18
Logic of Results

• Fundamentals and expectations for yields are
  captured from the model
• Remaining mispricing may be attributed to a
  bubble
• Following estimation of this model, fitted values
  were calculated for each city
• Proportionate difference between actual yields in
  each city and those estimated from model
• Greater difference then larger over- or under-
  pricing of office real estate

19
Office Market
20
Retail Market
21
Percentage Differences between Actual and
Estimated Office Yields for Individual Cities
1981-2006
22
(No Transcript)
23
Percentage Differences between Actual and
Estimated Retail Yields for Individual Cities
1981-2006
24
(No Transcript)
25
Commentary

• Late 80s London more susceptible to bubble
  effects but overpricing across markets
• Actual prices rising above fundamentals towards
  the end of the time period before financial
  crisis apparent in both office and retail
  sectors and across cities
• Clear individual patterns over time presumably
  reflecting localised demand and supply
  expectations
• Asymmetry in these effects is perhaps less
  straightforward

26
Conclusions

• Empirical analysis shows substantial and
  systematic cyclical bubble effects
• Prices will always be over-priced in a boom and
  underpriced in a bust
• Differences in time lags between office and
  retail did not seem to significantly affect
  results
• Hence traditional exposition of the property
  cycle with its time lags needs to be augmented to
  encompass such bubble effects
• Barras property cycle as a special case that adds
  time lags to a more generic model of cycles
  centring on adaptive expectation stimulated by
  external shocks
• But careful interpretation needed as further
  analysis required to improve model performance