System Dynamic Approach for Analyzing Cyclic
Mechanism in Land Market and Their Effect on
House Market Fluctuations
Hanieh Mohammadi ,Reza Kazemi , Hesam Maghsoudloo , Erfan Mehregan and AliNaghi Mashayekhi
Abstract—In this article we have developed a simple dynamic
model to portray a cyclic producing mechanism in land market
and in the following we have probed the effect of land market
oscillation on house market price, which has not been addressed
before. In this model, in the beginning two cyclic produc-
ing mechanisms (including Speculation Effect and Cumulative
Mechanism) in land market are introduced and their effect on
house market is elaborated in detail. As we combine the land
market model with house market model, to develop an integrated
mode! that offers better understandings of house market trends.
The model showed that, in contrast to common perception
which presumes house market fluctuations (in demand sector)
to be totally intrinsic, they are mainly due to cyclic producing
mechanism in land market. Our work uncovers the rich dynamic
complexity of the real estate system and can serve as a good
example of applying systems thinking principles to complex real
world problems. Moreover, we have taken advantage of classic
mass-spring systems, to model the house and land market and
thus a simple powerful tool is introduced to predict the effect
of various mechanisms affecting the house market (e.g. capital
market)and it can be a great help to understand the complex
house market system, in more depth.
Keywords. House Cost, PortFolio, Investment, Inflation.
I. INTRODUCTION
The housing market intrinsically leads to a cycles (Malpezzi
2004) that can be observed in the economics trend of al-
most all countries all over the world (harris 2003, Bertrand
renaud 1995). These cycles are a major of special interest of
investors and researchers, since the cycles affect the business
and commercial cycles significantly(Fred E. Foldvary 2003),
besides they also influence the return of investment (ROI)
well as economic success and failure (Phyrr 2003). The last
but not the least, these cycles have an impact on investment
circumstances, banking, government policy and social (con-
ditionsMalpezzi 2004,Weiss 1991). Contemplating all of the
aforementioned factors, it is crystal clear that analyzing these
cycles is absolutely crucial. Various approaches to housing
issue along with different results from recent researches in this
field, manfiests the complexity of the topic (Grisson & Delisle
1999). The intricacy stems from housing market structure as
well as several factors affecting the market price (yean pin lee,
1996).
One of the influential factors, whose price remarkably affects
house price, is construction land (meikle 2001). Moreover
Department of Management and Economy Sharif Uni-
versity of Technology, Tehran, Iran, E-mail: haniehmoham-
madi @ gmail.com,rezakazemi @ieee.org,maghsoudloo@ gmail.com
there is close correlation between land demand and house de-
mand (Yean pin lee ,1996), as a result, a deep accurate insight
about land market can be an effective tool for economists and
managers to make better decisions in the field of developing
and controlling the land market. This knowledge can be a
great help in the process of land allocation and increasing the
land density in some vicinities (yean pin lee 1996). Therefore
land is on of the most essential factors in house market.
Dynamic Analysis of land market and house market is really
complicated, however the most effective way to analyze a
complex system is taking advantage of a dynamic model to
simulation the real world and probing the model outcome.
This method is effective mainly because the aforementioned
systems is described by high order non linear equations and no
scholars is able to solve these equations without simulations,
the simulation help the scholar the grasp the interrelation
between various elements in the systems (forrester, 1991). The
main focus of the literature is on dynamic modeling of house
market and the process of shaping house market cycles. to
name a few, Investigation of speculation intensifying effect
on house market cycles (Malpezzi,Wachter, 2003), studying
the fluctuating behavior in rent market (wheaton, 1999) and
investigation of cycles in house ownership market and their
causes, and also the mutual relation between rent market and
ownership market and its effect on house price cycles. In this
article, first of all we analyzed the dynamics of land market
and we will elaborate on its Cyclic Producing Mechanisms
(CPMs) . In the following, the effect of these mechanism in
land market variables is investigated (no previous article has
address this issue). Afterwards, land market along with house
market is dynamically analyzed and the land market CPM,
which plays a great role in land market cyclic behavior, is
closely investigated. At the end, the effect of the land market
CPM on house market parameters is studied and the effect of
ignoring them in some policies is clarified.
II. LAND MODEL
A. Structure
The model of land market is demonstrated in the figure 1. It
consists of 3 main sectors including supply, demand and price
that will be elaborated in the following Supply Sector, Since
required land for construction is majorly formed of time-worn
houses, land supply is solely provided by the land of these
houses. Demand Sector, it comprises of two main parts
¢ Construction Demand.
st cost
ar
Average Land |
“Labous aud Stutl
pies per meter quate
Fig. 2: The effect of speculation on land price fluctuation
¢ Speculative Demand.
The price of the house determines the total demand for
house. Multiplied by a coefficient, the total demand settles
the construction demand. The expected return of keeping the
house, determined by land price and its growth trend, decides
the speculative demand. Price, the supply-demand mechanism
regulates the price. In this model, congestion is defined as the
number of built floors and it is assumed to be constant.
B. Outputs and analyzing the model
1) The effect of speculative mechanism on land price fluctua-
tion: In the beginning we will inspect the effect of speculation
on land market cycles. In this phase, we assume house price
to be constant to eliminate its probable effect on land price.
As it is can be observed on Figure 2, the positive loop -which
is bold in figure 2- leads to oscillation, in the following we
will discuss this loop in more detail:
Suppose that the land supply decreases. Consequently, the
price rises and the rate of price increase goes down. As a
result, the expected return of maintaining the land rather than
selling will also falls. (Expected Return is defined as the ratio
of the price growth to the price, the greater the ratio, the greater
the desire for speculation.) This decline will decrease the
speculative demand and therefore the total demand will drop
and thus the land price diminishes. The opposite story happens
as the demand rises, in that case the price rises consequently
and this cycle begins all over again.
2) The effect of cumulative mechanism on supply and
demand: Now, we formulate a hypothesis of 4 phases which
explains the cyclical behavior of price in the absence
of supply lag and on the basis of interaction of theses
two mechanisms and dominance shifts between them. As is
clear from the following hypothesis, neither accumulation
mechanism nor price mechanism is a CPM. But their
combination, to which can serve as a CPM.
I) At the outset, demand is equal to supply and the price is
high, so it is lucrative for suppliers to supply even more. As
the supply increases, the cumulative supply increases and as a
result the price decrease, however the price still remains high
enough to encourage suppliers to increase the supply and this
will continue until sell rate become equal to destruction rate
market price of land
aM
aM
UU !
oc 40 8 12) 160 200 240 280 320 36) 400
‘Time (Year)
rmatket pie ofland Land pice with «psculatire
snatket pice of aad: Lens pre without seecuaive
a
Fig. 3: The effect of speculation on land price fluctuation
speculative land_supply, demand and price
2
o 5 1 1 2 2 30 35 «40 45 50
Tae (Year)
supply Land pve wth spective
{lel demand : Land price mith speculative
snatket price of aac! Lend pre with spaculekirs —
Fig. 4: The effect of speculation on land price fluctuation
of time-worn houses.
II) in this phase the price still declines and thus sales take
overweighs destruction and therefore supply will decrease
until supply becomes equal to demand (The cumulative
mechanism rules).
TIT) in the beginning of this phase demand equals to
supply and the price is at the minimum level. Since the
destruction rate is still less than sales rate, the supply keeps
decreasing and due to low price, the demand will grow and
it accumulates. Demand rise will result in rise in price. On
the other side, because the ratio of house price to actual cost
decrease, consequently the demand for the land decreases,
however the former factor which is rise in demand because of
low price, is more dominant and thus the price keeps rising.
IV) In this phase the price is growing and this in turn will
lead to escalation of supply. As supply increases the ratio of
destruction rate to land sales will decrease and therefore the
demand will drop. At the end the demand and supply will
finally reach each other and the price will reach to its highest
level. It is exactly similar to the first phase and this cycle will
begin all over again. .
in this phase we can see the same state like first one and it
can continuously make cyclic behavior ,that we named them
accumulative mechanism in land market, when these behaviors
Average
Aggregation
°
Building \Land Ready to
Land rate 4, Sell
normal gain,
<Avaerage Land
Area>
land demand from
house builders
<Price> > target houses to be
constructed
normal demand k
4 for land by
ratio unit cost
° °
Average Land
Area
., Labour and Stuff
land
expected gain |
of land
rate6
prices per meter square .
Fig. 1: Structure of Land Market
act with the same mechanism on house market(mashayekhi
2009) ,we encounter very complex system .
3) the effect of variables on system behavior: :
1) The first variable is Average Area which used for each
house:
This variable has 2 major effects on land price:
« Decreasing the average price: In the model, given the
congestion to be constant, decreasing the average area
simply leads to increase in the ratio of land supply to
land demand. Therefore the average price will decline.
However, if we define congestion as the number of floors
(This definition is currently used in Iran) as it can be
elicited from Figure 5, the average price will increase.
Damping the Oscillations altitude: Increasing the av-
erage price, which was explained in previous section,
causes the expected return to decline and consequently
it discourages the speculation. Thus the cumulative oscil-
lator mechanism diminishes and the altitude of oscillation
will drop.
2) the second variable is The ratio of land’s vacant time to
the house durability:
This variable shows how long it takes for a vacant land to be
utilized in construction cycle again. In other words, it reflects
the land turnover in the market. In contrast to the hypothe-
sis put forth in (Mashayekhi-Ghili 2009) that claims house
durability has an inverse relation with oscillation altitude, as
it can be elicited from Figure 7, if we let house lifetime be
constant and increase the vacant time of the lands, the altitude
of oscillation will reduce. Thus, in the house market it is
important to notice that destructing time-worn houses by itself
is not an effective policy and decreasing the speculation in land
market should also be taken into consideration.
normal price for
market price
market price of land
-
750,000
500,000 Ah
aman [ALN \ |
NON UU
oO
i} 60 80 100 120 140 160 180 200
‘Time (Yeer)
macket psice often -2 50
racket price offand 2150
‘maske: price ofland :1 50
racket price ofland :1 159
Fig. 5: Market Prices of Land Versus Time
Fig. 7: Market Price Of Land
3) the third variable is the effect of price elasticity of supply
on land market oscillations :
In land market, as the price elasticity of supply increases, the
difference between supply and demand level increase and it
intensifies the fluctuation. It is worth mentioning that just like
in house market, In land market, the cumulative mechanism
leads to unpredictable reaction of market to the elasticity,
however, because the speculation CPM effect is dominant, the
cumulative CPM effect is not noticeable. we can see this effect
on Figure 8,9.
III. HOUSE AND LAND MARKET
A. House Market Structure
We have utilized the model of ghili and Mashayekhi for the
House market model and it is connected to land market model.
The structure of this model is similar to the Wheaton’s rental
model but there exist certain differences between them. Some
are simple and small. For instance delay in construction is
modeled as a first-order delay rather than a fixed delay. In
addition, in this model, the price adjustment process has
been modeled completely in a different way and price is not
adjusted by supply and demand immediately. Rather, the price
is modeled as a stock variable whose flow is determined by
demand-supply ratio. If supply and demand are equal, the flow
is zero and if demand-supply ratio rises, the flow increases
(Mashayekhi, 2006; Sterman, 2000).
But the main difference between the two models is concerned
with the stock-flow structures. This difference arises from par-
ticular characteristics of owner-occupied market, as a durable
goods market, which are absent in rental market. Goods traded
in owner-occupied market are basically different from those of
rental market. Goods traded in rental real estate markets are
not real estate but the use of real estate for a certain period
of time. Seeking simplicity we can say one year use of real
estate. This is not durable goods, because it endures less than
one year. The very houses are supplied and traded.
Consequently, in such a market, real estate is transferred from
sellers to households. On the basis of this transfer, there is a
stock variable named ”Occupied Houses” whose flow, "sales
rate”, transfers houses from Vacant Houses” into this stock.
sales rate” is equal to demand divided by transaction time
when there is not supply shortage in the market. When supply
(i.e.’Vacant stock”) is less than demand, sales rate is equal
to supply divided by transaction time. One can formulate this
logic using a fuzzy minimum. This is the difference between
stock-flow structures of owner-occupied and rental markets.
Here, supply and demand are stocks in nature. Supply is equiv-
alent to vacant houses stock, whose flows are “construction
completion rate” and sales rate”. Demand is a function of
“homeless families”, which is determined by subtracting a
constant (i.e. all families) from a stock (i.e. occupied houses).
Therefore, “depreciation rate of houses” and "sales rate” are
flows which change the number of homeless families. Such
a stock-like nature can result in the accumulation of supply
or demand over a period of time. If construction completion
rate is greater than sales rate, ” Vacant Houses” (i.e. supply) is
accumulated. On the other hand, if sales rate is less than de-
preciation rate of houses, homeless families are accumulated.
In this model, for the purpose of simplification, it is supposed
that there is no speculative demand in the market, and like the
rental market model (Wheaton, 1999), sellers supply all of the
houses regardless of price. In addition, population is supposed
to be constant, but including depreciation in the model is
equivalent to incorporating population growth or some trend
in demand (Wheaton, 1999).
market price of land
2M
0
0 40 80 120 160 200 240 280 320 360 400
Time (Yeas)
ke pric of nd shih ice cy, 050 >
‘ake reo ad I pe ey, 050 d= 10
Fig. 8: Tension Versus Price with assumption that ratio of substance
to steady time is big
B. House and Land Market Structure
Linking the two models for house and land market (which
were introduced in previous parts) together, we will reach to
an aggregated model that has 3 main sectors. These 3 sectors
are as follow:
market price of land
0 10 20 30 40 50 60 70 80 90 100
Time (Year)
Fig. 9: Tension Versus Price with assumption that ratio of substance
to steady time is small
« The rate of time-worn house depreciation divided by
congestion coefficient will determine the inflow of ready
to be sold lands. The congestion is assumed to be
constant in this model to understand the internal oscillator
mechanisms in the market. In prospect studies, one might
consider population growth in the model to make it more
accurate.
Built lands multiplied by congestion coefficient will de-
termine the construction start rate.
« The ratio of house price to actual unit cost ( actual unit
cost is calculated by multiplying the average area of lands
by sum of 1 square meter land price and cost of labor
and equipment is an influential actor in land demand.
C. Analyzing Model output
1) the effect of land speculative cpm on house market:
In this part in order to fully understand the effect of land
market oscillation on house market, we have decreased the
house lifetime to lessen the impact of cumulative loop, which
was discussed in Ghili’s paper. As it is illustrated in Figures
10,11 although we have eliminated the house market oscillator
elements in the model, house market prices still fluctuates.
In the following we will elaborate more on the causes of
this behavior. The positive and negative loops that cause the
oscillations are depicted in Figure13 with different thickne:
2) the effect of land market on influential factors is house
market: effect of the 3rd effective house market CPM on
supply price elasticity:
One of the interesting functionality of the 3rd CPM in house
model is reducing the sensitivity of the house model to price
elasticity of demand. To state the matter in a different way, in
low price elasticity range, the model does not show oscillation
and changes in model fluctuation due to changes is price
elasticity only occurs in high value of price elasticity. This
clearly shows that the 3rd CPM in house market, which
is formed by adding the land market model, is far more
influential than the mechanisms ruling in land market alone.
As a result, if we neglect the land market effect in developing
the policies for rent market or house market (which are said to
be interelated [Ghili’s Paper]) we may make a huge mistake
supply-demand-price
6,000
40M
3,000
30M.
o
20M.
demand : house and land
vacant stock. house and land
Price : house andl land
Fig. 10: Supply demand Price
housing & land price
o 5 0 15 2 25 3% 35 a0 45 SO
Tine (Peer)
Price house and land
market price of land : house and land
unit cost: house and land
Fig. 11: Housing and Land Price
and by changing a parameter such as construction rate (which
reflects the price elasticity of demand) we may not reach our
desired goal.
IV. CONCLUSION AND PROSPECT STUDIES
To give a better understanding of the insights of this model,
let’s start with the classic dynamic system, spring-mass
system. If a ma: attached to some parallel springs, it will
oscillate less than it would in the case that the mass was
attached to each of the springs. On the other hand, if the
springs are placed in series the system will oscillate more
intensely comparing to the case that the mass was connected
to each of the spring.
As it can be seen in parallel and series spring-mass systems,
when the oscillator elements (the springs in our example) are
directly connected to the oscillating factor (the mass in our
example), the altitude of oscillation will decrease, while in
the case of indirect connection (as in series connection) the
altitude of oscillation will rise. Now let’s get back to the house
market problem. Using the same rule, if we analyze each cyclic
producing mechanism individually and its connection to house
market (i.e. Direct or Indirect) we can predict the combined
effect of all CPMs together. For example, two CPMs (Supply
Lag and durability mechanism) which are inside house market,
are both connected directly to the house market. On the other
hand, land market oscillation is serried with one of the house
Tard
Demand
Fig. 12: The Structure That Generate Our Third CPM
market CPMs (Supply Lag). Given the spring-mass metaphor,
if we deactivate the two house market CPMs, the land market
CPM will be connected directly to the house market and still
there will be oscillation in house market, and it is verified
by the model results. This finding introduce a new CPM for
house market, that was neglected in previous researches.
In land market just like in house market, cumulative mecha-
nism along with speculation effect leads to oscillation in the
market. These mechanisms in the land market, may affect
other variables in this market and misapprehending of these
variables may lead to mistakes in establishing policies. Fur-
thermore, land market oscillation has nonlinear effect on house
market fluctuation and thus it may mislead us in recognizing
the sources of these oscillations. Hence, it is inevitable to
analyze house market along with land market and taking
land market into consideration in developing policies for rent
market as well as house market is absolutely vital.
Finally, the last but not least, Analyzing house market is really
complicated and there are several effective factors in this mar-
ket. Analyzing all of these factors is absolutely sophisticated
and arduous. In addition to all mechanisms and factors that
were discussed in this article, there are many other factors and
markets such as Investment market that affect house market,
and they all dynamically changes and these changes may affect
the house market as well.
REFERENCES
u
2
DiPasquale, D. and W. Wheaton (1994), Housing Market Dynamics and
the Future of Housing Prices, Journal of Urban Economics, 35: 1-27.
Wheaton, W. (1999), Real Estate “Cycles”: Some Fundamentals, Real
Estate Economics, 27(2): 209-30.
‘Ali Naghi Mashayekhi, The impact of exchange rate policy on inflation
rate in an oil-exporting economy.
‘Ali Naghi Mashayekhi, Transition in the New York State solid waste
system: a dynamic analysis.
Jac A. M. Vennixa Jay Wright Forrester Prize Lecture, Group model-
building: tackling messy problems , 1999.
Yaman Barlas and Stanley Carpenter,Philosophml roof8 of model vali-
dation: two paradigms.
Barry Richmond Systems thinking: critical thinking skills for the 1990s
and beyond.
John D.Sterman,System Dynamics Modelling : Tools for learning in a
complex world,
B
(4
(5
(6)
7
(8
9
[10]
uu
U2]
U3]
James M. Lyneisa, System dynamics for market forecasting and strue-
tural analysis .
James M. Lyneis,a Kenneth G. Coopera and Sharon A. Elsa, Strategie
management of complex projects: a case study using system dynamics
George P. Richardsona, Problems in causal loop diagrams revisited.
Sterman J. D., Business Dynamics: System Thinking and Modeling for
a Complex World, Boston, MA, McGraw-Hill Companies, 2000.
Jay W.Forrester, System Dynamics and the Lessons of 35 Years, MIT
Working Paper, 1991.
total number of
famillies
pe homeless
familly
‘ i
demand
Average
\ familly
7 Aguregation
sf
average life
under a By] vacant loceupied
constracti ak stock ;
contdiction ‘on Stock cocstrattion stock | saidates |_stock, depfiate under constraction
star] rate completion\rate Stock:
constraction
transaction
time
time
cenpied
[git
Average
stock
Ye
total honsing
Aggregation
Building and Ready to
Land Sell scant
“Ss gtogk
__— normal price for
land
nonmal gain
Avaerage Land
Area
kpeculative
demand
total demand
Jand demand from
house builders
expected gain
Puce —— target houses tobe tale
fo > constmeted
market price
of land
Pe
noumal demand 7
for land '
ratio. Be geil cost
Average Land
Area
Labour and Stuff
prices per meter square
Fig. 6: Land and House Market Structure
