Ichiro Kawabata
Licensed real estate appraiser
Member of Japanese Association of Real Estate Appraisal
President
Kawabata Real Estate Research Institute Co., Ltd.
Mailing address; Ken-Keizai Center2F, 26, Nisimigiwacho, Wakayama City, Wakayama Prefecture
640-8227, JAPAN
Url; http://www31.ocn.ne.jp/~reap/
E-mail address; reap@bb.mbn.or.jp
1. Introduction
A few miles south of Soledad, the Salinas River drops in close to the hillside bank and runs deep
and green. The water is warm too, for it has slipped twinkling over the yellow sands in the sunlight
before reaching the narrow pool. On one side of the river the golden foothill slopes curve up to the
strong and rocky Gabilan mountains, but on the valley side the water is lined with trees."
This is a quote from 'Of Mice and Men' by John Steinbeck.
In Japan there is also an area as beautiful and sacred as this in California. We call this the
Koya-Kumano area in the Kii Mountain Range . This area was added to Unesco's World Heritage List of
"Sacred Sites and Pilgrimage Routes in the Kii Mountain Range" in 2004.
The Kii Mountain Range is located to the south of Kyoto and Nara, ancient capital cities that ruled
Japan for over 1300 years. The mountains occupy most of the area known as the Kii Peninsula, are
covered with a dense blanket of green forest and have been Japan's spiritual heartland through the
ages, a sacred place to where, it is said, the gods of Shintoism and Buddhism descended to reside.
I am a real estate appraiser in Wakayama prefecture where the Koya-Kumano area belongs. There is
Shirahama town on the south coast of this prefecture. This is nationally famous as a holiday resort
place. As of 2003, the town had an estimated population of 19,646 with a total area of 64.73 km2.
Shirahama is also known for its spa and its beautiful white beach. The main Shirahama beach is lined
with hotels like a mini Waikiki. Shirahama is about two and a half hours from Osaka by Japan Railway's
"Ocean Arrow". There are also daily flights from Tokyo's Haneda Airport. The flight is less than 1
hour. However, though this town is famous as a tourist resort, the land value standard is never high
as it belongs to the rural district in the prefecture. It is about 150,000 yen/u at the ceiling price.
I am also in charge of the fixed asset tax evaluation in this town. Although this is a small town in
this prefecture, it is very difficult to evaluate every household. Therefore we choose some standard
places and evaluate them first. The standard property points rise up to about 2,200 places.
To evaluate these many properties in a well-balanced fashion, the multiple regression analysis on the
statistics is very useful. I want to talk you about my impressions using this analysis.
2. The multiple regression analysis
The regression analysis is one of the methods of statistics to formulate a dependent variable by other
explanatory variable (the regression equation) and to analyze each other variable's relation. The
regression equation is expressed by the least squares method to make the difference (the residual)
between the predicted values and the observed values the least. The precision of this analysis is judged
by the coefficient of determination. In other words, if the data (the observed values) has a tendency,
it will be expressed by the linear equation as the following.
Y= a X + b
There is some estrangement between the predicted values gotten by this equation and the observed values.
This estrangement is called residual. The residual sometimes shows a positive value, other time it shows
a negative one. So it is necessary for us to look into the degree of the estrangement by squaring the
residual. It is called the least squares method to get a and b of this equation by minimizing the sum of
these squared residuals.
Then what is the coefficient of determination? It is closely related with the variance which shows the
degree of the scatter which the distribution of the data has. Doing a regression analysis, the important
terms include the following.
(1)Mean : The numerical value divided the sum of all data by the number of the data is called mean.
(2)Deviation: The difference between the data itself and the mean of the data is called deviation.
(3)Variance : The numerical value divided the sum of the squares of deviation of each data by the number
of the data (or n−1) is called variance.
If the meaning and usage of these terms can be understood, the regression equation can be set up and the
coefficient of determination can be grasped. 'a and b' in the regression equation and the coefficient of
determination can be expressed as the following.
As stated above the equation,  is the sum of squares of the residual. The lesser this
is, the closer the coefficient of determination to 1 and the more suitable this regression equation is
for the analysis. In this way it is called the least squares method to seek the regression equation as
the coefficient of determination is closer to 1. The multiple regression analysis is this application.
In this simple regression analysis the explanatory variable is only one but in the multiple regression
analysis the variables become more than one. The multiple regression equation is expressed as follows.
3. The gap rate
When we appraise a piece of real estate, we must analyze the value influences of the real estate. Above
all, while performing an analysis of the subject neighborhood, we must grasp the market-specific value
influences and quantify them numerically. The quantitative analysis is expressed as the gap rate of the
market-specific value influences which are useful to apply the appraisal approach, especially the sales
comparison approach.
The land value is generally considered the product of various value influences. Value influences are
divided into general value influences, market-specific value influences and property-specific influences.
By the way, why do we have to quantify the market-specific value influences numerically? Though there are
various market-specific value influences, they are generally classified into 4 groups in Japan. They are
called the street condition, the traffic approaching condition, the environmental condition and the public
administrative condition.
For example, the width and the kind of the street are the street condition. The character of the closest
railway station and the distance from the station are the traffic approaching condition. The building to
land ratio and the floor area ratio are the public administrative condition. We look into the real data of
these influences and substitute the gap rate for this data. We don't use the real data itself as the gap
rate because the land value does not show the same response to the real data's change.
For example, the ratio of two distances from the closest railway station, 50m and 500m, is expressed in
the pair of two numbers '1 to 10', using the common quantity 50m of the two. But it is not appropriate to
adopt these real data as the gap rate which is used for the explanatory variable. Because the adequate
ratio fitted to the explanatory variable may be expressed even in the pair of two numbers '1 to 20'. So
it is necessary for us to standardize the real data for the gap rate which is fitted to the appraisal.
Only the environmental condition is not shown as the numerical data but as the situation which is called
residential area or commercial area and so on. So instead of the environmental situation we adopt the
quantification method on statistics and make it the numerical rate. In this way all data are replaced
for the gap rates and are used to calculate the land value.
However, even if the gap rate is appropriate in the same value influence, it does not indicate how much
each condition influences the whole market-specific value. The multiple regression analysis is very useful
in solving this problem.
4. The multiple regression analysis and market-specific value influences
Though the multiple regression analysis is different from our appraisal, the analysis will be helpful to
us. The multiple regression equation is a linear expression. Replace all variables with logarithms. As the
linear expression means addition, it is easy to change addition to multiplication by changing all
variables with the logarithms. The land value is generally considered the multiplication of the value
influences. So it is necessary to change addition to multiplication. That is to say, it is very convenient
to compare the variables replaced by the logarithms with the value influences. The multiple regression
analysis will not be a good guide to our appraisal until the linear expression is replaced by the
multiplication.
According to the multiple regression analysis, the market-specific value influences on Shirahama town are
shown as the following equation. In the equation each variable is replaced with a common logarithm.
Even if each variable is expressed by the common logarithm, there is not any change in the tendency. That
is to say, if the explanatory variables are directly proportioned to the dependent variable, replaced
logarithms also have the same tendency. The coefficient of determination is also 0.91 and the accuracy of
this equation is very high.
Let's change this equation from the addition into the multiplication.
This equation closely resembles the following one which we generally use in case of the appraisal.
Let's pay attention to the difference of these two equations. Except for the point being made by the
logarithms, it is the most different point that each condition has a different exponent in this
multiplication which is replaced from the multiple regression equation. The left side of both equations is
also different, but as the market-specific influences are a part of the land value, we don't take care of
the difference. These exponents are called partial regression coefficients in the linear regression equation.
This partial regression coefficient shows the contribution to the dependent variable of each explanatory
variable. Since each explanatory variable means a group of the market-specific value influences, try to
examine each group closely. The partial regression coefficient of the street condition, the traffic
approaching condition and the public administrative condition is very small. On the other hand the
environmental condition's is very large. What does this mean?
This means that the street condition, the traffic approaching condition and the public administrative
condition are not as important as the environmental condition in the market-specific value influences.
The environmental condition has the most influence on the market-specific value. Of course, this is true
of Shirahama town which belongs to the rural area and is not necessarily true of any municipalities.
Why does the traffic approaching condition have only a small impact on the market-specific value
influences? It seems because of motorization. The importance of the closest railway station has become
smaller and smaller in recent years. Almost every home in the local area has a car and it is more convenient
to travel by car. It is also one of the reasons why there are fewer railway systems.
The commercial area in front of the railway station doesn't prosper as it did in the past. In the days
when the car was uncommon, the area prospered mainly because of the souvenir stores. In those days the
area functioned as the commercial center of the town. The closer the subject place was to the railway
station, the more business it had. This tendency is no longer true. This is one of the reasons that the
traffic approaching condition has only a small impact on the market-specific value influences.
You may be surprised by the fact that the ratio of the street condition in the market specific value
influences is smaller than you think. By Building Standard Law of Japan, a building can not be built in
the area if the width of the road is less than 4 meters. The gap rate must be very large according to
whether the width of the road is more or less than 4m in theory. But in this analysis it is not really
very large. The reason is considered that a part of the influence by the width of the road is included in
the environmental condition.
Though it is not so much as multicollinearity, there are few places in which the width of the road is
more than 4m in the farmhouse area in Japan. On the other hand, in the usual residence area where there
are many detached houses most of the houses face the street where the width is equal to 4 meters or more.
For example, suppose that there is a 1 point gap in the traffic approaching condition and there is also
a 1 point gap in the environmental condition. We tend to take it for granted that there is a 1 point gap
in either condition and each gap is on a level. However, as stated above, each gap is not equal in
Shirahama town because the ratio of the traffic approaching condition in the market-specific value
influences is much smaller than that of the environmental condition. We often compare an influence with
another one without noticing this difference.
5. The general condition of Shirahama town
If we misunderstand the market-specific value influences, it has a bad effect on the general value
influences and the property-specific value influences. So I will show you the characteristics of the
general influences and the property-specific value influences on this town.
There are a lot of leisure homes, resort condominiums and hot spring hotels as well as a large-scale
theme park which has giant pandas in this town. This town had been very busy as a tourist resort. But
since the collapse of the bubble economy this business has declined. Large Japanese companies used to
have many resort houses in this town. Now most of these houses have been sold and there are some
bankrupted hotels. Although there are some surviving hotels, they don't contribute any efforts to
reinvigorate the local economy. The guests of the hotels can receive every sort of service from the
hotels themselves. So restaurants and amusement facilities in the town don't have a lot of customers.
Shopping stores are open for business, but are doing virtually none at all even in peak season.
There are representative lands for publication of land price in Japan. There is one of them in the main
commercial area in this town. The transition of the land price is the following.
In Japan a tourist resort always includes a spa because Japanese like it. In this town there have been
many spas of very good quality, leisure homes with hot spring bathrooms used to be sold frequently. But
recently the sales have dropped off. Since the collapse of the bubble economy people have been very
careful to find spas of good quality.
It is very important for us to look into the property-specific value influences of the leisure home area.
The influence which has a negative impact on the value of the general residence area sometimes becomes a
positive advantage for the leisure home area. For example, it is an advantage for the leisure home if the
view of the leisure home is beautiful. Even if it is located close to the sea and may be flooded when a
storm comes, someone may buy it because he can easily go in and out by boat.
Because of the many inns and hotels, the commercial area in this town is declining, but the effort to
reinvigorate the local economy is being made slowly. For example the fisherman's cooperative of this town
has opened a store serving both as a restaurant and a gift shop and a spa intended for the day tripper has
also been started. Recently Nanki-shirahama airport has begun to accept flights from foreign countries.
The expressway from Osaka has been extended near here. In the future this town will be the base not only
for traveling but also for sightseeing along the south of the Kii Peninsula which includes World Heritage
Sites.
6. The conclusion
The general value influences and property-specific value influences in the resort area are full of
variety and are apt to be affected by business trends. So it is very important to understand the
market-specific value influences in analyzing the value influences.
It is common for us to replace the gap of each influence in each condition with the gap rate by our
intuition. But, as mentioned above, we must pay attention to the different exponent (or coefficient) of
each condition. This exponent (or coefficient) may be called the weight of each influence.
Economics say that utilities can not be counted but can be only expressed as an ordinal number.
Nevertheless we need the utilities shown as a cardinal number in order to appraise the land. Therefore we
substitute the cardinal gap rate for the ordinal value influences. But we must pay attention to the fact
that each influence has a different weight even if each gap rate is the same. It is as if the volume was
the same at a glance but the weight was different when you compared a spoonful of water with a spoonful of
milk.
In this analysis of Shirahama town the environmental condition has the most weight in the market-specific
value influences. It is not too much to say that the land value depends on the environmental condition if
there is not much gap of the land price in a city or town.
Generally speaking, the broader the width of the road is or the closer the railway station is, the more
the land value may increase. So when we substitute the numerical gap rate for the real data of these
influences, it may be correct in the condition. But we must recognize how much weight the condition has
in the total market-specific value influences. So we need to apply the multiple regression analysis.
Recently personal computers have come into wide use and various functions of the spreadsheet software
have also been improved. So we can use the statistical analysis which includes this multiple regression
analyses. And we must recognize the character of the marker-specific value influences in the area where
the subject property exists and make the most of the appraisal opportunities.
The multiple regression analysis is suitable for not only understanding market-specific value influences
but also judging whether the evaluation of many spots is well-balanced. Please try to use the multiple
regression analysis in your appraisal.
There is a proverb in English "You cannot see the city for the houses".
|
