Tiểu luận what affects people s choices of purchasing apartment

We have found that due to the result of model testing, our seniors have removed “direction” variable- which Gretl has listed as insignificant.. We built multiple regression model with so

FOREIGN TRADE UNIVERSITY

ECONOMETRICS ASSIGNMENT

What affects people’s choices of purchasing apartment?

Table of Contents

Introduction 3

Methodolody 4

Econometrics Model I Model 4

II Variables meaning 1 Specific explanations of variables .4

2 Data descriptions 5

III Results 1 Running Model 7

2 Testing Model 7

a Testing Normality 7

b Testing Multicollinearity 8

c Testing Heteroskedasticity 9

d Testing Autocorrelation 11

e Testing Specification Bias 11

IV Conclusions 1 Final Regression Model 12

2 Explanation 12

V Sugesstions 13

Appendix 14

INTRODUCTION

Housing nowadays seems to be a never-ending problem of society due to the rapid growth of population while lands are limited It leads to an obvious result that the

Page | 2

prices of real estate, especially in big cities like Ho Chi Minh City or Hanoi, go up and down unstably The fluctuation of real estate price confuses customers, who are largely newly married couples seeking for their own nests

Since the scarcity of usable areas are still questioning land planners, a lot of

sensitive investors start investing in apartments, which seems to be the best

solution to housing problems at the moment Apartment brings a lot of advantages

to both contractors and customers: modest area required, cheaper cost, more people contracting at the same time, better security……

To figure out what affects the price of an apartment, we have done a research on which factors the price is dependent to bring a clearer view and support to

everyone who is planning to buy an apartment The question given is: “Which

elements affecting Hanoi’s apartment price?”

A METHODOLOGY

So far we have collected date from some reliable internet sources like

www.nhadatvideo.com, www.nhadat24h.net as well as consulted some

assignments of senior students to clarify what we should implement and develop and continue researching We have found that due to the result of model testing, our seniors have removed “direction” variable- which Gretl has listed as

insignificant Though in reality, most Hanoi people when deciding to buy a house

or an apartment usually care whether its direction is suitable to the owner’s lunar age

We collected about 120 reliable observations and then chose the most sufficient once again Our group ended up with 86 observations We built multiple

regression model with some assumed variables which we suppose may have direct effect on Hanoi apartment’s price per square, constructing process, floor or

square… Our group has tested 5 assumptions of the model that are normality, multicollinearity, heteroscedascity, autocorrelation and specification bias

B ECONOMICTRICS MODEL

Prps = ß1 + ß2 *area + ß3*pro + ß4*flr + ß5*sqr + ß6*dir + ui

1 Specific explanation of variables

- Prps: price per square (million VND/ m2)

- Area: whether that apartment is close to city center or convenient or not ( if

yes = 1, if no =1)

- Pro : constructing process – whether the building is completed or not ( if

yes = 1, if no = 0)

- Flr : floor – where the apartment is located

- Sqr : square (in m2)

- Dir : direction ( if the direction is south, south east or south west => =1; if

the direction is north, north east , north west, west =>0)1

Expectation in signs

- ß1 : it depends

- ß2 : positive ( apartments near city center tend to be more expensive than

those in suburb )

- ß3 : positive ( people prefer building that has been finished so that they can

directly judge if the apartment they are going to buy is beautiful or not)

- ß4 : it depends ( some people like low floor so that it would be more

convenient for them to go around, some like higher floor in order to avoid noise and city pollution)

- ß5 : negative ( the larger the apartment is, the less expensive per square)

- ß6 : positive ( if the apartment faces good direction, its price will be raised

higher)

2 Data description

1 This direction is only based on the weather, not referring to owner’s lunar age

Page | 4

a Price per square

Summary statistics, using the observations 1950/01/02 - 1951/08/20

for the variable 'prps' (86 valid observations)

Summary Statistics, using the observations 1950/01/02 - 1951/08/20

for the variable prps (86 valid observations)

b Area

Summary Statistics, using the observations 1950/01/02 - 1951/08/20

for the variable area (86 valid observations)

c Process

- Completed: 61/86

- Uncompleted: 25/86

d Floor

Summary Statistics, using the observations 1950/01/02 - 1951/08/20

for the variable floor (86 valid observations)

e Square

Summary Statistics, using the observations 1950/01/02 - 1951/08/20

for the variable sqr (86 valid observations)

f Direction

- Good direction: 48/86

- Bad direction: 38/86

III RESULTS:

1 Running Model:

Using Gretl, we have the result as above As it shown from the result, we have some initial conclusions:

- R- squared is 0.600373 meaning that approximately 60% of the variables

can be explained by the model

- There are five independent variables which can explain the dependent

variable

- All of these five independent variables are statistically significant.

The initial Regression Model can be defined as:

Prps = 31.2517 + 7.67236*area + 5.04203*Pro – 0.285302*Flr – 0.107574*Sqr

+ 3.80662*Dir + ui

2 Testing Model:

a Testing normality:

From the model, we test normality by using command Tests Normality of residuals in Gretl The result is:

The p-value from the result is 0.0718 > 0.05 The residuals are normally distributed.

Page | 8

b Testing Multicollinearity:

There are two ways to test Multicollinearity by using Gretl

First of which is using the VIF ( variance inflating factor) :

From the result, we can see that VIF < 10 Model is free from Multicollinearity

Or we can use the Correlation Matrix:

The correlation between the variables is smaller than 0.8 Multicollinearity does

not exist in the model

c Testing Heteroskedasticity:

Using White test in Gretl, we can see if the Model is suffered from Heteroskedasticity

or not The result is:

As can be inferred from the result, the p-value is 0.203444 > 0.05 The Model does not suffer from heteroskedasticity.

Page | 10

d Testing autocorrelation:

The p- value in this result is 0.951 > 0.05 There is no autocorrelation in this

model

e Testing Specification bias:

Using Ramsey’s RESET test, we will have the result as below:

The p-value is 0.143 which is greater than 0.05 Model is adequately specificated

IV CONCLUSION:

1 Final Regression Model:

Prps = 31.2517 + 7.67236*area + 5.04203*Pro – 0.285302*Flr – 0.107574*Sqr +

3.80662*Dir

2 Explanation:

- R- squared is 0.600373 meaning that approximately 60% of the variables can

be explained by the model

- Remarks on variables:

Variable name Conjecture of

Trend of impact

True Trend of impact Explanation

taken into consideration, the dependent variable reaches the estimated

value of 31.2517.

city center will be 7.67236

million VND more expensive than that in the suburb

Page | 12

Process (Pro) Positive Positive People have a tendency to

prefer completed apartment

to unfinished ones A m2 of

a finished one will be

5.04203 million VND more

expensive than that of unfinished one

Floor (Flr) It depends Negative Lower floor seems to beat

higher floor in terms of safety in people’s thinking For one extra floor will cost

you 0.285302 million VND

less

Square (Sqr) Negative Negative One extra m2 will safe you

0.107574 million VND.

Direction (Dir) Positive Positive In reality, for Asian people,

direction will always sit on top of other problems when

it comes to purchasing estate And a m2 of an apartment facing good

direction will take 3.80662

millionVND more out of your wallet in comparison with that which faces bad direction

V SUGGESTIONS:

 To buy a good apartment, customer should have a close enough check on 3

main elements : contractor's reputation, products' quality and contract in order for you to make a right purchase

 Firstly, contractor's reputation plays a very important role in your decision

making It is better if you have a basic understanding about his/her ability or financial statement to avoid being tricked or waiting continuously without

knowing when you can get your apartment

 Secondly, product's quality is also very essential Since real estate is a kind of special products, it involves with many external factors: legal base, certificate, neighboring areas, surrounding environment The more you have knowledge

of your about-to-buy apartment, the more advantages you have when

negotiating about pricing

 Last but not least, contract matters are always one of the most concerns to customers You should carefully check all the terms before signing Process delayed is the most popular risk when you buy an apartment which has not been finished, so in this situation, fine is the best way to urge the contractor to

complete the contract on time Also, you should check the bias between the apartment's are in reality with that in contract to avoid future complaints

APPENDIX

Customers and the taste of apartments, together with the price

Page | 14

29 0 1 14 69 0

43.5 1 0 6 40 1

Page | 16