ISSS608 2016-17 T1 Assign1 Shishir Nehete

Abstract

The Project focuses on the latest trends in resale of the public housing supply (HDB). To start with analysing the trends in sales of the HDB, these are the key points to note:

• Flat type
• Town
• Storey range
• Resale price
• Demand and supply as per applications and registrations

The project gives insights about these factors and the meaningful conclusions that can be made using these insights.

Problem and Motivation

The initial step for buying or selling the property is the analysis of the trends of resale prices across multi variables, such as Flat type, Town, Storey range, Age of the flat. The potential buyers or sellers can study these trends to be aware of the present market scenario. Thus appropriate visualization of this data will enable them to understand these stats in a quick and efficient manner, and help them make well-informed decisions. This project will majorly focus on below points to gain insights from the data.

• What are the shares of the resale public housing supply in 2015 (at least three observation)
• What are the distribution of the resale public housing prices in 2015 (at least three observation)
• With reference to the findings, compare the patterns of the first-half of 2016 with the patterns of 2015

Approaches

Data Acquiration

As knowing the dataset for compilation is the first step of visualization, the data needs to be obtained from a reliable source. The data used for this project is found at Government database (https://data.gov.sg). Following datasets are used for further analysis.

• Resale Flat Prices (Based on Registration Date) from March 2012 Onwards. (https://data.gov.sg/dataset/resale-flat-prices)
• Housing and development board resale price index quarterly. (https://data.gov.sg/dataset/hdb-resale-price-index)

Data Preaparation

The data acquired is clean and no missing values are noticed. However, the date field is in the string format which needs to be changed to date format. Hence we use the following formula for creating a new Calculated Field for Month Dimension.

• DATE(""+left([Month],4)+"-"+RIGHT([Month],2)+"-01")

Moreover, the resale price is given in units, so to analyse better, a new calculated field is created which specifies the Resale price per square meter. The formula for this calculated field is as below.

• [Resale Price]/[Floor Area Sqm]

Data Exploration

Data visualization techniques are used in this project for designing graphs and to depict a clear picture of the public housing supply over the year 2015. The shares of the resale public housing supply and the distribution of the resale public housing prices can be depicted across different variables. Also, the data is present for the first-half of 2016, so for comparing the patterns of first-half of 2016 with the pattern of 2015 we will consider data of first half of 2015 only.

Tools Utilised

1. Tableau – To explore and visualize data for finding insights from the visualizations
2. Power BI – To create visualizations and interactive graphs for counts of flats registered across the town in particular month of 2015.
3. Microsoft Powerpoint – To create Infographics.

Results

Infographics

Demand and Supply of the flats over the year 2015

Demand and Supply of flats over 2015

The left bar graph shows the number of resale applications made over the year 2015 and the graph at right shows the number of successful registrations of the resale flats. This graph shows that most of the applications made were registered with the success rate of 92.11% overall (Total number of Applications: 19306, Total number of Registrations: 17784). Though, we can also take an important insight that most applications were for the flat type 4-room, followed by 3-room and then 5-room respectively.

Distribution of Flat Types by Town

The above graph depicts the registered flat types across the towns in Singapore. Major contributor to the count of the flats is Jurong-West followed by Tampines. Also, the 4-room flats are the most registered flat across the whole city.
The following graphs provide insights of the number of registrations of resale houses which were registered per month. The registrations can also be visualized on an area basis. The month of October had the maximum number of registrations followed by June. On an area level, Jurong-West had the maximum registrations in the year 2015. We used the count of flat types in the graph to visualize the number of registrations. By selecting the filter on the right side, we can achieve a count of registration on a monthly basis for each area. This can be viewed in the screen video in the following link. (https://youtu.be/BmXG4JnxFNQ)

Distribution of Resale prices over flat types

The above graph gives us a clear picture of the median prices of the flats as per the flat types. The 4-room flats are highly priced with median value around 2,705,103 per sq meter. We also notice that for the 3-room flat, there are outliers that affect the mean. But in whole the price distribution across the flat types can be obtained using the above box plot distribution.

Resale Price across the town

The above graph shows the Pareto distribution for the major contributing towns regarding the resale price. From the graph, we get an insight that the 5 towns, namely Jurong-West, Tampines, Bedok, Sengkang and Woodlands contribute to the resale prices and thus are the major towns for the resale market.

Distribution of Resale prices across different Storey Ranges

As seen in the heat map, the 4-room flats at Storey Range 04-06 are highly in demand and make the costliest flats. High storey flats are less in demand and thus cost less compared to lower storey flats.

Distribution of Resale prices across the Streets

The above box plot gives an insight about the median values of the resale prices of different flat types across the streets. There are some outliers at that affect the mean of the resale prices across the flat types. Some of the examples of these outliers are shown in the graph, for example the flat at Yishun Ring Rd has a Resale Price of 2,405,102 per sq meters which is high compared to the median value of the 4-room flat type which is 219,253.

Comparison of Patterns of first half of 2015 and those of 2016

The above graph uses the data of first half of 2015 and 2016. The Pareto chart of count of flat type shows the demand for the resale flats in the first half of 2015 and 2016. As compared, the trend is quite similar over the flat types in both years. However, there is a slight increase in demand for the 4-room flat in the year 2016 as compared to that of 2015. An important insight is that the demand for the 3-room flat has reduced in the year 2016, and the demand for 5-room has increased. This can be seen in the cumulative % values. The next Pareto chart shows the distribution of Resale prices across different flat types in the first half of year 2015 and 2016. Similar to the inference from above Pareto chart, as the demand for 5-room has increased gradually, and 3-room flat has reduced, the cumulative % values show an increase in resale prices for 5-room flats in 2016 as compared to those in 2015.

Resale Price Index Pattern for previous 10 years

The above graph shows the resale price index pattern for the last 10 years. The price index went up in the year 2013 to the maximum value but since then it has followed a decreasing pattern. As clear from the graph, the price index has more or less stabilized in the last 4 quarters with minimal fluctuations.