IS428 2016-17 Term1 Assign1 Zheng Xiye

Problem & Motivation

With limited amount of land available for construction and high demand for private properties, Singapore has always been a 'low-hanging fruit' for real-estate players to gain substantial benefits from constructing and selling excessive number of private properties.

By referring to Morgan Stanley Research on Singapore Residential Supply, private properties supply reached its peak around millennium, decreased to historical minimum around 2008 Global Economic Crisis and re-bounced to around 40K units in 2014. The rapidly inflating real-estate market flashed 'red-light' upon Singapore government, urging policy makers to structure regulating measures in preventing the 'real-estate bubble' from bursting.

In order to facilitate policy makers' decision making process, this study aims to identify both supply and pricing patterns of private properties across the entire Singapore. By doing so, policy makers will be having a 'touch and feel' on the landscape of real-estate market. Effectively identifying real-dangers and potential risks, policy makers will be empowered to structure preventive mitigation actions accordingly, which ensures the well-being of Singapore real-estate market.

Tools

• Microsoft Excel: mainly for data pre-processing: data cleaning, sorting and transforming.
• Tableau 10.0: creating charts that effectively identify patterns hidden in huge data-sets and providing insights that may be proven helpful in facilitating policy makers' decision making process.

Approaches

Share of Private Property Supply 2015

Private Property Sale Overview 2015

After merging data downloaded from REALIS namely: Residential Supply 2015 q1-q4, I came to realize that in order to trace the trend of private property sale across 2015, an additional column, Quarter is required.

Chart plotted above illustrates change in percentage of total no.of completed and sold private properties across 4 quarters of 2015. Both of which reflect reducing heat in private property supply as indicated by the down-going percentage trend. Assuming that the reducing heat is not attributed to a seasonal effect and positive correlation between total completed and sold, we are able to conclude that private property supply will continue to shrink in 2016.

Private Property Sale Type Overview

Merged data on Residential Supply includes 6 types of Private Properties in total namely:

1. Condominiums
2. Executive Condominiums
3. Detached
4. Semi-Detached
5. Apartments
6. Terrace

However, I noticed some of the categories (i.e. Semi-Detached & Executive Condominiums) include very few number of inputs. As a result, I decided to merge categories: Condominiums & Executive Condominiums, Detached & Semi-Detached by making use of Tableau's 'Create Calculated Field' function.

By referring to the chart plotted, it is obvious that Condominiums and Apartments account for most of the Private Property Sales across 4 quarters of 2015. On top of which, the general decreasing trend in Private Property Sales should be attributed to the decrease in the number of Condominium Sales, followed by Apartment Sales.

Private Property Geographic Distribution

After plotting base map with given Longitude and Latitude Measures, I have placed Dimension Postal Code as Detail so as to plot all Private Property locations on the base map. After which, I have placed Planning Region Dimension under Color such that all Private Property Plots on the base map will be colored accordingly. As shown in the chart plotted, Private Properties are clustered around their geographic locations. Yet, cluster density around Central and North East Region is substantially higher than that of other regions, suggesting these 'best-sell' areas contributing to the steadily increasing Private Property Sale.

Distribution of Private Property Price 2015

Planning Region Unit Price (psm) Overview

After browsing through merged Transaction datasheet, I found that all both Unit Price ($psf) & Unit Price ($psm) are whole numbers. If I were to plot all of which to any chart without pre-processing it, I will be dealing with tens of thousands of Unit Prices. As such, I have again made use of Tableau's 'Create Calculated Field' function based on Unit Price ($psm) to merge all Unit Prices into 4 categories namely:

1. 0-10k
2. 10k-20k
3. 20k-30k
4. 30k-40k

After which, I managed to plot a horizontal bar chart with both Dimensions: Price psm & Planning Region as Rows and SUM(Number of Records) as Columns.

Looking into each of the Price psm categories, I noticed that Central Region is leading in terms of price count percentage in most of the high Price (psm) categories followed by North-East Region. As such, we are able to imply that the high sale quantity in both Central & North-East Region can be attributed to the high selling price in both regions.

Tenure Price Relationship Overview

Column Tenure of merged Transaction datasheet is a case by case situation whereby thousands of tenure limits are included in the column. As such, by adopting the same approach as the previous analysis, I have created a new calculated field: "Tenure Free or Not". As the name suggests, it includes two categories:

1. Free
2. Tenure Limit

After which, I placed newly created 'Tenure Free or Not' field in Columns, SUM(Number of Records) in Rows and 'Price psm' in Color.

Plotted chart shows that generally Private Properties free from tenure need to pay higher price as compared to those with tenure limit. This is reflected by 'Free' category's low percentage of 0-10k psm and high percentage of 10k-20k and 20-30k psm.