Real Estate Investment Analysis of Toronto Neighbourhoods
Leveraging the Foursquare API, Folium & Data Science for intelligent investment decision making
Real Estate Investment involves purchasing, managing, renting and selling real or physical property, for a profit. Such an investment is characterized by a large amount of capital, and involves careful planning and decision making, for it to be a successful one.
Toronto is the largest and most populous city in Canada, known for being diverse, multicultural and home to world class-amenities, making it a great place to invest in a house!
When investing in a property, an overall market analysis will provide a good idea of the trends, but will not suffice. It is imperative to buy property in the right neighbourhood, because even if the overall market is great, a wrong location may lead to decreased property value in the future.
Let’s use the power of Data science to perform a Neighbourhood-wise Analysis of Residential Real Estate in the 140 Neighbourhoods of Toronto, to determine which Neighbourhood would potentially offer a good return on investment.
For this analysis, Demographic data from the City of Toronto’s Open Data Portal, Real Estate Data from Community Reports & Market Watch of TRREB(Toronto Real Estate Board) & Realosophy, and Location data using the Foursquare API is used.
Analyzing Demographic Features
The Population, Educational Attainment, Crime Rate and Median Income Levels of a Neighbourhood, have a great impact on the value and investment potential of the houses in it. A Neighbourhood with good population growth, high educational attainment, greater median income, low unemployment rate and low crime rate, is generally considered to be an ideal Neighbourhood. Let’s visualize these variables using Folium and Seaborn.
Population Growth
The neighbourhoods with the highest population growth are Waterfront communities-The Island and Niagara. Followed by them are Bay Street Corridor and Henry Farm.The neighbourhood with highest population decline is Weston-Pellam Park.
Crime Rate ( per 1000 people)
The neighbourhoods with the highest crime rate are Bay Street Corridor, Moss Park and Lambton Baby Point. Even the surrounding areas have a high crime rate i.e. Church-Yonge Corridor, Kensington-Chinatown and University.
Median Income
Bridle Path-Sunnybrook-York Mills and the surrounding areas have the Highest Median Income. Apart from them, the neighbourhoods of Kingsway South and Princess-Rosethorn also have high median income.
Educational Attainment
Neighbourhoods with more educational attainment are considered more valuable, as education has a great impact on the outlook of a society. Highly educated areas naturally have higher income and employment, making them favourable for investment. We can see that, the least educated neighbourhoods are Beechborough-Greenbrook, Maple Leaf, Caledonia-Fairbank, Glenfield-Jane heights and Rustic, with their highest education being High School mostly. The most educated neighbourhoods are Waterfront Communities-The Island, Willowdale East, Annex, Bay street Corridor, Church-Yonge corridor and Niagara, with their highest education being a Bachelors Degree and above.
Analyzing Real Estate Features
Real Estate Data such as House Values, Structure Type, Number of Bedrooms, Year of Construction, Rental Rates, Total Sales and Sales Volume, Sales Price to List Price Ratio and Average Days on Market are factors that influence the value of a property. Let’s visualize some of them with folium.
House Prices
As you can see, the most expensive houses are at the heart of the city. The neighbourhoods with the highest house prices are Bridle Path-Sunnybrook-York Mills, Lawrence Park South and their surrounding areas. Away from the centre, the neighbourhoods of Princess-Rosethorn and Kingsway South have expensive homes.
Median Rent
Again, the highest rents are at the centre of the city. The neighbourhood with the highest rents is Bridle Path-Sunnybrook-York Mills. Apart from that, the areas of Princess-Rosethorn, Cabbagetown-south St.James Town and Guildwood also have high rents.
Sales Price to List Price Ratio
It is a key indicator in real estate investment analysis. It is the Final Sale Price/Last List Price * 100 . A SP/LP above 100 means that the property is sold over the list price. A SP/LP of 100 means that the property is sold at the list price. A SP/LP below 100 means that the property is sold below the list price. The areas with the Highest Sales Price to List Price are Blake Jones, North Riverdale, Player Estates-Danforth and the surrounding neighbourhoods. Apart from them, even Runnymede-Bloor West Village has a high SP/LP. Bridle Path-Sunnybrook-York Mills has the lowest SP/LP.
Average Days on Market
Lower Average Days on Market, means that the property is sold very quickly, and is of high demand. Higher average Days on Market, means that the property isn’t sold quickly, and has lesser demand. The properties which are listed for the highest number of days on the market belong to Bridle Path-Sunnybrook-York Mills.
Analyzing Location Data
This is the most important factor in Real Estate Investment. By location, we mean the proximity of the property to various Amenities.
Amenities are enhancing features, which benefit a location, contribute to its enjoyment and increase its value. A good location has close proximity to transport, schools, recreation, shopping, employment and various other Amenities. A property which has accessibility to Public Transport, Roads, Hospitals, Schools, Convenience Stores….and the like, is much more valuable than a property that doesn’t.
The location data has been collected using the search and explore endpoint of the Foursquare API. I’ve used the Search Endpoint to collect data on the most important amenities, which are Schools, Groceries/Convenience Stores and Hospitals and the Explore Endpoint to collect data on amenities like Recreation, Shopping, Restaurants, Transport, Banks…. in each Neighbourhood, and then sorted & grouped them into relevant categories.
Let’s visualize the amenities obtained:
The Foursquare API returns mostly restaurants as venues, as observed above.
- The highest number of Schools are in Mount Pleasant East.
- The highest number of Groceries are in Little Portugal and Corso Italia-Davenport.
- The highest number of Hospitals are in Bay Street corridor and Church-Yonge Corridor.
- The highest number of Transport amenities are in Waterfront communities-The Island and Clairlea-Birchmount.
- Banks/Gas facilities are fairly well-distributed across neighbourhoods.
- The highest number of Parks, Beaches and Outdoors are in High Park-Swansea, The Beaches and Church-Yonge Corridor.
- The highest number of Health,Fitness and Sports amenities are in Niagara and Church-Yonge Corridor.
- The highest number of Shopping and Recreation facilities are in Niagara, Church-Yonge Corridor, Trinity- Bellwoods and Bay Street Corridor.
- The highest number of Food and Restaurants are in Church-Yonge Corridor, Mount Pleasant West and Bay Street Corridor.
K-Means Clustering
Now that we have a good idea about the key features of the neighbourhoods of Toronto, let’s apply the k-means algorithm, which is an unsupervised machine learning algorithm, to cluster the neighbourhoods, in order to reveal their investment potential. K-Means will split the neighbourhoods in such a way that, the neighbourhoods within a cluster will be very similar, and across different clusters will be dissimilar.
Let’s split the neighbourhoods into 3 clusters:
Now, we can profile our neighbourhoods based on the clusters above, and assign them appropriate labels based on the common characteristics of the neighbourhoods. After carefully observing the clusters made by the k-means algorithm, the three clusters can be described as:
Cluster 1 : Less Populated, Vintage
This cluster consists of 95 neighbourhoods, but has the lowest average total population of ~14,000, and a low population growth of ~2.9%. It comprises mostly of families and seniors, similar to Cluster 2. The educational attainment is the high, with 46% of the people having their highest education as ‘Bachelors Degree’ and 21% as ‘Masters and Above’. They have the highest average income of ~$74,000, and the good Employment rate of ~60%. The crime rate is the lowest, when compared to other clusters. The houses are mostly Apartments/Condominiums and other attached dwellings, making up 72% of the neighbourhood houses. All number of bedrooms, from 1 to 4, are common. The number of owned and rented houses is almost equal. Most of the houses are very old, with ~45% being constructed pre 1960’s and 34% pre 1980’s. It has sharply decreasing new constructions. The houses prices are very expensive, and the highest compared to the other clusters. The houses are mostly sold over the list price. The rents are the highest, compared to the other clusters. They have moderate amenities compared to the other clusters.
Cluster 2 : Mid-Populated, Archetypal
This cluster consists of 40 neighbourhoods which have an average total population of ~30,000, and very less population growth. It comprises mostly of families with children and seniors, which account for 33% of the total population. The educational attainment is less, with 41% of the people having their highest education as ‘High School’ and 42% as ‘Bachelors Degree’. They have an average income of $64,940, lower than the other clusters, and a high unemployment rate. The crime rate is moderate, when compared to other clusters. The houses are mostly attached dwellings, making up ~69% of the neighbourhood houses, but also has a good number of detached houses i.e. ~31%. Houses with 2–3 bedrooms are more common, accounting for 58% of the total houses. There are also more owner occupied homes rather than rented ones. Most of the houses are old, with 32% constructed in the pre-1960 and 43% in the 1961–80 period. It has few new constructions, and the number of new constructions has declined sharply 1990 onwards. The houses prices are mid-ranged to expensive, depending on the structure & bedrooms. The houses are mostly sold over the list price, an get sold very quickly. The rents are mostly mid ranged, but can be high depending of the house. They have fairly fewer amenities compared to the other clusters.
Cluster 3: Densely populated, Bustling Urban
This cluster consists of only 5 neighbourhoods, but has the highest average total population of ~40,000, and extremely high population growth of ~30%. It has the highest working-aged population i.e. 68%, and most of them seem to be single or couples without children. The educational attainment is the highest, with ~56% of the people having their highest education as ‘Bachelors Degree’ and ~27% as ‘Masters and Above’. They have an average income of ~$65,000, and the highest Employment rate of ~68%. The crime rate is the highest, when compared to other clusters. The houses are mostly Apartments/Condominiums, making up ~84% of the neighbourhood houses. Houses with 1–2 bedrooms are the most common, accounting for ~90% of the total houses. There are also more rent occupied homes rather than owned ones. Most of the houses are new, with ~63% being constructed post 2000. It has rapidly increasing new constructions. The houses prices are low to mid-ranged, and much more affordable compared to the other clusters. The houses are mostly sold a little less than or equal to the list price. The rents are mostly mid ranged, and don’t vary much. They have the highest amenities compared to the other clusters.
Investment Strategies
The Neighbourhood Analysis of the 140 neighbourhoods of Toronto have revealed the investment potential for these neighbourhoods, based on their characteristics. Let’s look at it from the point of view of both individual investors and Real Estate companies.
a. For an investor looking to invest in an area with high development & growth, close proximity to commercial/business outlets, new modern constructions, and at the same time mid-ranged capital investment, the neighbourhoods of Cluster 3 are the most ideal.
- An Individual investor can opt for buying a Condominium, for rental property investment, as they are the most popular house structure here, and fetch good rents. However, if an individual wants to opt for a buy and hold strategy, it may or may not profitable, as houses in these neighbourhoods usually sell a little below or equal to the list price. Moreover, houses don’t sell as quickly as the other clusters.
- A Real Estate Company can find it profitable to invest in an Apartment, as most of the inhabitants in these neighbourhoods live in rented homes. Another key point is that smaller homes with 1–2 bedrooms are more in demand than bigger homes in such neighbourhoods.
A downside to investing in these areas is the extremely high crime rate, probably because of the presence of Commercial establishments, markets and businesses, as some crimes take place in bustling areas i.e.Robbery. This may explain why there are less families and more singles’ living in these neighbourhoods.
b. For an investor looking to invest in an Archetypal, family-friendly neighbourhood, with great value, the neighbourhoods of Cluster 2 are favourable.
- An individual investor looking for a home to live in with his family, or to earn rental income, can invest in the properties of these neighbourhoods. These neighbourhoods have a good capital appreciation, and are very profitable in the long run. They have a low crime rate, and have a pleasant ambience. Most of the homes are owner occupied, and increase in value overtime.
- A Real Estate Company can invest in both Apartments or Detached houses. There is a high demand for homes in such neighbourhoods as they get sold very quickly, and mostly above the list price. Big homes with more than 2 bedrooms are high in demand, due to family size.
A downside of these neighbourhoods is that the educational attainment is less, and unemployment is high. This may have a negative impact on timely payment of rents, in case of rental property investment.
c. For an investor looking to invest in a swanky,peaceful & posh neighbourhood, Cluster 1 has many neighbourhoods of this kind!
- An individual investor wanting to live in a well-educated, high-class neighbourhood, or simply wanting high capital appreciation, can invest in suitable neighbourhoods in this cluster. Rental property investment is also a great option, with high return on investment. It has the safest neighbourhoods, with the lowest crime rate. There is an equal distribution of owned and rented houses, so both buy and hold strategy and rental property investment are great, depending on the requirements of the investor.
- A Real Estate Company can invest in Apartments, of any size, big or small, as rents are high for both, or in big houses like mansions in certain areas, which are sold for high prices. There is high demand for residential real estate and homes are sold mostly above the list price.
A downside of these neighbourhoods is that there is the presence of working class neighbourhoods, as the k-means algorithm has classified some neighbourhoods on the basis of Total population. So, before investing, one must carefully observe the neighbourhoods individually.
Conclusion
The goal of this analysis was to identify the neighbourhoods of Toronto with high investment potential and return on investment.
With the help of Demographic, Location and Real Estate data, insightful visualizations, and then ultimately clustering the neighbourhoods, we could identify areas for potential investment, depending on the preferences of the investor.
Throughout the analysis, a neighbourhood which constantly stood out was Waterfront Communities-The Island. In my opinion, it has the highest investment potential, no matter who the investor is. It has the highest development, population growth, education, proximity to amenities…..and so on. Moreover, the prices are very reasonable.
Some other neighbourhoods with great investment potential are Niagara, Willowdale East and Runnymede-Bloor West Village. I think that these neighbourhoods would exceed the investors expectations in the short as well as long run.
However, the ultimate decision, to invest or not to invest, depends on the investor, as everyone has their own needs and preferances.
Hopefully, this analysis will be helpful to anyone who wants to answer the question, “Is this area worth a closer look?”
Note: You can find the complete code in my notebook on Github.
