COVID’s Impact on the Housing Market


The COVID19 outbreak is rapidly changing the housing market, with real estate companies quickly adopting digital strategies and alternative methods of doing business. In Ontario, for example, agents have been forced to change. The Real Estate Council of Ontario (RECO) has reported that, while real estate services are still considered essential, the government prohibited open houses as of April 4, 2020. Firms now have to sell digitally if they hope to attract buyers. Moreover, the agents with the most foresight in the business may also not be migrating online as a temporary measure – many expect the industry to change permanently after things settle down.

Reporting through major news networks such as the CBC, Toronto Star, Morningstar and others, real estate analysts seem to agree that the outbreak will impact housing prices as well. Toronto Real Estate Board (TREB) data showed that for the first 17 days of April 2020, MLS-based sales dropped 69 per cent compared to the same period last year. “Since COVID-19 forces a change in how people buy and sell, and the volume at which they do so, there is likely to be an expected change in housing market price behaviour,” says Amir Madadi, CEO of Gnowise.

The Gnowise intelligent home valuation system


Madadi and his team believe they can help. Gnowise offers a complete, neighbourhood-focused home price forecasting and valuation tool. They are also offering adjustments buyers can make based on their own preferences and experiences with the price. Complete with risk simulations under various conditions, Gnowise has grown its user base substantially since launching in 2018. The firm isn’t planning to cut out the experts either – which many AI/ML systems aim to do. Madadi says that “human interaction is the largest driver of trust in the markets – whether real estate or financial. And without trust, the market collapses. People often forget that when building complex models and API access for developers.” For that reason, Gnowise also offers a “human-in-the-loop” approach, with certified appraisers verifying AI/ML valuations from their platform.

That approach is proving particularly valuable right now, where traditional algorithms don’t take “unforeseen” risks into account the same way that experts do, he says. While Gnowise uses extensive amounts of open data to model prices off of every variable available, including distance to schools, macroeconomics, capital markets, jobs and so on, it can also adjust when needed.

Gnowise’s CTO, Faraz Arbabi, a PhD graduate of Engineering at the University of Toronto, specializing in numerical and computational modelling, says that this allows the firm to rapidly gather accurate, real-time data while researching COVID19’s impact on the housing market. “We’re in early stages of researching the exact price changes that occur due to the new shifts in the economy. We’re integrating more data every day, through the Centre of Excellence in Next Generation Networks (CENGN) Cloud Computing system based out of Communitech’s Data Hub. Soon, we expect to be able to forecast such impacts through our automated API.” Learn more about Gnowise here, and join them in the next CENGN cohort.

Using the Gnowise API


To use the Gnowise API, users must first obtain an API key provided by the firm, which can be requested along with pricing terms here. Note, your API call should pass information associated with a property. Gnowise then returns results associated with that property in the API response in less than one second. All Gnowise API calls are HTTPS POST requests. The response is JSON data contained in the body. An Intelligent Valuation Models (IVM) API sample can be found below:

Request:


POST /api/session-query HTTP/1.0

Host: https://pwlq74oz0j.execute-api.ca-central-1.amazonaws.com/gnowise-ivm-api Content-Type: application/json

x-api-key: [client key]

Content-Length: 116

Body: {“address”: “99 Cordella Ave”, “unit”: na, “zip”: “M6N2J8”, “bedrooms”: “3”, “washrooms”: “3”, “den”: 0, “age”: “15”, “rooms_area”: “2200”, “lot_area”: “3100”, “community_code”: “01.W03.0360”, “parking_spcs”: “1”, “pool”: na, “type1”: “Semi-Detached”, “style”: “2-Storey”, “basement1”: “Finished”, “basement2”: “Sep Entrance”, “kitchen”: 1, "condition": "3" "chartID": "e239d259-3837-4c6d-898e-1a59994888c3",}

Response:


{ "gnowise_value:": “8733400”, "value_low": "8177600", "value_high": "8733400, "attributes": [“location”, “bedrooms”, “den”, washrooms”, “age”, “rooms_area”, “type1”, “basement1”],

"price_forecast":[{"d":"2020-04-30T00:00:00.000Z", "v":8733400.0 }, ... ],

"best_time_to_buy": "2020-06-30 00:00:00", "best_time_to_sell": "2020-04-30 00:00:00", "risk_of_decline": 55.0,

"chartID": "e239d259-3837-4c6d-898e-1a59994888c3"}

 

Required Input Parameters


The table below lists required API Input parameters (for IVM):

Parameter Name Type Descriptions
org_id CHAR 32 A unique client ID generated by Gnowise for your organization
bedrooms INT 16 Number of bedrooms
washrooms INT 16 Number of washrooms
den INT 16 den
age INT 16 Age of property
rooms_area FLOAT 16 Sqft of the property
community_code CHAR 32 Code of the neighbourhood
parking_spcs INT 16 Number of parking spaces
type1 CHAR 32 Type of property
style CHAR 32 Style of property
Basement1 CHAR 32 Type of basement 1
Basement2 CHAR 32 Type of basement 2
kitchen CHAR 32 Number of kitchens
chartID CHAR 32 Chart ID


 

Recommended Input Parameters


The table below lists recommended API input parameters, which will enable Gnowise to analyse inputs in an optimal manner and will facilitate search functionality in the Gnowise control portal. Please note that it is a requirement to pass at least one account identifier as part of the API call.

 

Parameter Name Type Descriptions
address CHAR 128 Property address
postal_code CHAR 16 Number of bedrooms
unit CHAR 16 Number of bathrooms
pool CHAR 32 Type of pool if exist
condition INT 16 Condition of property


 

API Output Parameters


The Gnowise API result can return up to 10 different attributes.

 

Parameter Name Type Descriptions
gnowise_value INT 16 Predicted price
value_low INT 16 Minimum predicted price value
value_high INT 16 Maximum predicted price range
attributes ARRAY Attributes effective on price
price_forecast ARRAY Price forecast
best_time_to_buy DATE Best date to buy this property
best_time_to_sell DATE Best date to sell this property
risk_of_decline FLOAT 16 Risk of price decline
chartID CHAR 32 Chart ID