How we measure property value
The Property Value score is a 0–100 score built from within-city percentile ranks of municipal assessment rolls, permit activity and the legacy Census shelter-cost ratio. It describes how a neighbourhood's assessed values, growth and investment activity rank within its city. It is not built from MLS sold prices, and an assessed value is not a sale price.
Last updated
What we measure
Assessed values come from official municipal assessment rolls: the City of Calgary's property assessments, BC Assessment values published through Vancouver's open-data portal, and the City of Edmonton's assessments. For each neighbourhood we compute the median, quartiles, mean and spread of current assessed values, plus year-over-year and five-year growth from the historical roll years.
Toronto is different, and we state it rather than hide it. Ontario's Municipal Property Assessment Corporation (MPAC) does not publish per-parcel values as open data. Toronto values on PickYourPlace are estimates: the Statistics Canada Census 2021 median dwelling value of each parcel's dissemination area, projected forward with a TRREB House Price Index multiplier per property type (detached, semi-detached, townhouse, condo apartment and others) so the estimate lands in the same roll-year vintage as the other cities. In the current run 98.3% of Toronto's roughly 984,000 parcels carry a projected value; the rest sit in areas where Statistics Canada suppresses the median.
Renovation activity is measured from building permits and new-build activity from development permits — the two are kept separate because they signal different things. Counts and total project value are computed per neighbourhood per year.
The incumbent four-city scoring policy retains a Census shelter-cost ratio, historically labelled rental yield: annualized median tenant shelter costs divided by owner-estimated dwelling value. Shelter costs include applicable utilities, and the numerator and denominator describe different households. This ratio does not establish landlord rental income or investment yield.
Montréal and Québec City publish independent assessment ranks. Their Value composite is published only when observed components reach 40% of the fixed model weight: assessment supplies 25%, and new-build and renovation permits add to it where a city's permit evidence is observed. Two disclosed estimates also count where their evidence exists: growth is the previous triennial roll value annualised over its three-year cycle (not year-over-year change), and rental yield is CMHC zone average rent times 12 over the median assessed value (gross, and blank where no CMHC zone covers the area). Historical Census tenant shelter costs remain separate context and are never used for rental yield.
Data sources and update cadence
- City of Calgary property assessments (data.calgary.ca)
Per-parcel assessed values and roll-year history for Calgary
CadenceAnnual assessment roll
- BC Assessment values via Vancouver open data (opendata.vancouver.ca)
Per-parcel assessed values and roll-year history for Vancouver
CadenceAnnual assessment roll
- City of Edmonton property assessments (data.edmonton.ca)
Per-parcel assessed values and roll-year history for Edmonton
CadenceAnnual assessment roll
- Statistics Canada Census 2021 + TRREB House Price Index
Toronto value estimates: dissemination-area median dwelling values projected by per-property-type HPI factors
CadenceCensus vintage 2021; HPI factors updated when the model is re-tuned
- Municipal building and development permits (Calgary, Vancouver, Edmonton, Toronto open-data portals)
Renovation activity (building permits) and new-build activity (development permits)
CadenceTypically weekly at the source; ingested on the pipeline cadence
- CMHC Housing Market Information Portal
Rental-market time series used in rental context
CadenceCMHC survey cycle, typically annual
How the score is computed
Five sub-metrics are computed per dissemination area: median assessed value, average year-over-year growth, development-permit activity, renovation-permit activity, and the legacy Census shelter-cost ratio. Each is ranked against every other dissemination area in the same city with PERCENT_RANK, giving 0–100 percentiles, which are combined with fixed weights into the Value score. Dissemination areas roll up to the neighbourhood by population-weighted averaging.
A score of 50 means the neighbourhood is roughly typical for its city on this blend. Because the blend is not itself a rank, a 50 is not exactly the city median. Growth carries the largest weight at 30%, so the score favours neighbourhoods where assessed values are rising relative to the rest of the city, not simply neighbourhoods that are expensive.
Assessment methodologies differ between cities — Calgary assesses with sales-comparison and income approaches, Vancouver uses BC Assessment, and Toronto values are census-based estimates — so the score is only meaningful when comparing neighbourhoods within one city.
Fixed shares of the Property Value model. Québec retains these weights and the minimum 40% available-weight requirement. A Québec Value composite is published only when observed components reach it; assessment alone supplies 25%.
What this score does not capture
Stated plainly, because a score is only worth trusting if its limits are.
- No MLS sold prices. Sale and transaction data are not in the open-data feeds this lens is built on. Assessed values track relative patterns well but they lag the market and are not appraisals.
- Toronto values are modelled estimates, not assessments. Every parcel in the same dissemination area and housing-type bucket shares one estimate, and year-over-year growth is not computed for Toronto because there is only one census baseline.
- Scores are not comparable across cities. Assessment methodologies differ, and Toronto's figures are census-derived, so only comparisons within one city are meaningful.
- Assessed values are set on a cycle. An assessment reflects the roll-year valuation date, not the market on the day you look.
- The legacy Census shelter-cost ratio is not measured rental yield. Tenant shelter costs include applicable utilities, while owner-estimated dwelling values refer to a different population. The ratio cannot establish rental income, return or a current market rent. Québec rental ranks are withheld.
- The incumbent four-city policy retains legacy missing-data fallbacks. Québec missing or unsupported scores remain unavailable; assessment facts require their own observed coverage and do not inherit a neutral 50.
Definitions
- Assessed value
- The value a municipal assessment authority assigns to a property for taxation, tied to a roll-year valuation date. Not a sale price and not an appraisal.
- Roll year
- The assessment cycle year a value belongs to. All four cities' values are aligned to the same roll-year vintage, Toronto via HPI projection.
- Year-over-year growth
- Percent change in a property's assessed value between the current and previous roll year, averaged across a neighbourhood.
- Five-year growth
- Current assessed value divided by the value five roll years earlier, minus one. Null where fewer than five years of history exist.
- Building permit
- A municipal permit for work on an existing structure — renovations and alterations. Used as the renovation-activity signal.
- Development permit
- A municipal permit for new construction or a change of use. Used as the new-build-activity signal, kept strictly separate from building permits.
- Legacy Census shelter-cost ratio
- Median monthly tenant shelter costs × 12 ÷ median owner-estimated dwelling value × 100, from the 2021 Census. Applicable utilities and different household populations prevent interpreting this as rental income or yield. Withheld for Québec.
- Percentile score
- A sub-metric's rank among the same city's dissemination areas, 0–100, where 50 is the city median. The Value score blends these ranks and is not itself a rank. Percentiles do not transfer between cities.
Frequently asked questions
Does PickYourPlace use MLS sold prices?
No. The Property Value lens is built from municipal assessment rolls, permits and census data — all public sources. Sold prices are not in those feeds, and tools that show them license the data separately. An assessed value tracks relative patterns well but is not a sale price.
Why are Toronto property values estimates?
Ontario's MPAC sells per-parcel assessments under licence rather than publishing them as open data. PickYourPlace estimates Toronto values from Statistics Canada Census 2021 dissemination-area medians, projected forward with per-property-type TRREB House Price Index factors. 98.3% of Toronto parcels carry an estimate; the method and its error against TRREB district medians are validated on every run.
What does a value score of 70 mean?
A strong result on a weighted blend of the neighbourhood's within-city ranks for median assessed value, growth, permit activity and the legacy Census shelter-cost ratio. Québec does not publish this composite until enough supported inputs qualify. It does not mean the neighbourhood ranks above 70% of its city. Because growth carries the largest weight (30%), a 70 is closer to "appreciating and active" than to "expensive".
How current are the property values?
Assessment rolls are annual, so values reflect the most recent roll year, not today's market. Permit data refreshes roughly weekly at the source. Every report shows the data's last-refreshed timestamp.
Why do building permits and development permits count separately?
They signal different things: building permits capture renovation of existing homes, development permits capture new construction. A neighbourhood scoring high on one and low on the other is telling you something specific, so the score weights them separately at 15% each.