What a percentile means for net worth
A percentile places one family's net worth on a scale relative to every other family in a comparison group: standing at the 65th percentile means sixty-five out of a hundred comparable families hold less net worth, and thirty-five hold more. As with income, a net worth percentile is only meaningful relative to the specific comparison group it was computed against.
On this app, that comparison group is a province and an age band considered together, since net worth in the underlying survey data is not broken down any further than that — no split by gender or education exists in the source data at all, a limitation covered in the guide to net worth by age. A family's net worth percentile here answers one specific, narrower question: where does this particular family sit among other families of a similar age living in the same province, not among Canadians generally and not among some narrower slice the underlying data simply cannot support.
The real data behind these percentiles
Unlike the income percentile estimate described in a separate guide, net worth percentiles on this app rest on real published data rather than a modelled log-normal curve. Statistics Canada's Survey of Financial Security reports net worth broken into quintiles — five equal-sized slices of the population, from lowest to highest — for Canada as a whole and for each province, along with the median net worth of the families within each quintile.
Those five within-quintile medians place naturally at five points along the percentile scale: the lowest quintile's median sits at the 10th percentile, the second quintile's at the 30th, the middle quintile's at the 50th, the fourth quintile's at the 70th, and the highest quintile's at the 90th. That gives a real five-point net worth curve for each province, built from actual survey results rather than an assumed mathematical shape — a meaningfully different footing than the income percentile estimate has, and worth understanding on its own terms.
Rescaling the curve to match age
The published quintile curve is specific to a province as a whole; it is not broken out further by age in the source data. To place a family's net worth in the context of their own age band, rather than their province's entire adult population, this app rescales the whole curve so its middle point lines up with the real province-by-age net worth median described in the net worth by age guide — the same median shown on that province's page for that specific age band.
The rescaling multiplies every one of the five published breakpoints by a single factor: the province-by-age median divided by the province's overall median, which is the curve's own 50th-percentile point. Multiplying every breakpoint by that same factor shifts the entire curve up or down to the right level for that age band while preserving the shape of the spread between its points exactly as published.
This carries an assumption worth naming plainly: age, and the province-by-age median, is treated as shifting the distribution's location — where its centre sits — while the shape of the spread around that centre is assumed to match the province's overall quintile spread, rather than being measured separately for each age band on its own. A younger age band, whose real net worth distribution might in fact be more compressed than the province's overall spread, is shown using a distribution shape borrowed from the wider population instead of one measured specifically for that band.
Linear interpolation between breakpoints
Once the five breakpoints are rescaled, a specific family's net worth is placed on the curve by straight-line, or linear, interpolation between whichever two rescaled breakpoints it falls between — a value that sits halfway in dollar terms between the 30th and 50th percentile breakpoints is reported as roughly the 40th percentile, and the same logic applies along every other segment of the curve.
Linear interpolation, rather than the log-scale interpolation used for this site's income percentile estimate, is a deliberate choice for net worth specifically. Net worth can be zero or negative — a family whose debts exceed its assets, described in the net worth by age guide — and a logarithmic scale is undefined at zero and cannot represent negative values at all. A straight-line method handles the full range of real net worth values, including negative ones, without needing the kind of transformation a log-based method would require near the low end of the distribution.
The extrapolation caveat above the 90th percentile
The published curve stops at five points: 10, 30, 50, 70 and 90. For a net worth value that falls below the 10th-percentile breakpoint or above the 90th-percentile breakpoint, there is no second real data point on that side of the curve to interpolate between, so the app extends the line formed by the nearest two breakpoints outward instead of stopping at the edge of the known data. This is standard extrapolation — continuing a trend beyond the range the data actually covers — rather than interpolation between two known points, and it grows less reliable the further a value sits beyond the last real breakpoint.
This matters most above the 90th percentile specifically, because Canada's wealthiest families are exactly the group whose net worth is most unevenly distributed among themselves, and the published data provides no second breakpoint within that top group — no 95th or 99th percentile figure — to anchor an extrapolation against. A family placed at, say, the 96th percentile by this method has a straight line drawn through the 70th and 90th percentile breakpoints extended outward to reach their net worth; it is not a value read from any real 96th-percentile figure, because no such figure has been published for this site to read from.
The same extension applies, in principle, below the 10th percentile, though it matters less in practice: net worth near the bottom of the distribution is bounded below by how much debt a family can plausibly carry relative to its assets, which limits how far an extrapolation in that direction can drift from the real data compared with the effectively unbounded upside above the 90th percentile.
Two different methods, two different footings
Income percentiles and net worth percentiles on this site are computed by genuinely different methods, and that difference reflects what data is actually available for each measure rather than any deliberate inconsistency between the two. Income percentile thresholds exist only through an interactive tool that cannot be bulk-downloaded, so income percentiles here are modelled from a log-normal curve fitted to a real median and average. Net worth quintile breakpoints, by contrast, are published directly in bulk-downloadable tables, so net worth percentiles here are computed from those real breakpoints, rescaled to the right age band and interpolated exactly as described above.
The practical result is that a net worth percentile from this app rests on firmer ground through most of its range than an income percentile does, resting on genuine survey breakpoints rather than a two-parameter mathematical fit. The one notable exception is the extrapolated territory above the 90th percentile described above, where the net worth estimate's reliability narrows to something closer to the income estimate's uncertainty throughout its own range, for the same underlying reason in both cases: past a certain point along each curve, there simply is no more real published data left for either method to anchor its number to.
Neither method is presented as more authoritative than the other in absolute terms; each is simply the closest estimate this site could build from what Statistics Canada actually makes available in a form suited to a comparison tool covering every province, age band and, for income, gender and education combination at once, rather than one figure looked up by hand.