I am Ming Ren, a Broker with RE/MAX Realtron.
The plain conclusion first: neither the average nor the median is "wrong" — they answer different questions. The mistake is using the wrong one for the job. For "what does a typical home cost," use the median. For the long-run trend, use the HPI benchmark price. The average is best reserved for totals, like overall dollar volume.
The math, plainly
Average price: add up every sale price in the month and divide by the number of sales.
Median price: line up every sale price from lowest to highest and take the one in the middle. Half the sales are above it, half below.
HPI benchmark price: the MLS® Home Price Index uses a statistical model to define a "typical" home of each type and tracks that typical home's price over time — like-for-like, rather than a mix of whatever happened to sell.
An invented example, only to show the arithmetic: an area records five sales in a month — $800,000, $850,000, $900,000, $950,000, and $3,000,000 (the last is a luxury home).
- Average = (800 + 850 + 900 + 950 + 3,000) ÷ 5 = $1,300,000.
- Median = the middle sale = $900,000.
Same five sales, a $400,000 gap. The average was dragged away by one luxury sale; the median does not care how expensive that sale was, only where it ranks. That is the core difference: the average is sensitive to extremes, the median is not.
Where each one misleads
The average's trap is mix. It can rise simply because luxury homes made up a bigger share of that month's sales, and fall because cheaper types sold more — while no individual home's price changed at all. In low-volume areas this is extreme: one or two luxury sales can lift a whole district's average.
The median's trap is that it only sees the middle. It tells you the middle sale's price and nothing about the ends: entry-level prices can be falling while luxury prices rise, and the median sits still. In very low-volume months the median itself jumps around, because the "middle sale" is literally a different house each time.
The HPI's limit is that it is a model. A benchmark price is a modelled typical home, not any real transaction. It is published with a lag, gets revised, and should never be used to price your specific home. Its one best use: the direction of like-for-like price trends.
How I use the three together
- Pricing or offering on a specific home: skip all three and use recent comparable sales of similar homes — the only evidence that applies to one property.
- Feeling the market's temperature: the median — it is closest to what the typical buyer actually paid.
- Judging trend direction: the HPI benchmark, year-over-year and over several consecutive months — it filters mix noise the cleanest.
One discipline on top: never conclude anything from a single month. If the average jumps 8%, ask first whether every home rose, or whether expensive homes simply sold more that month. Put volume, mix, and the HPI side by side and the answer usually shows itself.
Common misunderstandings
"The news says the average rose, so my home rose by the same percentage." The average mixes real price change with mix change; it cannot be applied proportionally to your home. Your home's movement shows in its segment's HPI and your area's actual comparable sales.
"The median is below the average, so one of them is cooked." No. A median below the average is the normal shape of housing prices — a minority of high-priced sales pulls the average up, and the middle sale naturally sits below it. A wide gap just means a wide price spread.
"One of the three must be the 'real' number." All three are real; they measure different things. Treating them as contradictions means the definitions have not been separated yet.
Action checklist
- Before quoting any price figure, confirm whether it is an average, a median, or an HPI benchmark — and do not cite it if you cannot tell.
- Trends: HPI over three months or more. Temperature: median. Average: reference only.
- Pricing or offering: go straight to comparable sales and skip all three.
- In low-volume areas and months, discount all three — small samples make every measure jumpy.
- When sharing market numbers with family or partners, state the measure alongside the number.
Trust Professional, Diligent Work. I am Ming Ren. If a market statistic does not make sense, send me the original wording and we will settle the definition before we discuss the conclusion.
This article is general information only, not financial or legal advice. Statistical definitions follow the publishing body's current methodology notes.
Primary sources: Toronto Regional Real Estate Board (TRREB) Market Watch definitions of average and median price; Canadian Real Estate Association (CREA) MLS® Home Price Index methodology (consulted November 2026). The $800,000–$3,000,000 figures are illustrative only, not real sales.


