Adjustment Support in Appraisals and CMAs: A Victoria Guide

Published October 2026. Market figures are from the Victoria Real Estate Board's September 2026 release; appraisal standards are current to CUSPAP 2026.

Every comparable sale tells you what one buyer paid for one home on one day. Turning that into a value for a different home, on a different day, takes adjustments, and an adjustment is only as good as the market evidence behind it. This guide explains how appraisers support each adjustment, works through a complete adjustment grid, and gives you a checklist for testing any CMA or appraisal you are handed in Greater Victoria.

Why every adjustment needs support

The sales comparison approach, which Canadian appraisers usually call the direct comparison approach, rests on substitution: an informed buyer will not pay more for a home than it would cost to buy an equally desirable alternative. Recent sales of similar homes are the best evidence of value, but no two homes are identical and no two sales happen on the same day. Adjustments bridge the gap. Each one answers a narrow question: what would this comparable have sold for if it had the subject's feature, on the subject's effective date?

The direction never changes: you adjust the comparable toward the subject. If the comparable is better in some respect, subtract; if it is worse, add. A comparable with a bigger lot gets a downward lot adjustment, never an upward one. Reversed signs are among the most common errors reviewers find.

What the standards require

Members of the Appraisal Institute of Canada work under the Canadian Uniform Standards of Professional Appraisal Practice (CUSPAP), whose 2026 edition took effect on April 1, 2026. CUSPAP does not prescribe a method for adjustments. It sets a test instead: would a reasonable appraiser, a peer working within an acceptable standard of care and on rational assumptions, have done the same? It requires a report to analyze the relevant data and detail the reasoning behind each approach, and it treats a report that misleads by omission as seriously as one that misleads by commission.

Fannie Mae's Selling Guide, the most widely cited U.S. lending standard, is blunter. Adjustments must reflect the market's reaction, not a rule of thumb; a bare statement that an adjustment was made, without support, is not acceptable; and the final value must fall within the range of the adjusted sale prices. Canadian lenders do not lend under Fannie Mae's rules, but its guide is a useful public benchmark of what lenders regard as defensible.

The elements of comparison, and the order they go in

Appraisers sort the differences between a comparable and the subject into ten elements of comparison. The first five describe the transaction; the last five describe the property. Order matters, because the transactional adjustments put every sale on a common footing (the same rights, cash-equivalent terms, an arm's-length deal, no hidden costs and today's market) before any physical difference is measured.

The sequence

1. Transactional adjustments first. Property rights, financing, conditions of sale and expenditures after purchase convert the price to that of a normal, cash-equivalent, arm's-length deal. In dollars, the order among these four matters little; as percentages, apply them in sequence, each to the price as already adjusted.
2. Market conditions last of the transactional group. The time adjustment carries the deal-adjusted price from the contract date to the effective date, so every comparable is measured against the same market.
3. Property adjustments on the time-adjusted price. Location, physical, economic, use and non-realty differences are then measured against that common baseline. Chattels and other non-realty items are often stripped out at the very start, since they were never part of the real estate.

Illustrative example. A comparable sold for $900,000. The buyer knew the roof was finished and replaced it right after completion for $30,000, and prices in that segment have since fallen 3%. In order, the roof is added first ($930,000) and time is applied to that figure: $930,000 x 0.97 = $902,100. Applying time first gives $900,000 x 0.97 = $873,000, plus $30,000, or $903,000. The $900 gap looks trivial, but sequence errors grow with larger time adjustments and more transactional items. Because the first four transactional elements are the hardest to measure, a sale that needs a large adjustment for one of them is often better given little weight, or dropped.

Percentages, dollars and sequence

Dollar or percentage?

A dollar adjustment adds or subtracts a fixed amount, such as a garage worth $24,000 to buyers. A percentage adjustment scales with price, such as a busy road costing about 4% of value. Features whose contribution does not grow with the price of the house (a carport, a bathroom, a deck) are usually better in dollars; attributes whose effect rises with value (location, market conditions) usually better as percentages. What is never right is treating the same feature in dollars on one comparable and in percent on another, or adjusting a 600 sq ft size difference on one sale while ignoring a 500 sq ft difference on another.

Which way does the percentage run?

"The comparable is 5% inferior" and "the subject is 5% superior" are different statements. Illustrative example: a comparable sold for $800,000. If the subject is worth 5% more than the comparable, the indication is $800,000 x 1.05 = $840,000. If the comparable is worth 5% less than the subject, it represents 95% of the subject's value, and the indication is $800,000 / 0.95 = $842,105. A careful report says which base it used and uses it consistently.

Added or multiplied?

Transactional adjustments are applied in sequence, each to the price as already adjusted, so in effect they compound. Physical adjustments are normally estimated one at a time, a size adjustment from one analysis and a view adjustment from another, and then added; compounding them would multiply effects that were measured separately. The exception is a statistical model built on multiplicative relationships, where compounding is part of the model.

Neither is wrong in itself. The test is whether the method matches how the adjustments were derived, whether the report says which it used, and whether every comparable is treated the same way.

The support toolkit

The strongest support comes from the market itself: pairs of sales, groups of sales, resales and statistical models. Cost, income and opinion evidence are most valuable as cross-checks. Well-supported appraisals use more than one method for the adjustments that matter most, and keep the working in the file even when a short-form report has no room to show it.

Paired sales

Find two sales alike in every material way except the feature you want to measure, sold under similar market conditions; the price difference is one indication of what buyers pay for it. Real pairs rarely differ in only one way, so every other difference must be adjusted out first, and each of those adjustments weakens the result. Sales months apart must first be brought to the same date. One pair is never enough, because either sale might be a forced sale or a family deal. If several pairs disagree, with one showing a premium and another a discount, that is evidence too: the feature may matter little to buyers, as with upper floors in a low-rise building where higher units gain no real view.

Resale analysis

A resale is the same property selling twice; if nothing changed in between, the price change measures the market. Define the segment, pull the resales for the period, and confirm from listing remarks and permit records that no renovation or addition happened between sales. The traps: a buyer's new kitchen inflates the "market" change, and a first sale made under pressure, or a builder's sale with incentives such as included GST, distorts the starting point. Days on market are a useful clue. Convert to a monthly rate only if prices moved steadily.

Grouped data analysis

Sort a larger set of sales into groups that differ in the feature (with and without a suite, corner and interior lots), bring them to a common date, and compare the group medians. More data makes the result less sensitive to one odd sale. The catch is that groups often differ in other ways at once: houses with suites may also be larger, so the raw gap mixes the suite premium with the size premium. Narrow the groups to a common size band and area, or move up to regression.

Statistical analysis, regression and graphs

Regression estimates what each attribute contributes to price across many sales at once. A regression of price on finished area yields a dollars-per-square-foot slope that can serve as a size adjustment; multiple regression adds lot size, age and time. The R-squared statistic shows how much of the price variation the model explains, and a low figure means other factors dominate. Regression needs dozens of clean sales, and real relationships are often curved: an extra square foot is worth less in a larger house, and land value per square foot falls as lots get bigger. Plotting price per square foot against size usually reveals the curve. Variables that move together, such as size and bedroom count, make individual coefficients unstable, and one mistyped square footage can bend the line.

Sensitivity analysis

Try a range of values for one adjustment and see which brings the adjusted comparables closest together. If a suite premium of $70,000 tightens the range and $40,000 or $100,000 widens it, the comparables themselves support something near $70,000. It is a good test of an adjustment produced another way. On its own it is circular: with enough adjustments any set of sales can be made to agree, and a suspiciously tight range after large adjustments is a red flag, not proof.

Depreciated cost

Estimate what a feature would cost new, then subtract depreciation. This works best for improvements that are new, typical for the area and addable: a garage, a deck, a bathroom, a kitchen renovation. It is a good reality check; if paired sales suggest a garage adds ten times its building cost, question the pairs. It fails where cost and value part company. Pools, elaborate custom finishes and over-improvements for the street often return far less than they cost, and depreciation is itself an estimate.

Income capitalization of differences

For a feature that produces rent, capitalize the rent difference with a multiplier taken from local sales. Illustrative example: if homes with and without a feature rent for $150 a month apart, and sales show a monthly gross rent multiplier of 140, the feature indicates about $150 x 140 = $21,000. The multiplier must come from actual sales, not a rule of thumb. Owner-occupiers, who buy most Greater Victoria houses, value a suite partly as income and partly as flexibility, and income from an unauthorized suite carries more risk than income from a permitted one; see the guide to suites and zoning.

Trend analysis, secondary data and interviews

Real estate board price indexes are the usual starting point for time adjustments; the Consumer Price Index is not, because it measures consumer goods and lags housing. BC Assessment values are mass-appraisal estimates as of July 1 of the previous year: the ratio between assessed land values in two areas can corroborate a location adjustment, but assessments are not sales. Interviews with the agents, buyers and sellers involved are the best way to verify a sale and confirm what was included, and the weakest way to measure an adjustment on their own.

When the numbers run out: the hybrid approach

Many good appraisals make quantitative adjustments only for the few attributes that drive most of the value and can be measured, typically time, location, lot size and finished area, and then rank each comparable as inferior, similar or superior on the rest, so the subject is bracketed between better and worse sales. If every comparable is inferior, or every one superior, there is no bracket; widen the search.

Time adjustments with a market index

In a moving market, time is often the largest single adjustment and the one casual CMAs skip most often. The Victoria Real Estate Board's Victoria Core detached benchmark was $1,306,500 in September 2025, $1,301,800 in August 2026 and $1,280,100 in September 2026: a year-old sale needs about a 2.0% downward adjustment, and 1.7% of that came in one month. September 2026 saw 463 sales against 3,811 active listings, a ratio of about 12%; the board treats anything below 17% as a buyers' market.

How the MLS HPI works

The MLS Home Price Index, maintained by the Canadian Real Estate Association and published locally by the board, prices a benchmark home for each property type in each area, built from the median quantities (living area, bedrooms, lot size) and most common qualities of homes there. Statistical models estimate what each attribute contributes to price. Because the benchmark's attributes hold steady, the index tracks price change better than average or median prices, which swing with whatever mix of homes sold that month. Current figures are on the monthly market statistics page.

Applying an index

1. Pin down each contract date: when the deal became firm, not the completion date. Fannie Mae's guide, for example, runs the analysis from each comparable's contract date to the effective date.
2. Pick the matching series: the same property type and the closest area with a reliable benchmark.
3. Compute the factor: the index or benchmark price at the effective date divided by the figure for the contract month.
4. Apply it to the transaction-adjusted price, after rights, financing, conditions of sale and expenditures.
5. Cross-check with resales and paired sales when the span is long or the market turned.

Illustrative example using the board's figures: a Victoria Core detached sale of $1,200,000 that went firm in September 2025, valued as of September 2026: $1,200,000 x (1,280,100 / 1,306,500) = about $1,175,752, a time adjustment of roughly -$24,250.

• The benchmark is not your house. Waterfront homes, entry-level condos and unusual properties can move differently, and small sub-areas produce noisy numbers.
• No straight lines across a turn. A 12-month change spread evenly misleads when prices rose for six months and then fell; Fannie Mae's guide notes the 12-month trend may differ from the change over a comparable's own window.
• No double-adjusting. Once a sale is index-adjusted, do not adjust it again because "spring was hotter". A sale that went firm in the effective month may need no time adjustment at all.

Gross, net and reconciliation

Net adjustment is the sum of all adjustments with their signs; gross adjustment ignores the signs. A comparable with a -$40,000 view adjustment and a +$45,000 lot adjustment has a net adjustment of only +$5,000 but a gross adjustment of $85,000. Net figures can hide big, offsetting judgments; gross figures show how far the appraiser has moved from the hard evidence of the sale price.

There is no official limit. Fannie Mae's guide sets none and says the size or number of adjustments alone is not a reason to reject a comparable. In practice, reviewers start asking questions as gross adjustments approach a quarter of the sale price, and a comparable adjusted by half its price or more is rarely more than a weak indicator. Large adjustments are not wrong in themselves; unsupported ones are.

Weighing the comparables

• Fewest and smallest adjustments, measured by gross percentage.
• Best-supported adjustments: one adjustment from three consistent pairs beats a guessed condition adjustment.
• Most similar location and property type, including the same building or complex for strata.
• Most recent contract date, or the smallest time adjustment.
• Verified, arm's-length data, confirmed with someone involved in the deal.

Do not simply average. Adjustments should narrow the range of indications; if the range widens, the adjustments or the comparables are wrong. The final value should fall within the adjusted range. And a range that is suspiciously tight after large adjustments can mean the numbers were worked backward from a target, the first thing a review appraiser looks for.

Double counting and contributory value

An adjustment should equal a feature's contributory value: what it adds to the whole property in buyers' eyes, not what it cost. Double counting adjusts for the same contribution twice. The usual suspects:

• Age and condition. A condition adjustment usually captures most of what age means to buyers.
• Price per square foot and size. If the comparison is already per square foot, a separate size adjustment is usually redundant, unless the size gap is large enough for the per-foot rate itself to change.
• Suite, finished area and bathrooms. A suite brings area, a kitchen and a bathroom; adjusting for each and then adding a full suite premium counts some of it twice.
• View and location. Do not credit the same ocean once as location and again as view.
• Lot size and zoning potential. If the lot adjustment came from sales that already reflect multi-unit potential, a separate "development upside" premium repeats it.
• Renovation and overall condition, which often describe the same work.
• Correlated variables in a regression, such as size and bedroom count.

A worked adjustment grid

Illustrative example. Every figure in this section is hypothetical, including the price index; it shows the mechanics, not local rates. The subject is a suburban detached house: 2,000 sq ft finished, a permitted one-bedroom suite, a quiet street, a double garage, updated throughout, with a lot similar to the comparables. The effective date is October 2026.

Step 1: the time index

Factor = 300.0 divided by the contract-month index. Each time adjustment in the grid uses the full factor and is rounded to the nearest $100.

Step 2: paired sales for the suite

In Pair 2 the home with the suite is 100 sq ft larger, so 100 x $150 = $15,000 comes off the $86,000 gap, leaving $71,000.

The pairs indicate $68,000 to $71,000; the grid uses $70,000. Pair 2 shows why other differences come out first: its raw gap overstates the suite premium by the value of the extra floor area. Because the pairs were matched on size, the figure measures the suite premium over and above its area. All three involve permitted suites; an unauthorized suite would need its own evidence.

Step 3: paired sales for a busy road

Each pair is two similar homes sold in the same month. They indicate $38,000 to $43,000, about 3.2% to 3.8% of the quiet-street price. The grid uses $40,000 for the one comparable on an arterial road.

Step 4: the other adjustments and their support

• Finished area, $150 per sq ft, from a regression of 40 sales in the 1,700 to 2,300 sq ft band (hypothetical R-squared 0.61). The size gaps in the grid are small, so a straight-line rate is reasonable.
• Parking, $24,000. The extra garage space would cost about $40,000 new; less 40% depreciation gives $24,000. Two paired sales indicated $20,000 and $26,000.
• Condition, -$35,000 for Comp 2. The seller reported spending $78,000 on a 2025 renovation, but two paired sales showed buyers paying only $30,000 to $40,000 more for a recent full renovation than for a well-kept, updated home. Cost overstates value, so the market figure is used.
• Expenditure after purchase, +$24,000 for Comp 3. The buyer had a $24,000 roof quote in hand when the price was agreed and replaced the roof right after completion. It is added before the time adjustment.

Step 5: the grid

On a phone, swipe the table sideways to see all four comparables.

Percentages are of each sale price. Comp 3 shows the sequence at work: the $24,000 roof is added to the $1,205,000 price first, and the time factor is applied to the resulting $1,229,000 ($1,229,000 x 300.0 / 309.0 = $1,193,204, rounded to $1,193,200). Area adjustments are the size gap times $150; for Comp 1, (2,000 - 1,850) x $150 = $22,500.

Step 6: reconcile

Before adjustment the sales spanned $1,185,000 to $1,268,000, a range of $83,000. After adjustment they span $1,219,800 to $1,271,600, a range of $51,800. Well-supported adjustments narrow the range, and these did.

• Comp 4: Smallest adjustments (1.7% gross); recent.
• Comp 1: Most recent, but relies on one large suite adjustment.
• Comp 3: Oldest sale; transactional and time adjustments in sequence.
• Comp 2: Largest gross adjustment; different street type; renovation judgment.

The weighted indication of $1,245,870 is reconciled at $1,246,000. It sits inside the adjusted range and is close to the mean ($1,242,500) and median ($1,239,300) of the adjusted prices, a useful sanity check. The weights are a judgment, but a written one that a reader can follow and challenge.

Where CMAs fall short

A comparative market analysis (CMA) is a pricing opinion, usually prepared by a licensee to help set a list price or shape an offer. It is not an appraisal under CUSPAP and has no standard form, so quality varies widely. A good CMA uses the same logic as an appraisal. A weak one shows the same handful of problems:

Pricing pressure can run either way, high to win a listing or low to frame an offer; the test is the same. In a multiple-offer situation, buyers often remove their financing condition, so if the lender's appraisal later comes in below the price, the buyer must cover the gap in cash. See the mortgage financing guide for how lenders use appraisals, and the guide to writing an offer for how subject clauses protect you.

Greater Victoria pitfalls

Price per square foot

Dollars per square foot falls as homes get larger, so a small house always looks "expensive" per foot. In older parts of Victoria, Oak Bay and Saanich much of a detached home's value is in the land, so the building's per-foot price says little on its own. Finished basements may be counted in total finished area on one listing and not on another, and a strata plan's unit area can differ from a listing's measurement. See how widely the figure varies even among new Victoria condo buildings.

View premiums

A view is worth what buyers pay for it, which depends on its quality (open ocean or a glimpse), its permanence (protected by topography or open to the next building) and where it is enjoyed. For condos, stacked units in the same building, sold close together in time, make the best pairs.

Suites

A permitted suite and an unauthorized one are different features with different evidence, and lenders, insurers and buyers treat them differently. Adjust from sales with the same legal status, and confirm that status rather than relying on a listing description; the guide to suites, zoning and Bill 44 explains how.

Lot size and the new zoning

Under the Province's small-scale multi-unit housing rules, most lots formerly zoned for single-family or duplex use now permit three to six homes, depending on lot size and transit. A lot with real redevelopment potential (width, a workable shape, lane access, a frequent bus route) may sell on land value rather than on the house, which is a use and zoning question and the point where lot-size effects stop being linear. Covenants and building schemes on title can still limit what is built; see the guide to title searches and covenants.

Strata versus freehold, and different areas

Do not mix strata and freehold sales. Strata comparables carry their own variables (fees, contingency reserve, special levies, the condition shown in the depreciation report), and a special levy the buyer must pay is an expenditure-after-purchase item. Likewise, the board publishes benchmarks by area and property type because the Core, the Westshore and the Saanich Peninsula do not always move together. Use the series that matches the comparable, not the region-wide headline. More in the guide to buying a condo.

Checklist: test any CMA or appraisal

• Are all the comparables completed sales, with contract dates shown, and any listings labelled as supporting evidence only?
• Are they from the same segment: property type, area, age band and legal status of any suite?
• Which sales were considered and rejected, and why?
• Was each sale verified as arm's length, with any concessions, chattels or known repairs identified?
• Is there a time adjustment, what supports it, and does it run from the contract date?
• Are transactional adjustments applied before physical ones, with time last of that group?
• For each adjustment, what is the support, and is it written down where you can see it?
• Do the adjustments run in the right direction: superior features subtracted, inferior ones added?
• Is each feature treated the same way on every comparable?
• Any double counting: age and condition, suite and area, price per square foot and size, view and location?
• What are the gross and net adjustments for each comparable, and did the adjustments narrow the range?
• How were the comparables weighted, and does the conclusion fall inside the adjusted range?
• Have suite status, zoning and title been confirmed, not assumed?

Questions and answers

Photo from Wikimedia Commons, used under the licence linked in the caption.