Question 10
Question 10
Why Comparable Sales Are Wrong for Business Valuation
The market approach is the most commonly used and least reliable method for valuing private businesses. The data is incomplete, the comparisons are false, and the conclusion depends on transactions whose circumstances are unknown. Here is why — and what should be used instead.
By Eric Jordan, CPPA — International Business Valuation Specialist | Expert Witness (Canada)
Fee Range: $1,500 – $15,000 | Basic Average: $3,500 | 877-355-8004 | Timeframe: 1 to 2 weeks from when documents are available
In This Guide
Why Comparable Sales Are So Appealing
The Seven Structural Flaws in Private Business Comparable Sales
The Survivorship Problem: What the Databases Never See
What Transaction Databases Actually Contain — and What They Do Not
The Fair Market Value Problem: Why Most Comparable Sales Do Not Qualify
The Intangible Asset Problem: Why No Two Businesses Are Comparable
What Canadian Courts Say About Comparable Sales Evidence
The Proper Role of Market Data in a Business Valuation
What Should Be Used Instead: The Evidence-Based Alternative
Frequently Asked Questions
The logic of comparable sales is borrowed from real estate appraisal, where it works well. A three-bedroom house in a specific neighbourhood is genuinely comparable to another three-bedroom house in the same neighbourhood. The physical characteristics are observable, the sales are publicly recorded, the financing terms are disclosed, and the market is relatively liquid. A buyer can verify the comparable.
The appeal of applying this logic to business valuation is obvious. Instead of projecting future cash flows, selecting discount rates, and making assumptions about growth, the valuator can point to real transactions and say: “Similar businesses sold for this multiple, therefore your business is worth that multiple applied to your earnings.” The method is efficient, intuitive, and has an air of objectivity. It feels like market evidence rather than estimation.
The problem is that none of the conditions that make comparable sales work in real estate exist in private business valuation. And the appearance of objectivity conceals a foundation built on unverifiable data, false comparisons, and undisclosed circumstances.
Flaw 1: The circumstances of the sale are unknown
Fair market value requires a transaction between willing parties, neither under compulsion, both with reasonable knowledge. But private business sales frequently involve compulsion the database does not disclose. An owner selling because of illness, divorce, partnership dissolution, or financial distress is a motivated seller whose price does not reflect fair market value — it reflects urgency. An employee or family member purchasing at a discount is not an arm’s-length buyer. A sale forced by a bank calling a loan is a liquidation, not a market transaction. The database reports a price. It does not report the reason for the sale. Without knowing the reason, there is no way to determine whether the price meets the fair market value standard.
Flaw 2: The deal structure is not disclosed
A business that sold for $1,200,000 with $1,200,000 in cash at closing is a different transaction than one that sold for $1,200,000 with $400,000 in cash, $400,000 in seller financing at below-market interest, and $400,000 in earnout payments contingent on performance. The present value of the second deal is materially less than $1,200,000. Yet the database reports both as “$1,200,000 sales.” Deal structure — seller financing, earnouts, holdbacks, working capital adjustments, consulting agreements, non-compete payments — can account for 20% to 40% of the reported price. When the structure is unknown, the reported price is meaningless as a comparable.
Flaw 3: What was included in the sale is ambiguous
Did the $1,200,000 include real estate, or was the real estate leased? Was $300,000 in equipment included, or does the buyer need to invest separately? Was inventory included at cost, at market, or not at all? Was a non-compete agreement part of the consideration? Transaction databases vary in whether they disclose these details, and many do not. A valuator comparing a $1,200,000 sale that included real estate to a business with leased premises is not comparing comparable transactions — they are comparing different bundles of assets at different prices.
Flaw 4: The intangible asset composition is never disclosed
Two dental practices with the same revenue and the same number of patients can have entirely different intangible asset profiles. One may have a patient base that schedules with “the office” and will remain after a change of ownership. The other may have a patient base loyal to the specific dentist who will follow them to a new practice. The first has high enterprise goodwill. The second has high personal goodwill. The first is worth significantly more to a buyer. The database reports a multiple for each. The multiple tells you what someone paid. It does not tell you what they received — and whether what you are selling is the same thing.
Flaw 5: Owner dependency varies enormously between businesses
A comparable sale of a landscaping company at 2.5 times seller’s discretionary earnings tells you nothing about whether that company had a functioning management team or was entirely dependent on the owner. If the comparable had low owner dependency and a transferable client base, and the subject business has high dependency and an owner who is the sole client relationship, the multiple is not transferable. Owner dependency can reduce value by 25% to 50%, and transaction databases do not measure it.
Flaw 6: The data is self-reported and unaudited
Transaction databases rely on brokers, buyers, and sellers to report sale details. The data is not audited, not verified, and not standardized. Reporting is voluntary — transactions that do not close, or that close at unfavourable terms, are less likely to be reported. This creates survivorship bias: the database overrepresents successful transactions and underrepresents distressed sales, failed deals, and transactions at the low end of the market. The multiples derived from the database are therefore biased upward.
Flaw 7: The sample size is almost always too small
For a comparable to be meaningful, the transactions should be in the same industry, of similar size, in the same geographic market, and recent enough to reflect current conditions. When these filters are applied to any transaction database, the number of remaining comparables typically shrinks to 3 to 8 — and sometimes to zero. Drawing conclusions from a sample of 5 transactions, each with unknown circumstances, unknown deal structures, and unknown intangible asset compositions, is not statistical analysis. It is anecdote dressed as data.
The compound effect: Each of these seven flaws might be individually manageable. But they are not independent — they compound. A comparable sale with unknown circumstances, undisclosed deal structure, ambiguous asset inclusion, unknown intangible composition, unmeasured owner dependency, unaudited data, and a sample size of five is not evidence of value. It is a number wrapped in seven layers of uncertainty, presented as a market-derived conclusion.
The seven flaws above describe problems with the data that is in the databases. There is a more fundamental problem: the vast majority of business outcomes never enter a transaction database at all. The databases capture only a narrow, self-selected slice of the market — and that slice is systematically biased toward outcomes that do not represent the full picture.
Only about 25% of listed businesses actually sell
Industry data consistently shows that roughly 70% to 80% of businesses listed for sale never close a transaction. The owners eventually delist, retire, or shut down. These businesses had a value low enough, a dependency high enough, or a problem severe enough that no buyer was willing to pay a price the seller would accept. They represent the majority of the market. They are invisible to the database. Every multiple derived from completed sales excludes the businesses that could not sell at any price — which means the database systematically overstates what “the market” pays, because the market also includes the businesses it rejected.
Businesses that liquidate are never counted
For every business that sells through a broker, there is another that simply closes. The owner retires and locks the door. The lease expires and the equipment is auctioned. The clients drift away when the owner stops answering the phone. These businesses had value at some point — they generated revenue, employed people, served customers — but their value was not transferable, typically because of extreme owner dependency, undocumented processes, or personal goodwill that could not survive a transition. These outcomes represent a real and significant portion of the business lifecycle. They are entirely absent from transaction databases. A market approach that ignores every business that failed to sell or chose to liquidate is not measuring “the market” — it is measuring the winners and calling it the average.
The best businesses are sold privately, under unknown terms
The most valuable, most transferable businesses — the ones with strong management teams, diversified client bases, documented operations, and low owner dependency — often never appear in a broker’s listing. They are sold to insiders: a long-time employee who has been groomed for succession, a family member who has been involved in the business for years, a key client or supplier who understands the value intimately, or a competitor who approached the owner directly. These transactions happen privately, under terms negotiated between people who know each other, with deal structures (seller financing, transition consulting, earnout arrangements, non-compete provisions) that are never reported to any database.
The result is a double distortion. The worst businesses (those that could not sell) are excluded from the bottom of the database. The best businesses (those that sold privately to insiders) are excluded from the top. What remains in the database is the middle — businesses that were marketable enough to attract a buyer through a broker but not desirable enough to be acquired privately. The multiples derived from this middle band are then applied to all businesses, including the ones that would never have appeared in the database in the first place.
Only broker-reported sales are counted
Transaction databases are populated by business brokers who voluntarily report the details of their closed transactions. Sales conducted without a broker — the owner who places a local advertisement, the transaction negotiated through the owner’s accountant or lawyer, the handshake deal between the owner and a trusted employee — are not reported. These non-brokered transactions represent a significant share of all business transfers, and they occur under conditions (personal relationships, informal negotiations, non-standard terms) that are likely to produce prices different from brokered transactions. Excluding them introduces another layer of selection bias into the data.
Consider the filters that must be applied to produce a genuinely reliable comparable:
By the time every necessary filter is applied, the number of genuinely comparable, verifiable, arm’s-length, non-compelled, fully disclosed transactions in a relevant industry and geography is not a percentage of the database. It is a handful at best — and often zero. The honest estimate is that the databases capture perhaps 2% to 5% of all business outcomes in a given industry, and that within that 2% to 5%, the proportion of transactions that meet every condition of a reliable comparable is a fraction still. The valuator is drawing conclusions about what a specific business is worth from a data set that represents a small, biased, and unverifiable sample of what actually happens in the market.
The question to ask any valuator who relies on comparable sales: “What percentage of all business outcomes in this industry — including businesses that liquidated, businesses that could not sell, businesses sold privately to insiders, and businesses sold without a broker — are represented in the database you used? And of those that are represented, how many can you verify were arm’s-length transactions without compulsion?” If the answer to both questions is “I don’t know,” the comparable sales are not market evidence. They are a curated subset presented as if they represent the whole.
The left column is the input. The right column is the context needed to interpret the input. A multiple without context is a number without meaning. It is like knowing that a house sold for $800,000 without knowing whether it had three bedrooms or eight, whether it was a foreclosure or a bidding war, and whether the roof leaked or was just replaced. Nobody would value a house on that basis. Yet this is the standard of evidence commonly accepted in business valuation.
Fair market value is defined as the highest price available in an open and unrestricted market between informed, prudent parties acting at arm’s length, with neither under compulsion. This definition has five requirements, each of which must be met for a transaction to qualify as a fair market value comparable:
Any transaction that fails one or more of these requirements does not represent fair market value. In the private business market, the majority of transactions fail at least one. Using these transactions as evidence of fair market value is a methodological contradiction: the data does not meet the standard it is being used to support.
The comparable sales method assumes that businesses in the same industry with similar revenue or earnings are substitutable — that a buyer choosing between them would pay approximately the same price. This assumption may hold for standardized assets (a barrel of oil, a government bond, a square foot of warehouse space), but it fails for businesses, because the majority of a business’s value is intangible, and intangible assets are unique to each enterprise.
Consider two accounting practices, each generating $800,000 in annual revenue and $250,000 in seller’s discretionary earnings:
A transaction database would report both practices as “accounting firm, $800K revenue, $250K SDE.” They would be treated as comparables. But Practice A might be worth $750,000 to a buyer and Practice B might be worth $350,000 — a difference of more than 100% — because the intangible asset composition, owner dependency, and transferability are entirely different. The database cannot distinguish between them because it does not measure the factors that determine value.
The fundamental problem: Comparable sales work when the assets being compared are standardized. Business value is not standardized. It is created by intangible assets that are unique to each enterprise, observable only through on-site inspection, and measurable only through systematic identification. A method that cannot see these assets cannot compare them. And a comparison that cannot account for the most significant components of value is not a comparison — it is a coincidence of financial metrics.
Canadian courts accept market approach evidence, but they subject it to scrutiny that exposes the same problems described above. Courts have consistently required that:
The comparables must be genuinely comparable. The valuator must demonstrate that the businesses used as comparables are similar in size, industry, geography, and financial characteristics — not merely in the same SIC/NAICS code. Courts have rejected comparable sales evidence where the “comparable” businesses were in different markets, of different sizes, or facing different competitive conditions.
The source and reliability of the data must be disclosed. A valuator who cannot explain where the comparable data came from, how it was verified, or what the database does and does not contain will face difficulty under cross-examination. The opposing expert will ask: “Do you know whether this sale was arm’s length? Do you know whether seller financing was involved? Do you know the owner dependency level?” If the answer is no to each, the comparable is effectively impeached.
Adjustments for differences must be explained. Applying an unadjusted multiple from a comparable to the subject business implies they are identical in every relevant respect. Courts expect the valuator to identify differences and adjust the multiple accordingly. If no adjustments are made, the court may conclude that the analysis is superficial.
In practice, when comparable sales evidence is weak or unverifiable, Canadian courts give primary weight to the income approach. The market approach survives not as the basis of the conclusion but as a secondary reasonableness check — which is the role it should play.
Comparable sales data is not useless. It is misused when it carries the primary burden of the valuation. Its proper role is as a secondary cross-check — a test of whether the income-based conclusion is within the range of observed market activity.
Used properly, market data answers this question: “Is the value I have concluded using the income approach consistent with what buyers are generally paying for businesses with similar financial profiles?” If the income approach produces $1,200,000 and market multiples suggest a range of $900,000 to $1,500,000, the conclusion is within market bounds. If the income approach produces $2,500,000 and every market multiple suggests $800,000 to $1,200,000, there is a discrepancy that must be explained — either the income analysis has a flaw, or the subject business has characteristics that genuinely distinguish it from the comparables.
This is a fundamentally different use than deriving the value from the comparable. It does not require the comparables to be identical to the subject business. It does not require the transaction circumstances to be known. It requires only that the market data provides a plausible range against which the primary conclusion can be tested. In this role, the limitations of the data are manageable because the data is not carrying the weight of the conclusion.
If comparable sales cannot reliably serve as the primary basis of value, what should? The answer is the income approach — but only when supported by the evidence that makes it reliable.
The income approach alone has its own weaknesses. The capitalization rate is sensitive to small changes. The normalized earnings depend on adjustments that can be subjective. The growth assumption is inherently uncertain. These weaknesses are real, but they are addressable — through the kind of evidence that comparable sales databases cannot provide.
The 25 Factors Affecting Business Valuation
The 25 Factors framework systematically identifies and measures the value drivers and risk factors that determine what a specific business is worth. Each factor corresponds to an observable, measurable characteristic: client base composition, owner dependency, workforce depth, operational documentation, competitive positioning, brand strength, technology assets, contractual assets, regulatory exposure, and others.
These are the factors that explain why Practice A above is worth $750,000 and Practice B is worth $350,000 despite identical financial profiles. They are the factors that transaction databases do not capture. They are the factors that determine the company-specific risk premium in the income approach — the single most subjective and influential input in the valuation formula. When these factors are measured through direct observation rather than assumed from comparable data, the income approach becomes evidence-based rather than assumption-based.
The 5 Senses Inspection Report
The 25 Factors cannot be populated from financial statements. They require on-site presence. The 5 Senses Inspection Report is the methodology for gathering this evidence: observing who clients interact with, how decisions are made, where institutional knowledge lives, what the physical workspace reveals about operational reality, and how the business functions when the owner is not directing every action.
The on-site inspection produces the evidence that comparable sales databases do not contain and cannot contain: the operational reality of the specific business being valued. It converts the company-specific risk premium from a guess into a documented finding. It identifies the intangible assets that constitute the majority of value. It assesses owner dependency through direct observation rather than inference. And it provides the evidentiary basis for every input in the income approach that would otherwise rest on assumption.
How this changes the valuation
A valuation built on the 25 Factors and supported by the 5 Senses Inspection does not need comparable sales to establish value. It has something better: direct evidence about the specific business being valued. The evidence supports the normalized earnings (because the normalization is informed by what the valuator observed, not just what the financials reported). It supports the capitalization rate (because the company-specific risk factors are identified and documented). It supports the intangible asset values (because each category is assessed through observation). And it supports the conclusion (because the conclusion is traceable from the evidence through the methodology to the number).
Comparable sales can still serve as a secondary cross-check. But they are no longer required to carry the weight of the conclusion — and the valuation is stronger for it.
The core argument: Comparable sales tell you what someone paid for a different business under unknown circumstances. The 25 Factors and 5 Senses Inspection tell you what a specific business is actually worth based on observable evidence. One is inference from incomplete data. The other is conclusion from direct observation. In every context where accuracy matters — court, CRA, lending, acquisition — evidence beats inference.
It is a starting point, not a conclusion. A broker using a multiple from a database is giving you a range based on what other businesses in a broad category have reportedly sold for. It does not account for your specific client base, owner dependency, intangible assets, workforce, or operational risk. The multiple might be correct. It might be 50% too high or 50% too low. Without understanding the characteristics of your specific business, there is no way to evaluate whether the multiple is appropriate. An independent valuation examines your business specifically, not a category it happens to fall into.
Professional standards require valuators to consider all three approaches (income, asset, and market) and explain why each was used, modified, or rejected. The requirement is to consider, not to rely on. A valuator who considers the market approach and explains that reliable comparables are not available — disclosing the data limitations, the lack of comparable transactions in the relevant industry and geography, and the absence of verifiable deal terms — has satisfied the professional standard. The standard recognizes that market data is not always available or reliable, and does not require its use when it is not.
Comparable sales are most reliable for businesses that are relatively standardized in their operations and where transaction data is abundant and well-documented. Franchised businesses with prescribed operating models, standardized gas stations, coin laundromats, and certain retail formats have more homogeneous characteristics than professional practices, technology companies, or service businesses. Even in these cases, however, the comparables must be verified for deal structure, asset inclusion, and transaction circumstances. The more unique the business’s intangible asset profile, the less reliable the comparable sales method becomes.
This is an opportunity under cross-examination. The opposing valuator can be asked to identify the source of each comparable, verify whether the transaction was arm’s length, disclose the deal structure, explain the intangible asset composition of the comparable business, and describe the owner dependency level. If they cannot answer these questions — and they typically cannot, because the data does not contain this information — the foundation of their conclusion is exposed as incomplete. A valuator whose methodology is evidence-based rather than database-derived is in a stronger position in adversarial proceedings.
CRA expects a genuine attempt to determine fair market value, not a specific methodology. If the valuation explains why comparable sales data is unreliable for the specific business — because adequate comparables do not exist, because the available data does not meet the fair market value standard, or because the subject business has characteristics that distinguish it from the comparables — and instead uses a well-supported income approach with thorough intangible asset identification, CRA is unlikely to challenge the methodology. CRA is more likely to challenge a valuation that relies on unverified comparables than one that uses a supported alternative and explains the basis for doing so.
Not necessarily. Extracting multiples from a database is fast. But the on-site inspection, intangible asset identification, and evidence-based risk assessment that replace the comparable sales methodology also serve every other section of the report — the business description, the financial normalization, the owner dependency assessment, and the conclusion. These are not additional steps; they are the steps that a thorough valuation should include regardless of which approach is used. A valuation that skips these steps to save time and cost is saving time and cost in exactly the places where accuracy requires investment.
Speak Directly With the Valuator
Contact Eric Jordan, CPPA — Expert Witness (Canada)
Toll-free & available 24/7 · Canada-wide
877 355 8004
[email protected]
© Eric Jordan — International Business Valuation Specialist | Expert Witness (Canada)
Contact Eric Jordan, CPPA — Expert Witness (Canada)
Toll-free & available 24/7 · Canada-wide