A company whose revenue has fallen should not be valued on a blended average that mixes the peak with the collapse.
A company whose revenue has fallen should not be valued on a blended average that mixes the peak with the collapse. Identify whether the decline is seasonal or cyclical, structural in the market, or caused by the owner’s pricing, staffing, or neglect. Weight the periods that represent maintainable earning capacity, insert known cost shocks such as a rent reset, and compare the going-concern indication to an orderly asset floor. If cash flow after those adjustments cannot support a buyer and a lender, the value is closer to tangible property and transferable rights than to a historical multiple.
What to know before using the headline number
- Do not average a peak year with a collapse unless you can prove the peak is a normal point on a repeating cycle.
- Name the decline: cyclical, structural, or owner-caused. The method follows the diagnosis.
- Rebuild units, price, mix, and cost before you pick an income, market, or asset approach.
- Known shocks such as a rent reset belong in the base case. They are not optional sensitivities.
- When going-concern cash flow cannot clear a buyer’s stack, the indication moves toward an orderly asset floor.
Averaging a peak and a collapse is not a method. It is a disguise.
Sellers facing two down years often propose a three-year average so the old peak still influences the price. That arithmetic is only defensible when the peak is part of a repeating cycle and the current year is a normal trough. It is not defensible when covers have fallen from 88 to 73, food cost has moved from 31.8 percent to 37.1 percent, and a lease reset is nine months away. Revenue Ruling 59-60 asks for earning capacity as of the valuation date, not for a memorial service for a better year. If the capacity is gone, the average is a wish.
Build a year-by-year and monthly bridge before any multiple is discussed. Show units, price, mix, and the cost lines that moved with them. Census seasonal-adjustment methodology is a useful reminder that recurring calendar effects should be separated from underlying movement. A restaurant that always slumps in January is not “declining” in January. A restaurant that has lost 22 percent of peak-year revenue over two years, with a matching drop in covers, is not a seasonal story. Label the pattern before you weight it.
Separate cyclical, structural, and owner-caused decline
Cyclical decline can reverse without a new concept: a contractor who is light on bids because local housing turns are down, a dry cleaner whose office corridor is half empty until leases re-fill, an ecommerce seller riding a product fad that has a measurable remaining life. Structural decline is different. A dry cleaner still dependent on perchloroethylene in a tightening regulatory market, an ecommerce catalog whose paid-traffic economics broke, or a restaurant whose traffic source was a now-closed office tower will not be cured by “better marketing.” Owner-caused decline is different again: hours cut, recipes cheapened, reviews ignored, the chef-owner absent, prices frozen while vendors were not.
The valuation treatment follows the diagnosis. Cyclical names may justify a cycle-average or a trough-to-mid-cycle case with a higher risk rate. Structural names belong closer to an asset floor plus any remaining transferable rights, or to a turnaround case that is explicitly a project, not a historical multiple. Owner-caused decline can be the most salvageable—if a buyer can see that the demand is still there and the neglect is reversible—but the buyer, not the seller, usually captures that upside. Do not charge a going-concern multiple for a turnaround the owner failed to execute.
Rebuild the driver stack before choosing income, market, or asset methods
For a restaurant, the drivers are covers, average check, food and labor cost, occupancy, and remaining lease term. For a dry cleaner, they are piece counts, solvent or wet-clean mix, route versus plant work, and environmental condition. For an ecommerce brand, they are contribution margin after ads, repeat-purchase rate, platform dependence, and inventory aging. A top-line percentage decline that is not decomposed into those drivers cannot support a capitalization rate or a multiple. You cannot know whether you are looking at a price problem, a traffic problem, or a cost problem.
Normalization still matters, but it cannot rescue a broken run rate. Adding back a one-time repair does not restore lost covers. Removing the owner’s car does not change a rent reset. Use the same adjustment policy across years so you can see the real trend. Then decide which method is even applicable. An income approach on a collapsing run rate needs a defensible maintainable number and a risk rate that admits the trend. A market approach needs companies with a similar trajectory, not peak-year listings. An asset approach becomes the floor when the going concern is no longer the most credible premise.
The asset floor is a discipline, not a threat
Orderly liquidation or adjusted book value is not an insult. It is the amount a buyer could realize from furniture, equipment, transferable licenses, remaining inventory, and any assignable leasehold interest after selling costs. IRS tangible-property rules help separate routine repairs from betterments and restorations; a buyer will make a similar cut when deciding what must be capitalized on day one. Fully depreciated kitchen equipment can still have a used-market bid, and recently capitalized leasehold improvements can have zero transfer value if the lease is short or the landlord will not consent.
Compare that floor to the going-concern case after the rent reset and a realistic labor model. If the going-concern indication is only slightly above the floor, the extra amount is thin goodwill and should not be dressed up as a historical multiple. If the going-concern case is below the floor, the company may be worth more closed or sold in pieces than operated, which is an unpleasant but necessary conclusion. SBA close-or-sell planning includes winding down as a real option. A valuation that refuses to look at it is advocacy.
Put known shocks—especially a rent reset—in the base case, not the footnotes
A lease that steps from $7,420 a month to $10,350 a month in nine months is not a sensitivity. It is a scheduled $35,160 annual cost increase that the current owner has not yet had to carry for a full year. Any maintainable-earnings figure that ignores it overstates capacity. The same is true of a known insurance non-renewal, a minimum-wage step, a lost anchor tenant next door, or a required hood-system replacement. Buyers will insert those items. Sellers who insert them first keep control of the narrative.
Work the shock through contribution margin, not just through an expense line. If the restaurant is already at 37.1 percent food cost and declining covers, a $35,160 occupancy increase may eliminate remaining seller discretionary earnings after a chef replacement. At that point the going-concern premise is a turnaround budget: new rent, new labor, a capital list, and a period of negative or break-even cash flow. The price is then a function of how much cash the buyer must inject to stabilize, not of last year’s tax return.
Write a conclusion that a buyer can use without re-doing your work
State the diagnosis, the periods weighted, the maintainable cash-flow range, the known shocks included, the asset floor, and the premise of value. If you used a market indication, say why the guideline companies were declining or stable and why that is comparable. Reported listing multiples from marketplace commentary are context; they are not a company-specific conclusion and they are not a cure for a 22 percent two-year decline. Keep the valuation date visible. Later months can corroborate or contradict the trend, but they should be labeled as subsequent information.
Owners still have operating moves: cut a losing daypart, renegotiate the reset, restore food quality, or sell now to a nearby operator who can absorb the lease. Those moves change the file. They do not change the rule against averaging a dead peak into a live price. The honest range is almost always wider on a declining company, and that is a feature. A point estimate that pretends the collapse was noise is the number that dies in diligence.
Diagnose the decline before you weight the years
A percentage drop is a symptom. The valuation needs the driver, the likely persistence, and the cash the buyer will actually inherit.
| Issue | What the owner should assemble | What a buyer is likely to test | Why it changes the decision |
|---|---|---|---|
| Seasonal or calendar effects versus trend | Three or more years of monthly revenue, covers or orders, and a note of known calendar events. | Compare like months and like day-counts instead of annual totals, consistent with the idea behind Census seasonal adjustment: separate recurring calendar effects from underlying movement. | A seasonal dip is not a collapse. A two-year slide that survives seasonal comparison is a trend. |
| Cyclical demand | Local activity measures, customer-industry exposure, bid logs, and prior-cycle financials. | Ask whether a mid-cycle operator would still see this volume without a new concept or channel. | A true cycle can support a cycle-average with a higher risk rate. A one-way slide cannot. |
| Structural market change | Traffic sources, regulatory file, platform rules, neighborhood occupancy, and competitor openings or closings. | Identify the demand that is gone for good: a closed office tower, a solvent ban, a marketplace algorithm, a lost wholesale account that will not return. | Structural files move toward turnaround pricing or an asset floor, not toward a peak-year multiple. |
| Owner-caused neglect | Hours, reviews, pricing versus vendors, staffing, and maintenance logs compared with the years that worked. | Separate demand that is still present from quality, hours, or cleanliness the owner let slip. | Upside may exist, but it usually belongs to the buyer who will fund the repair, not to a seller multiple on the old peak. |
| Asset floor and required catch-up capital | Equipment condition, lease remaining term, inventory quality, and a tangible-property split between repairs and betterments. | Price an orderly sale of FF&E, inventory, and transferable rights, then add the cash needed to stabilize. | If going-concern value cannot beat that math, stop using historical earnings as the headline. |
Worked example: a restaurant down 22 percent with a rent reset in nine months
This example is hypothetical. A full-service restaurant did $1,094,000 of revenue and $218,600 of SDE two years ago. Last year it did $958,000 and $161,400 of SDE. The trailing twelve months are $853,000 of revenue, 22.0 percent below the peak year, and $117,800 of SDE. Covers moved from 88 a day to 73. Food cost moved from 31.8 percent to 37.1 percent. Current rent is $7,420 a month. In nine months it resets to $10,350. The owner wants a three-year average SDE of about $165,900 as the earnings base.
| Case | Earnings or asset figure | What it includes | Why it is or is not maintainable |
|---|---|---|---|
| Peak year SDE | $218,600 | 88 covers, 31.8 percent food cost, old rent | Not the valuation date capacity |
| Three-year average SDE | About $165,900 | Blends the peak with the collapse | Disguise, unless a proven cycle exists; none is shown |
| Trailing SDE | $117,800 | 73 covers, 37.1 percent food cost, old rent still in place | Closer, but still too high because the reset has not hit yet |
| Rent reset shock | $35,160 more occupancy a year | Scheduled step from $7,420 to $10,350 | Must sit in the base case, not a footnote |
| Going-concern after reset and a chef-owner replacement | Near break-even on current covers | New rent plus a market kitchen wage | Not a multiple-bearing earnings stream without a turnaround plan |
| Orderly asset floor | $51,000 FF&E plus $8,700 food inventory | Used-market kitchen and dining package; leasehold book of $64,000 is not assumed transferable | Floor if the going concern cannot be believed |
The three-year average is the most dangerous number in the file because it looks professional. It is still a blend of a dead peak and a live decline. Revenue Ruling 59-60 wants earning capacity as of the valuation date. Capacity here is a 73-cover restaurant with a damaged food-cost line and a $35,160 occupancy increase that has not yet appeared on a tax return. Weighting the peak year is advocacy.
Once the reset and a market wage for the chef-owner are inserted, the going-concern case is a turnaround budget, not a historical capitalization. A buyer might still pay above the $51,000 equipment floor for a transferable liquor right, a remaining lease with a real landlord conversation, or a location that a nearby operator can absorb. That extra amount has to be justified with a stabilization plan and cash to fund it. It cannot be justified with $218,600 of memory.
Tangible-property discipline helps the capital list: hood work and a replacement range may be restorations or betterments a buyer must capitalize, not items to add back as “one-time.” Marketplace listing commentary about restaurant multiples describes other companies. It does not recast this rent clause. Use it as context, if at all, and then return to the covers.
A six-step declining-revenue workup
The deliverable is a diagnosis and two numbers: a going-concern range that already contains the shocks, and an asset floor.
- 01
Build a monthly driver stack
For each of at least 36 months, show units or covers, average price, mix, and the cost lines that should move with volume. Do not start with annual totals.
Deliverable: Monthly driver workbook
- 02
Label seasonal, cyclical, structural, and owner-caused pieces
Write a short memo that assigns the decline. If more than one cause is present, estimate the share. Refuse an unlabeled “down year.”
Deliverable: Cause memo with evidence citations
- 03
Re-cast every year on the same policy
Apply one normalization policy so the trend is visible. Do not add back neglect, deferred maintenance, or a wage the buyer must pay.
Deliverable: Consistent multi-year earnings bridge
- 04
Insert known shocks in the base case
Rent resets, wage steps, insurance, lost contracts, and required capital go into maintainable cash flow now. Show the month they hit.
Deliverable: Shock schedule with dollar effects
- 05
Compute the asset floor
Bid used equipment, remaining inventory at net realizable value, and any assignable rights. Treat unassignable leasehold improvements as a warning, not an asset.
Deliverable: Orderly asset-floor exhibit
- 06
Write the two-premise conclusion
State going-concern value after shocks, asset-floor value, and which premise is more credible on the valuation date. If they converge, say the goodwill is thin.
Deliverable: Two-premise conclusion page
Where otherwise credible analyses break down
Averaging the peak because “buyers always use three years”
Why it matters: A custom is not a method. Mixing a collapse into an average overstates capacity and sets up a diligence failure.
Better approach: Weight the periods that represent current capacity. Use a cycle average only with evidence of a cycle.
Calling every drop seasonal
Why it matters: Seasonal patterns repeat. A two-year slide in like months is a trend, and Census seasonal-adjustment logic exists to keep people from confusing the two.
Better approach: Compare like periods, then talk about seasonality only for the residual.
Adding back the costs of the decline
Why it matters: Emergency discounts, extra promotions, and catch-up repairs are often the new normal, not one-time noise.
Better approach: Treat stabilization spending as part of the turnaround budget a buyer must fund.
Ignoring the asset floor because it feels pessimistic
Why it matters: A going-concern multiple on empty covers can be lower than used equipment plus inventory. Refusing to look leaves money on the table or invents goodwill that is not there.
Better approach: Always show both premises on a declining file.
What a defensible owner decision looks like
I will not average a restaurant’s best year with its collapse and call the result independent. If covers are down, food cost is up, and a rent reset is on the calendar, the valuation date has already told us what earning capacity is. The peak year is research. It is not the base.
The same discipline applies to a dry cleaner losing piece counts and to an ecommerce brand watching contribution margin die in the ads account. Name the cause. Put the shocks in the base case. Print the asset floor. If a nearby operator can still use the lease or the platform, we can talk about a price above scrap. We cannot talk about it as if the last good tax return were still the business.
SBA close-or-sell planning includes winding down. That is not a threat I use for leverage. It is a real alternative when the going concern is a project the current owner already failed. Owners who face that early still have time to cut a daypart, renegotiate, or sell to someone who can absorb the overhead. Owners who hide inside a three-year average donate that time to diligence.
Questions owners ask
Should I use a three-year average if last year was down?
Only if the down year is a normal point in a documented cycle. If the decline is structural or owner-caused, an average of the peak and the collapse overstates maintainable earnings.
When does a declining business become an asset-sale file?
When cash flow after replacement costs and known shocks cannot support a buyer’s capital stack, or when the going-concern indication falls to or below an orderly asset floor.
Does a national listing multiple still apply to a shrinking company?
Not as a plug. Marketplace ranges describe reported sold companies as a group. A company with a documented two-year decline needs a company-specific earnings base and risk assessment, not a healthy-company multiple on a peak year.
Can a buyer still use a market multiple on a declining company?
Only if the guideline transactions are actually similar in trajectory and the earnings base is the current maintainable number, not the peak. A healthy-company multiple on a shrinking run rate is a mismatch.
What if revenue has declined but cash flow has not?
Then the driver stack should show why: mix, pricing, or cost cuts. Durable margin with lower volume can still be a going concern. Margin that is being propped up by skipped maintenance or unpaid owners is not.
Does closing the business have tax consequences that should affect the floor?
Asset sales, cancellations, and wind-downs have federal filing consequences described in IRS closing-a-business guidance and related publications. Those are after-tax proceeds questions. They do not justify keeping a going-concern multiple on empty volume.
Sources and review date
Last reviewed: September 3, 2026. Sources are linked for context; a national benchmark is not a substitute for local comparable sales or a purpose-specific appraisal.
- IRS valuation job aid and Revenue Ruling 59-60 — Appendix A reproduces Revenue Ruling 59-60 and its closely held business valuation factors; the job aid itself states that it is not legal authority.
- U.S. Census Bureau: Quarterly Services seasonal-adjustment FAQs — Explains why recurring calendar effects must be identified and separated before interpreting changes in an economic time series.
- U.S. Small Business Administration: Close or sell your business — Current owner guidance on sale planning, valuation approaches, sale agreements, transfer choices, professional advice, and maintaining required records.
- IRS: Tangible property final regulations — Explains the federal tax framework for distinguishing supplies, routine repairs, maintenance, betterments, restorations, and capital improvements.
- BizBuySell industry valuation benchmarks — Reported Main Street sold-business data. A national category range is context, not a company-specific conclusion.
- IRS: Closing a business — Identifies federal filing considerations when a business closes or its assets are sold.
- International Business Brokers Association glossary — Professional definitions for SDE, transaction terms, and Main Street business brokerage concepts.