Compliance
409A Safe Harbor: What Actually Earns It, and Why the Price Tag Doesn't
The 409A safe harbor is a conditions test, not a price tier. What the regulation actually checks, where cheap reports fail, and who pays when one does.
By 409.AI Team - 2026-08-19
Put three 409A quotes side by side and you'll typically see something like $500, $2,400, and $8,000. The next question is always the same one: which of these is actually safe harbor compliant?
None of them, and all of them. Price is not a variable in the regulation. You can read Treasury Regulation §1.409A-1(b)(5)(iv) end to end and never find a dollar figure, a fee floor, or any suggestion that a more expensive report earns more protection. What you find instead is a short list of conditions that are either met or not met, and a rebuttable presumption that follows from meeting them.
That distinction matters more than most founders realize, because the thing you are buying is not a PDF. It's a shift in who has to prove what if the IRS ever asks.
What the safe harbor actually gives you
Start with the default rule, because the safe harbor only makes sense against it.
For stock that isn't readily tradable on an established securities market, fair market value means "a value determined by the reasonable application of a reasonable valuation method," judged on the facts and circumstances as of the valuation date. The regulation lists what a reasonable method takes into account: tangible and intangible assets, the present value of anticipated future cash flows, market values of comparable companies, recent arm's length transactions in the stock, and other relevant factors including control premiums and [discounts for lack of marketability](https://409.ai/articles/discount-lack-marketability-dlom-409a-valuation).
Under that default, if the IRS challenges your strike price, you carry the burden. You have to show the method was reasonable and applied reasonably.
The safe harbor flips that. Use one of three specified methods and your valuation is "presumed to result in a reasonable valuation," and the Commissioner can only rebut the presumption "upon a showing that either the valuation method or the application of such method was grossly unreasonable" ([26 CFR §1.409A-1(b)(5)(iv)(B)(2)](https://www.govinfo.gov/content/pkg/CFR-2025-title26-vol6/pdf/CFR-2025-title26-vol6-sec1-409A-1.pdf)).
Grossly unreasonable is a far harder standard for an examiner to meet than merely unreasonable. That gap, not the report itself, is what you're paying for. Our [guide to what a 409A valuation is](https://409.ai/articles/what-is-a-409a-valuation-a-comprehensive-guide) covers the groundwork this piece assumes.
Three doors, and what each one checks
The regulation names three qualifying methods. Each comes with conditions you can check off before you sign anything. None of them mentions cost.
The independent appraisal
The most common route: a valuation of a class of stock determined by an independent appraisal meeting the requirements of section 401(a)(28)(C), as of a date no more than 12 months before the transaction it's applied to, such as an option grant date.
Follow that citation and you get somewhere useful. [IRC §401(a)(28)(C)](https://www.govinfo.gov/content/pkg/USCODE-2024-title26/html/USCODE-2024-title26-subtitleA-chap1-subchapD-partI-subpartA-sec401.htm), the ESOP appraisal rule, defines an independent appraiser as one "meeting requirements similar to the requirements of the regulations prescribed under section 170(a)(1)." Those regulations, at [26 CFR §1.170A-17](https://www.govinfo.gov/content/pkg/CFR-2025-title26-vol4/pdf/CFR-2025-title26-vol4-sec1-170A-17.pdf), define a qualified appraiser as an individual with verifiable education and experience in valuing that type of property: either relevant professional or college-level coursework plus two or more years of experience valuing that property type, or a recognized appraiser designation from a professional appraiser organization.
Verifiable education. Documented experience. A designation. That's the test, and it attaches to a person, not to an invoice.
The formula method
A valuation based on a formula that would be treated as fair market value under the §83 rules if used as part of a nonlapse restriction, provided the same formula is used consistently for every transfer of that class of stock to the issuer or to any 10 percent owner, other than an arm's length sale of substantially all the company's stock.
This one is rare in venture-backed companies, and the consistency requirement is why. A formula you use for the option plan but abandon the moment a buyer offers a real number is not a formula you can rely on here.
The illiquid start-up valuation
The path most often misdescribed. It permits a valuation "made reasonably and in good faith and evidenced by a written report" of illiquid stock of a start-up corporation, and the regulation defines that term with some precision.
The first condition: the corporation has no material trade or business that it or any predecessor has conducted for a period of 10 years or more. Read that carefully, because it is not the same as "the company is less than 10 years old," which is how you'll usually see it summarized. On the text of the condition, the clock runs on the business, not on the incorporation date, so a twelve-year-old corporation that pivoted five years ago into a genuinely different business is in a different position from a three-year-old company built on a predecessor's decade-old operating business. If your history has a pivot, a rollup, or a predecessor entity in it, that is a question for your counsel and your appraiser rather than a box to tick.
The remaining conditions are more mechanical: no class of equity securities traded on an established securities market; the stock isn't subject to any put, call, or similar purchase right or obligation, other than a right of first refusal on a third-party offer or a lapse restriction; and neither the company nor the service provider reasonably anticipates a change in control event within 90 days of the action the valuation is applied to, or a public offering within 180 days of it.
The written report must be prepared by a person the corporation reasonably determines is qualified based on significant knowledge, experience, education, or training. The regulation gives a benchmark: significant experience generally means at least five years in business valuation or appraisal, financial accounting, investment banking, private equity, secured lending, or comparable experience in the company's line of business.
Note what this path does not require. It doesn't require independence, and it doesn't require you to pay anyone. A qualified CFO can write it. The conditions are strict enough that most companies with a term sheet in hand or an exit in sight fall out of eligibility, which is usually why they buy the appraisal instead.
The only place the rules talk about fees
There is exactly one fee rule in this chain, and it runs in the opposite direction from the one founders assume.
Section 1.170A-17(a)(9), "Prohibited appraisal fees," states that the fee for a qualified appraisal "cannot be based to any extent on the appraised value of the property," and treats a fee as based on appraised value if any part of it depends on the value the IRS allows after examination.
The concern the regulations encode is that the appraiser's pay might be tied to the answer. That is a conflict problem, not a budget problem. Nothing anywhere sets a minimum. A $500 report and a $15,000 report from appraisers with identical qualifications sit in exactly the same position under §1.409A-1(b)(5)(iv)(B)(2), on the day they're issued.
So where do cheap reports actually go wrong?
They go wrong on the substance conditions, which is a different failure than the one people shop against.
Two sentences in the general rule do most of the damage. First: a valuation method is not reasonable "if such valuation method does not take into consideration in applying its methodology all available information material to the value of the corporation." Second: a previously calculated value is not reasonable as of a later date if the calculation fails to reflect information available afterward that may materially affect value, or if it was calculated as of a date more than 12 months earlier.
That's where a thin engagement fails. Not because it was cheap, but because nobody asked about the acquisition conversation that started in March, or the term sheet on the table, or the [secondary sale that put a real price on your common stock](https://409.ai/articles/tender-offers-secondary-sales-409a-valuation). A report built from a cap table upload and last year's financials, with no one asking questions, can miss information that was sitting in the founder's inbox.
The same applies to the calendar. A 409A doesn't quietly expire at month 13 so much as it stops being a reasonable basis for a grant, and any material event resets that clock early. Our piece on [how often you need a new 409A](https://409.ai/articles/409a-valuation-frequency-how-often-should-you-get-one) walks through the triggers.
The methodology has to hold up too. Choosing between [an OPM, a PWERM, or a backsolve](https://409.ai/articles/409a-allocation-methods-opm-pwerm-backsolve) and defending that choice is judgment that either shows up in a workfile or doesn't, and our breakdown of [what the IRS asks for in a 409A exam](https://409.ai/articles/irs-audit-409a-valuation-document-request) shows how quickly that gets tested.
What a failure costs, and who pays it
Say you grant an engineer 60,000 options at a $0.30 strike based on a report that missed a material event, and the defensible fair market value on the grant date was $1.10. Those options were granted at a discount, so they lose the exemption that properly priced options get and become deferred compensation subject to §409A. The bill then arrives on vesting rather than on exercise: the deferred amount is includible in income once it stops being subject to a substantial risk of forfeiture, whether or not she has exercised anything or sold a single share.
Per the IRS [Nonqualified Deferred Compensation Audit Technique Guide](https://www.irs.gov/pub/irs-pdf/p5528.pdf), amounts includible under §409A are reported separately on Form W-2 in box 12 with code Z, and are subject to an additional 20 percent income tax plus a second tax based on an imputed interest underpayment, the premium interest tax. If the spread across that grant is $48,000 by the time it has fully vested, the 20 percent additional tax alone runs to $9,600, on top of ordinary income tax and before any premium interest.
The guide is explicit about who absorbs it: those additional taxes "are assessed against the employee/service provider and not the employer/service recipient." The company saved a few thousand dollars on the report. The engineer pays the penalty on a grant she had no way to price-check. That asymmetry is the real argument against shopping this on cost alone. If you suspect a past report has this problem, [correcting an incorrect 409A](https://409.ai/articles/dealing-with-incorrect-409a-valuations) is solvable, and far cheaper to solve before an examiner finds it.
Five questions that actually separate providers
Since price tells you nothing, ask about the conditions instead.
Who signs the report, and what are their credentials and years of valuation experience? The safe harbor attaches to that person.
Is the fee fixed, or does any part of it depend on the value produced? Contingent structures are the one fee arrangement the regulations single out.
What did you ask me about the last twelve months? A provider that never asked about pending term sheets, acquisition approaches, secondaries, or major customer losses cannot claim to have considered all available material information.
Which allocation method did you use and why? "The model picked it" is not a defense of a methodology choice.
What's in the workfile if this gets examined? Get the answer before you need it.
The takeaway
The safe harbor is a conditions test, and every one of its conditions is checkable before you sign an engagement letter. Independence and qualification, or company eligibility plus a qualified author, plus a method applied to all the information that was actually available. Cost isn't on the list, and no invoice amount adds a condition or waives one.
Where price does matter is downstream of the presumption: it tends to correlate with how many questions get asked, how the methodology gets defended, and what exists in the workfile if the presumption is ever challenged. Buy that work, not the line item. When you compare [409A valuation providers](https://409.ai/products/409a), put the credentials of the signer and the depth of the intake at the top of the comparison, and treat the number at the bottom of the quote as what it is: the least informative thing on the page.