Khaled Azar 
Your business pays you $1 million a year after expenses. A buyer offers $6 million. Six years of income in one wire, an easy yes.
Except the wire is not $6 million. After federal and state taxes it can be closer to $3.8 million. And your $1 million a year was never $1 million either: after taxes you keep about $580,000 of it. So the real comparison is $3.8 million once, versus $580,000 for every year you keep the keys. That is six and a half years of take-home income, not six years of profit, and one tax provision can stretch it to nine.
That comparison, both sides after tax and risk-adjusted, is the sell vs hold analysis most founders never actually run. In 2026, a California founder without tax protection can lose more than a third of their gain: 20% federal, 3.8% net investment income tax, and up to 13.3% state. Here is how to run the numbers properly.
The Terms That Decide This
Five terms carry this whole decision. Plain English first, math after.
| Term | What it is | Why it matters here |
|---|---|---|
| Net check | The cash that actually lands in your account: price minus debt, adjustments, held-back money, fees, and taxes | This is the sale side of the comparison, not the headline price |
| Deferred revenue | Money customers prepaid for service you have not delivered yet, like the unused months of annual subscriptions | Buyers inherit that obligation, so it often reduces the price at close |
| Working capital target | The amount of day-to-day operating cash the buyer requires to be left in the business at close | Come in under the target and the difference comes out of your check |
| Escrow | A slice of the price held back, usually 12 to 18 months, as insurance against surprises | Part of your money arrives late, and only if nothing goes wrong |
| QSBS | Qualified Small Business Stock, a federal rule that can exempt some or all of your gain from federal tax | The single biggest swing factor in the sale-side math |
The Apples-to-Oranges Error in Sell vs Hold Analysis
The mistake is not bad math. It is comparing two numbers that live in different worlds.
The sale side starts with the headline price, but you do not deposit the headline price. Buyers deduct debt. Deferred revenue often gets negotiated out of the price at close, the exact treatment varying by deal. The working capital target adjusts the check again. Escrow delays another slice. Only what survives all of that gets taxed, and only what survives the taxes is your net check.
The hold side has its own distortion. Founders quote sale multiples pre-tax but quote their own distributions post-tax, because that is the number they see land in their account. The result is a comparison that flatters whichever option the founder already prefers.
Run both sides to the same finish line: cash in your account after taxes. Anything else is a comparison between a brochure and a bank statement.
The Sale Side: What the Check Actually Nets in 2026
Start with federal long-term capital gains. For 2026, the IRS set the 20% bracket to begin above $545,500 of taxable income for single filers and $613,700 for joint filers, per Tax Foundation’s 2026 bracket tables. On a multi-million dollar gain, nearly all of it is taxed at 20%. The 3.8% net investment income tax stacks on top, and its $200,000 threshold has never been indexed for inflation, so it catches most founder exits structured as stock sales.
Then comes your state. California taxes capital gains as ordinary income at rates up to 13.3%, with no discount for holding period. Texas and Florida take zero. The same $6 million gain nets roughly $800,000 more in Austin than in Los Angeles: state residence is a valuation lever most founders never price.
Taxes are only the last bite. The deal mechanics from the terms table each take theirs first. Here is the same $6 million offer run through the full waterfall, with illustrative round numbers for a debt-free, single-filer California seller:
| Waterfall step | Amount | Running total |
|---|---|---|
| Headline price | $6,000,000 | $6,000,000 |
| Deferred revenue adjustment | -$120,000 | $5,880,000 |
| Working capital true-up | -$80,000 | $5,800,000 |
| Advisor and legal fees | -$320,000 | $5,480,000 |
| Taxes on the gain (about 36% combined) | -$1,970,000 | $3,510,000 |
| Escrow, held about 15 months | -$580,000 held back | $2,930,000 wired at close |
The wire at close is about $2.9 million. The escrow arrives around 15 months later if nothing goes wrong, bringing the total to roughly $3.5 million. Notice that the $3.8 million from the opening assumed taxes were the only cost. Run every lever and the no-QSBS payback drops toward six years of net distributions, which strengthens the hold case further. This is why the waterfall, not the headline, is the number that belongs in your decision.
The counterweight is Qualified Small Business Stock. The 2025 tax law expanded Section 1202 in ways that change exit timing math directly.
For stock issued after July 4, 2025, the per-company QSBS cap rose from $10 million to $15 million, and a new tiered schedule excludes 50% of gain at a 3 year hold, 75% at 4 years, and 100% at 5 years.
One nuance buyers’ accountants know and founders often do not: California does not conform to Section 1202. A fully QSBS-excluded federal gain is still taxed in full by Sacramento. Whether your stock qualifies at all depends on entity history and deal structure, which is exactly the ground covered in our guide to QSBS, asset versus stock treatment, and the tax planning founders should do before a sale.
The Hold Side: Your Distributions Are Taxed Worse Than Your Gain
Holding means your company keeps paying you, and those dollars are taxed as ordinary income. The top federal rate is 37%, against 23.8% for capital gains. Pass-through owners get relief from the qualified business income deduction, which the 2025 law made permanent: it can cut the effective federal rate on qualified profits to roughly 29.6%. Add your state and a California operator keeps somewhere near 55 to 60 cents of each distributed dollar.
So the hold scenario earns its money on repetition, not rate. The question is how many years of net distributions it takes to match the net check, and what can go wrong while you collect them.
Here is the illustration for a single-filer California founder with a $6 million all-cash offer on a business distributing $1 million a year. To keep the illustration clean, assume the $6 million is already the price after debt, deferred revenue, and working capital adjustments, with zero tax basis and escrow paying out in full. Real deals move each of those levers, always run your own waterfall. These are rounded, directional numbers, not tax advice.
| Scenario | Gross | Taxes | Net |
|---|---|---|---|
| Sell, no QSBS | $6.0M gain | ~$2.2M (20% + 3.8% + ~13%) | ~$3.8M once |
| Sell, 100% QSBS | $6.0M gain | ~$0.8M (state only) | ~$5.2M once |
| Hold | $1.0M per year | ~$0.42M per year | ~$0.58M per year |
Without QSBS, the sale equals about 6.5 years of net distributions. With full QSBS it equals about 9. That single variable moves the answer by two and a half years, which is why the tax structure question belongs at the start of a sell vs hold business analysis, not the end.
Risk-Adjusting the Hold: The Payback-Years Test
Six years of distributions only match the check if the business still looks like this in year six.
I call the framework the Payback-Years Test. Divide the net check by net annual distributions to get raw payback years. Then ask one question: what is the probability this business distributes at today’s level for that many more years, with you still willing to run it?
The hold scenario carries every risk you already own. One large customer leaving can cut distributions and the resale value in the same quarter. Churn creep, a platform shift, your own burnout: the hold case is a concentrated, undiversified bet that compounds those risks annually, while the sale converts them to cash at today’s multiple. Waiting has a price even when the business holds steady, because multiples move and buyers reprice momentum.
A reasonable risk adjustment discounts hold-side cash meaningfully. If raw payback is 6.5 years and you weight each successive year at 85% odds that distributions hold, compounding annually, the risk-adjusted payback stretches past 8 years. Holding does preserve one asset the simple version ignores: the option to sell later. So add a realistic terminal value at the end of your horizon, priced at a conservative multiple, rather than assuming the business is worth zero in year six. When risk-adjusted payback still runs well past your honest planning horizon, sell. When it lands inside it and the business is genuinely durable and transferable without you, holding is not sentimental. It is just the better trade.
The Hold Model Is Also Your Negotiating Position
A pattern from real deals, kept general: a founder models the hold case properly, discovers the offer on the table nets less than a few years of keeping the keys, and stops negotiating like someone who needs the deal. The counter gets firmer. Deferred revenue treatment, escrow size, and earn-out terms all move, because the buyer is no longer bidding against hope. They are bidding against a documented alternative.
That only works when the hold model is credible: post-tax, risk-adjusted, and built on the same numbers a buyer will diligence. It works even better inside a competitive process with more than one bidder, where the hold case sets your floor and the market sets the ceiling.
Build the hold model even if you intend to sell. It is the cheapest negotiating asset in the entire process.
Frequently Asked Questions
How do I compare selling my business to holding it?
Take both sides to post-tax cash. Compute the net check: price minus debt, deferred revenue and working capital adjustments, escrow, fees, and taxes. Divide it by your net annual distributions to get payback years, then risk-adjust those years for concentration, market, and personal factors.
What taxes apply when a founder sells a SaaS company in 2026?
Federal long-term capital gains of 20% above $545,500 of taxable income for single filers, plus the 3.8% net investment income tax, plus state tax. California taxes the full gain as ordinary income at up to 13.3%, while states like Texas and Florida charge nothing.
Does QSBS still apply if I sell before five years?
For stock issued after July 4, 2025, yes: the 2025 tax law added a 50% exclusion at a 3 year hold and 75% at 4 years, with 100% at 5 years and a $15 million per-company cap. Stock issued on or before that date keeps the old all-or-nothing 5 year rule.
How many years of profit should a sale price equal?
There is no universal number: the useful test is post-tax payback years, not a headline multiple. Compute how many years of net distributions the net check replaces, risk-adjust those years, and compare the result to how long you honestly want to keep running the business.
Next Steps
Before you answer an offer, know your real numbers on both sides: the net check a sale would put in your account and the risk-adjusted value of keeping the keys.