The defining feature of incentive stock options (ISOs) is what doesn’t happen when you exercise them: no regular taxable income, no withholding, nothing on the W-2. Exercise 10,000 ISOs at a $10 strike while the shares are worth $110, and for regular tax purposes the $1,000,000 of built-in gain simply isn’t income yet.
The catch is a parallel tax system called the Alternative Minimum Tax, and it’s the reason ISO exercises often produce surprise tax bills. This article covers how the AMT interacts with ISO exercises, and how modeling an exercise before making it can shrink that bill, often to zero.
How the AMT sees an ISO exercise
The AMT is a second tax calculation that runs alongside the regular one. It starts from your regular income, adds back certain items (called preference items and adjustments), applies its own exemption and its own rates, and produces a “tentative minimum tax.” Each year you effectively pay the higher of the two calculations. Most people never notice the AMT because their tentative minimum tax comes out lower than their regular tax, so the regular tax wins.
An ISO exercise changes that. The spread between the strike price and the fair market value at exercise (the “bargain element”) is invisible to the regular tax but counts as income for the AMT. In the running example:
- The bargain element: $110 value minus $10 strike is $100 per share. Exercise all 10,000 options and that’s $1,000,000 of AMT income in a single year, with no cash received and no shares sold.
- The resulting bill: AMT rates run 26% to 28% on income above the exemption. A full exercise of this position could produce a tax bill in the low-to-mid six figures, due the following April, funded entirely out of pocket.
- A moving target: Tax law changes effective in 2026 lowered the income thresholds where the AMT exemption phases out and made the phaseout steeper. Exercises that stayed under the radar in prior years can trigger AMT under the current rules, which makes this worth calculating rather than eyeballing.
The AMT is usually a timing cost, not a permanent one
Before getting to the planning, one piece of context that changes how the whole problem feels: AMT paid on an ISO exercise generally comes back.
- The AMT credit: AMT attributable to timing items (which an ISO exercise is) generates a credit that can offset regular tax in future years, subject to limits. The money isn’t gone; it’s prepaid.
- The basis adjustment: Shares exercised through AMT get a higher basis for AMT purposes, which reduces the AMT gain when the shares are eventually sold.
- The timing catch: Recovery can take years, and the credit only helps in years when regular tax exceeds tentative minimum tax. Paying six figures of AMT is a significant cash-flow event even if it’s not a permanent cost. And if the stock later collapses, the AMT was paid on value that never materialized (the credit still exists, but may not be unlocked without income to release it).
So the AMT question is less “how do I avoid this tax forever” and more “how much am I willing to prepay, and for how long.”
The breakeven exercise
Here’s where the planning gets concrete. Because you pay the higher of regular tax and tentative minimum tax, there’s room between the two in most years: regular tax typically exceeds tentative minimum tax by some margin. An ISO exercise eats into that margin dollar by dollar. The number of options that can be exercised before the margin runs out is the AMT breakeven, and exercises up to that point incur no AMT at all.
- The modeling: Project the full year’s income (salary, bonus, RSU vesting, investment income, deductions), calculate both tax systems, and solve for the bargain element that makes them meet. That’s the year’s free exercise capacity. In the running example, the answer might be 900 options one year and 2,500 the next, depending entirely on what else is happening on the return.
- Multi-year spreading: Repeating a breakeven exercise every year moves the position steadily into shares (each starting its own long-term holding period) without ever writing a check to the IRS for AMT. The tradeoff is time: a large grant can take years to work through, and the 409A may keep climbing while you wait, shrinking each year’s capacity.
- Expanding the capacity: Counterintuitively, more ordinary income means more room for AMT-free exercises, since regular tax rises faster than tentative minimum tax. Years with a big bonus, an NSO exercise, or other income spikes are often the best years for ISO exercises. Some plans go further and pair exercises with same-year sales of a portion of the shares (a disqualifying disposition, covered in Part 3), which converts some spread into ordinary income and can fund and enlarge further exercises at the same time.
The other direction: exercising through the AMT on purpose
The breakeven approach minimizes tax outlay, but it isn’t automatically the right plan. When a company looks headed for a rapid run-up (say, IPO preparation), there’s a case for the opposite: exercise a large block now, pay the AMT, and start the one-year clock toward long-term capital gains treatment on the whole position while the spread is still relatively small.
- The tradeoff: Prepay a known AMT bill today in exchange for long-term rates on a potentially much larger gain later, with the AMT itself gradually recovered as a credit.
- What it requires: Comfort fronting the cash for an extended period, and conviction in the company. This is the version of the plan where the investment judgment from Part 1 does the heavy lifting; the tax math only makes sense if the shares are worth owning at this price.
What careful planning looks like here
The pattern across all of this: the AMT consequences of an ISO exercise are knowable in advance. A projection of the full year’s income, run before exercising rather than after, turns the AMT from a surprise into a dial. Modeled well, an exercise can be sized to incur exactly $0 of AMT, or deliberately sized past that point with the bill quantified and the credit recovery mapped out.
The projection itself has real moving parts (exemption phaseouts, state treatment, interaction with other income events, the credit mechanics in later years), and the breakeven shifts every year. This is the kind of analysis where a tax advisor who models equity compensation regularly earns their fee: the difference between a guessed exercise and a calculated one is often the entire tax bill.
This article is for educational purposes only and is not tax, legal, or investment advice. The scenarios and numbers are illustrative and don't describe any actual client. Tax outcomes depend heavily on individual circumstances; consult a qualified tax professional about your specific situation.