If you've spent any time researching startup equity, you've seen the ISO vs NSO comparison. ISO good (lower tax!), NSO simpler (no AMT!). Most articles end there. Most articles also assume you can sell your shares within a year of exercise. At a private company, that assumption breaks down — and the tradeoff changes.
The textbook comparison
- ISO at exercise: no immediate income tax, but the spread (FMV − strike) is an AMT preference item.
- NSO at exercise: spread is ordinary income, immediately, with payroll withholding.
- ISO at sale (>1yr after exercise, >2yr after grant): entire gain at long-term capital gains rates (20% top federal + 3.8% NIIT).
- NSO at sale: spread at exercise was already ordinary income. Any subsequent gain is capital gains.
The illiquidity twist
At a public company, you can exercise an ISO and sell the shares the same day. AMT becomes negligible because you have cash from the sale to cover it. The ISO advantage is fully realised if you can hold for >1 year.
At a private company, you exercise — and then you sit. You may not be able to sell for 3, 5, even 10 years. Meanwhile, your AMT liability is real and due that tax year. People have gone bankrupt over private-company ISO exercises that triggered AMT they couldn't pay.
When ISO still wins
ISO wins when (a) you have cash to cover AMT without selling shares, (b) you expect a liquidity event within a reasonable timeframe, and (c) you can wait long enough post-exercise to clear the holding periods.
When NSO is actually better
NSO is better when you'd otherwise face large AMT and no way to pay it. The ordinary income at NSO exercise is also less than total ISO ordinary-rate equivalence in some scenarios — particularly when you're certain you'll sell within a year.
Rule of thumb: if you have ISOs and the spread × your vested shares is more than 25% of your liquid net worth, the AMT trap is real. Talk to a CPA before exercising.