Buried in the benefits packet, after the 401(k) match and the ESPP, sits a plan most employees scroll past: nonqualified deferred compensation. It lets you push salary and bonus into the future - before it ever hits your W-2 - invest it, and take it back later. No IRS contribution limit. An executive can defer $250k or more a year, sidestepping the extreme top marginal rates that make every extra dollar earned feel half-optional. It’s the last legal way to make a big chunk of income disappear from this year’s tax return.
The math. Defer $100k at a 50% combined marginal rate and you’re investing the whole $100k instead of the $50k that would survive income tax today. The benefit comes from avoiding annual taxable-account drag while the money compounds, and potentially paying a lower rate when it comes out. If the entry and exit rates are the same and the taxable alternative has no annual tax drag, the after-tax outcomes are equal at equal returns. One thing this isn’t: a payroll-tax deferral to payout. FICA generally applies at the later of service or vesting. The deferral is an income-tax play.
The state exit. Here’s the play nobody talks about: federal law blocks states from taxing retirement income of nonresidents when it’s paid in substantially equal installments over ten years or more. Defer while you’re earning in California, structure the payout as a 10-year installment, move to Nevada or Texas, and California’s 13.3% never lands. That’s not a loophole. That’s the statute.
The catch, and its real size. Your deferred comp is technically an IOU - you’re an unsecured creditor of your employer. But size it honestly: this is the small catch, the black swan. Enron and Lehman made the headlines precisely because it’s rare, and a rabbi trust covers the change-of-heart case. Worth knowing, not worth skipping the perk over.
The election trap. You choose your distribution schedule before you earn the money, and 409A makes changing it painful: re-elections push payments out five years, early withdrawals are plan-specific, and a 409A failure can trigger a 20% federal tax plus interest; key employees at public companies wait six months after separation. The flexibility is an illusion. Plan the exit at the entrance.
Who it’s for. This can be valuable if you’re in a high tax bracket today and have high confidence your bracket will be the same or lower when the money is paid out. A lower payout rate adds to the benefit; at the same rate, the advantage depends on avoiding the taxable alternative’s annual tax drag. Max the 401(k) and the mega backdoor first where available. Then weigh the deferral against your employer’s credit risk and the payout schedule you can actually commit to.
Assumptions
The perk is tax timing, and sometimes a lower future rate. It is not free money without employer, plan and tax conditions.
- Contribution examples: $250,000 is an illustrative annual deferral, not an entitlement in every plan. $100,000 x (1 - 50%) = $50,000 of current after-income-tax cash. The example ignores separate payroll taxes and assumes the plan permits the election.
- Compounding: With equal returns and the same 50% tax on exit, $100,000 x (1 + r)^n x 50% equals $50,000 x (1 + r)^n if the taxable alternative suffers no tax drag. A real advantage therefore needs different payout rates and/or tax drag on the taxable account’s investment earnings. Comparing gross deferred wealth with net taxable wealth would overstate the benefit. The article does not specify returns, horizon, annual taxable distributions or payout taxes for a numerical advantage.
- State exit: 4 U.S.C. 114 protects defined retirement income for nonresidents/non-domiciliaries, including qualifying substantially equal periodic payments made at least annually over at least ten years. A plan being called “10-year installments” and a move alone do not establish qualification. California’s 13.3% is a top marginal-rate scenario, not everyone’s rate or a guaranteed saving on all deferred income.
- Payroll tax: IRS guidance describes FICA’s special timing generally as the later of service performance or the end of substantial forfeiture risk. That can differ from the election year. FICA cannot simply be assumed postponed to payout or always due at the original election.
- 409A: The five-year re-election delay, six-month separation delay and 20% federal additional tax are rule references, not a complete plan-specific checklist. Re-election timing, specified-employee status, exceptions, interest and state consequences matter. The IRS NQDC guide is a starting source; compliant plan terms and professional review decide application.
- Employer risk: The model assumes the employer pays what it owes. An unsecured promise and a rabbi trust do not protect assets from employer creditors. No default probability or recovery model is supplied, so “small catch” is a judgment, not a quantified risk estimate. Plan stability, payout horizon and concentration in your employer all affect value.
Have better inputs, a missing cost or evidence that changes the result? Share it . We will update our assumptions and conclusions to reflect the most accurate representation.