The NUA trade is won on cost basis, not tax rates

Net unrealized appreciation grows every year a client leaves employer stock alone inside a 401(k) or ESOP, and its value depends less on the statute than on the client's filing cabinet. As Ed Slott lays out for IRAHelp readers, a participant who distributes appreciated company stock pays ordinary income rates only on the shares' cost basis, while the appreciation that accumulated inside the plan is taxed at long-term capital gains rates when the shares are sold. A $1 million position sitting on $200,000 of basis puts $200,000 in front of ordinary rates in the year of distribution and leaves $800,000 waiting for capital gains treatment at sale; without the NUA election, the full million would eventually be taxed as ordinary income, and the spread on the appreciation is the entire prize.

NUA is conditional, and the conditions deserve a place in the client conversation because few clients arrive knowing them: reaching age 59½, separation from service, disability, and death all trigger eligibility. The carve-outs matter because separation from service does not work for the self-employed while disability works only for the self-employed, a distinction that tracks how the client earns income rather than the size of the position. The plan also has to be completely emptied, and a total distribution event is what satisfies that requirement; selling the stock is not necessary, which is why the high-basis shares a client does not want treated as NUA can be rolled into an IRA as part of the same move.

Executing the distribution is easy; deciding whose basis absorbs the ordinary income treatment is harder, and plans typically answer with one blended figure. Average cost basis takes the total dollars used to purchase the shares, divides by the current value of the position, and applies that percentage to the current share price. In Slott's example, $200,000 of basis against $1 million of stock is 20%, so at a $150 share price every share is credited with $30 of basis. That is the number the recordkeeper reports, and for a client who accumulated shares over decades at very different prices, it is an approximation with a tax cost attached.

Specific identification is where the strategy earns its keep. A participant who bought some shares at $8 early in a career and others at more than $100 in recent years can separate the low-basis lots from the recent shares. The low-basis lots carry the widest gap between basis and current value, which means they generate the largest amount of appreciation eligible for capital gains treatment; the recent shares carry a thinner spread, which makes deferral inside an IRA the better home. Distributing the low-basis shares through NUA while rolling the high-basis shares to the IRA is how the participant maximizes the benefit. Slott is explicit that NUA is not all-or-nothing, since partial distributions are permitted, but specific identification carries a hard prerequisite: the actual purchase price of each share has to be documented. No per-share records, no lot-level election.

Run the arithmetic against the blend and the mismatch surfaces: a $30 average basis assigns $22 more of ordinary income to a share actually bought at $8 than its purchase price implies, while giving the recent shares bought above $100 a lower cost basis than they really carry. The default penalizes precisely the lots a client would want to distribute, which is why specific identification carries the strategy's real edge.

The rate spread between ordinary income and capital gains is fixed by law and identical for every client, so the variable that moves is the size of the ordinary-income piece, and cost basis sets it. That makes the client's records the one input still open to change, and records are a project rather than a signature. As this publication has argued, the last mile of retirement is where advisory value is being repriced now: claiming decisions, survivor income floors, decumulation. NUA belongs on that list because it is a retirement-transition decision most clients will only ever see as a plan-distribution formality.

Average cost basis is the default for a reason: per-share purchase records are frequently gone by the time a client retires or separates, which suggests the plan's blended number is often the only figure still standing. Rebuilding a career of purchase prices from old statements is the kind of task that tends not to get finished, and all the while the trigger event has been met and the plan has to be emptied on whatever terms the recordkeeper provides. That is a records problem before it is a tax problem, the same file-quality issue that runs through the succession wave.

The move at the desk is small and early: ask any client holding employer stock in a plan for the lot-level purchase history at the next review, well before a trigger event forces the distribution. In Slott's example, the basis is $200,000 and the appreciation is $800,000, and the rates that will apply to each are the same for every client. Which shares carry which of those two figures is settled by records the client either kept or did not.