The 'It's Not Gain Until I Sell' Fallacy

A senior IC with 8,000 RSU shares trading at $275 sees $2.2M of value on the account screen. Asked about the position, the employee says, “it’s not real until I sell it.” The framing sounds sensible: paper gains can evaporate, markets fluctuate, only cash-in-hand is certain. The framing is also financially incoherent and a primary driver of under-diversification.

The fallacy has two components. First, treating an unrealized gain as “not real” makes it feel like you have nothing to lose by holding. Second, treating realization as the moment of gain makes the act of selling feel like “creating” a loss if the stock falls afterward. Both framings lead to the same pathology: never sell. The paper wealth is never converted to diversified wealth, and the diversified alternative never captures the gain.

This article breaks down why the fallacy is wrong, what a correct framing looks like, and the specific mechanical moves that implement the correct framing.

The Accounting View: Value Is Value

In accounting terms, unrealized gains are real. They are reported on brokerage statements, counted toward net worth, pledged as collateral for securities-based loans, used as collateral for mortgages, considered in divorce settlements, taxed at death (under certain transfer taxes), and would be recognized if marked to market.

Tax law treats them as unrealized for purposes of income recognition under IRC §1001 (gain or loss is recognized only on disposition). But this is a tax-timing rule, not a statement about whether the gain exists. The gain exists in every economic sense. The tax is deferred.

A balance-sheet approach: your household balance sheet includes the brokerage account at current market value. Your net worth is calculated on that value. Banks lend against that value. Your quality of life decisions (can I buy this house, can I retire) depend on that value. Treating the value as non-existent until sold is inconsistent with how you actually use the wealth in decision-making.

The Tax-Deferral Illusion

The usual justification for the fallacy: “If I sell, I pay tax. If I hold, I defer tax and earn more.”

The math behind this argument is not actually in favor of holding in most realistic cases.

Scenario A: Sell $2M of concentrated stock with $1M of long-term capital gain. Tax: $238K (23.8% federal plus 3% NIIT for a high earner). Net proceeds: $1,762K. Invest in diversified portfolio. Grow at 8% annually for 10 years: $3,803K.

Scenario B: Hold $2M of concentrated stock. Grow at 8% annually for 10 years (on historical single-stock risk-adjusted return, probably 1-2% lower than diversified, but assume parity): $4,318K. Sell at year 10 with $3,318K of long-term gain. Tax: $790K. Net: $3,528K.

Scenario B holds more in the account during the 10 years and pays more tax at the end. Scenario A pays less tax but invests less. The difference: Scenario A is $3,803K, Scenario B is $3,528K. A is better by $275K.

The reason: deferring tax on a concentrated position does not help if the concentrated position’s risk-adjusted return is meaningfully lower than the diversified alternative. The tax savings from deferral are wiped out by the return penalty on concentration.

If the concentrated stock outperforms the diversified portfolio by 2% annually over 10 years, Scenario B wins. If it underperforms by 2%, Scenario A wins by a wider margin. The historical base rate for single tech stocks underperforming broad indices risk-adjusted is meaningful (perhaps 55-60% of individual tech stocks underperform the NASDAQ 100 over rolling 5-year windows based on 2010-2024 data).

Tax deferral is worth something. It is not worth enough to justify concentration risk on most single stocks.

The Stepped-Up Basis Exception

One case where “hold forever” has real tax logic: if the employee plans to hold until death, §1014 provides a basis step-up equal to FMV at date of death. Pre-death capital gains are permanently forgiven for income tax purposes (though included in the estate for estate tax purposes).

For an employee age 75 with a $5M concentrated position and a $500K basis, selling produces $1.1M of federal tax. Holding until death produces $0 of federal tax on the income side, leaving the full $5M (less estate tax if applicable) for heirs.

This calculus changes for anyone under age 65 with good life expectancy. Plan horizons of 20+ years mean concentration risk dominates tax deferral. Plan horizons of under 5 years (terminal illness, age 85+) mean step-up dominates.

A 40-year-old with a 45-year life expectancy does not benefit from plan-to-step-up reasoning. The 40-year-old who holds concentrated stock waiting for a death basis step-up is exposing $5M to single-stock drawdown risk for 45 years. The expected cost of concentration compounds dramatically over 45 years versus 10 years.

”Paper Gains Evaporate” Is a Real Risk, Priced Into the Sell Decision

The second component of the fallacy: “paper gains evaporate if I hold, and that’s why I should sell.” This is the correct observation that should drive the sell decision. Concentrated positions have high single-stock volatility (35-45% annualized for typical tech stocks). A 40% drawdown from $2.2M takes the position to $1.32M, an $880K loss.

The employee saying “it’s not gain until I sell” has correctly identified this risk but then, paradoxically, concludes that holding is safer because the gain isn’t “real.” This is backwards. The risk is precisely that the gain evaporates while held. Selling converts the risky paper to diversified paper (or cash), reducing the exposure.

The correct mental model: “I have $2.2M of wealth currently held in a risky single-stock form. I can convert it to a less-risky diversified form by selling and re-investing. The cost of conversion is the tax I pay on realized gain. The benefit of conversion is reduced concentration risk.”

Under this frame, the decision becomes a tradeoff between the present-value cost of tax and the present-value cost of concentration. For most tech employees with 20+ year horizons, the tradeoff clearly favors conversion.

Specific Moves That Encode the Correct Framing

Count the concentrated position at full current value in your net worth. Do not mentally discount it as “paper.” Track it as wealth.
Compute the “if I had this amount in cash, what would I do with it” question. Would you buy more of the same stock at current prices? Almost always no. That answer is the diversification signal.
Schedule a quarterly sell of a fixed percentage. 10b5-1 plan, 25% of vests, whatever. The scheduled sale converts paper to diversified on a drumbeat, regardless of current-quarter psychology.
Treat tax paid on realization as a fee for risk reduction. Like an insurance premium. The premium for taking $500K off a concentrated position might be $120K. If the diversification value of that move is greater than $120K, pay it.
Monitor concentration as a percentage of net worth, not as a dollar amount. Concentration at 50% of net worth is always too high. Concentration at 10% is usually fine. The percentage tells you whether to sell, not the absolute dollar amount.

The Framing Flip

Try this reframe: you do not have $2.2M of “unrealized gain.” You have $2.2M of wealth that happens to be stored in a single stock. The question is not “should I realize the gain?” It is “is storing this much wealth in this single asset the right allocation?”

Under the storage-form frame, the gain/loss context disappears. You are not choosing between “selling” and “holding the gain.” You are choosing between two storage forms: single-stock or diversified. The tax cost is a conversion fee between storage forms.

Most people would immediately answer the storage question correctly: “No, storing half my wealth in one company’s stock is too concentrated.” The only thing preventing the diversifying trade is the framing that treats the current storage form as privileged. It isn’t.

Frequently Asked

What if I do believe the stock will outperform? Then you are making a forecast-based bet, not a default-of-inertia. Size the bet explicitly. Most investors recommend no more than 15% of net worth in any single stock even on a high-conviction basis.

Doesn’t every sell-and-reinvest trigger tax? Yes, and the first reinvestment establishes a new basis. Subsequent rebalancing of the diversified portfolio is much lower tax drag because the diversified holdings have less individual appreciation.

What about §1031 or §1045 rollover? §1031 doesn’t apply to securities (limited to real property since 2018). §1045 applies to QSBS-eligible stock rolled into other QSBS within 60 days. Neither is a general-purpose deferral for concentrated stock.

What if I use an exchange fund? §721 exchange funds accept concentrated stock and deliver diversified portfolio exposure without current gain recognition. 7-year holding requirement. Fees typically 50-100 bps annually. Useful for large positions ($5M+) where the deferred tax is large enough to justify the complexity.

Is there a size where holding is clearly correct? Positions under 5% of liquid net worth generally don’t need immediate diversification. The concentration risk is small enough that the tax cost of diversifying may not be worth it until a rebalancing opportunity arises.