Your Company Stock Has to Beat the Market by 3 Points a Year to Be Worth Holding

A question that turns up on every investing forum, in some form: eight years of accumulated company stock, a large unrealized gain in a taxable account, a growing sense that this is too many eggs in one basket, and no plan. How fast should you sell, and where should the money go?

The usual answer is a list of tax tactics. Sell across tax years, use specific lots, pair with harvested losses, donate the low-basis shares. All of that is useful and all of it is downstream of a question nobody seems to answer: what is holding actually costing you?

Both sides of that are calculable. Deferring the tax keeps more capital compounding, which is worth a real and specific amount. Concentration drags your typical outcome down, which costs a real and specific amount. Subtract one from the other and you get the number that decides it: how much your stock has to beat the market by for holding to make sense.

This is a risk question with a tax constraint

The tax bill is visible, immediate, and quantifiable. The concentration cost is invisible, deferred, and probabilistic. That asymmetry is why the question feels like a tax problem, and why the instinct is almost always to wait.

The instinct is not stupid. Waiting genuinely does something for you, and the rest of this guide would not be credible if it pretended otherwise. So start there.

What deferring the tax is actually worth

Sell today and you pay tax on the embedded gain, so you reinvest less than the position is worth. Hold, and the whole position keeps compounding with the tax bill postponed. That head start is real.

Take a $1,000,000 position with a $100,000 basis at a combined 23.8% rate, against a diversified portfolio expected to return 7%. Selling costs $214,200 in tax and leaves $785,800 invested. Setting the two after-tax outcomes equal and solving for the return the stock needs:

So the deferral is worth a bit over a point a year. Two features of that number are worth noticing because they run against intuition.

It shrinks the longer you hold. Deferral is a one-time boost to how much capital is working, and spreading that boost over more years makes it smaller in annual terms. The common belief that a long horizon strengthens the case for holding has it backwards.

It shrinks as your basis rises. At a zero basis the deferral is worth 1.32 points a year; at a 90% basis it is 0.13. If your shares came from recent vests near the current price, there is very little to defer and the tax argument for waiting is close to empty. Our guide on selling RSUs at vest covers that case, where the basis is the vest price and the gain is usually small.

Waiting for a better price costs $3.20 per dollar saved

A specific version of waiting deserves its own arithmetic, because it is the one people actually do: holding through a decline, on the theory that a smaller gain means a smaller tax bill.

It does. The tax bill falls by the tax rate times the decline. Your wealth falls by everything else. The ratio depends on nothing but the tax rate:

At 23.8% that is $3.20. At a 15% rate it is $5.67. At 33.3% it is $2.00. On the $1M position, a 30% decline shrinks the tax bill by $71,400 and shrinks your after-tax wealth by $228,600. You would be trading three dollars for one, and you would be doing it on purpose.

The only way this trade improves is if your tax rate rises, which is the opposite of what someone waiting for a lower bill is hoping for.

What the concentration is costing

Now the other side. A single stock and a diversified fund can have the same expected return and still deliver very different typical outcomes, because volatility drags the median below the mean by roughly half the variance. That is the same variance-drag relation behind leveraged portfolios, applied to a position you already own.

Measured on daily returns from 2016 to 2026 for eleven widely held large-cap employer stocks:

The split matters more than the total. Between 41% and 75% of a single stock’s variance is idiosyncratic, and idiosyncratic risk earns no expected premium. The systematic part does earn one, and you keep it by owning the market. Selling the position sheds the unpaid risk and keeps the paid risk. Our guide on compensated and uncompensated risk develops that distinction.

For the median stock in that table, the idiosyncratic part alone drags the typical outcome by 4.06 points a year. Against a deferral benefit of 1.18. The evidence on what happens to individual stocks over long horizons points the same way: most of them underperform Treasury bills, while a tiny minority produce all the net wealth.

Putting the two together

Subtracting the deferral benefit from the volatility cost gives the hurdle: on the baseline case, the stock must be expected to beat the diversified portfolio by about 3.1 points a year, every year, to leave you no better off after tax.

That number is not universal, and the way it moves is the practical point. It depends far more on your stock’s volatility than on your tax rate or your basis.

Read the top row and the bottom row together. For a utility-like employer stock at 25% volatility, the hurdle is a third of a point and holding is a reasonable position to take. For a 50%-volatility growth name it is ten points a year, sustained, which almost nothing delivers. The concentration question is mostly a volatility question, and volatility is something you can look up rather than guess.

Put your own position in:

When holding is the better call

Four cases genuinely favor holding, and it is worth being specific about them rather than treating diversification as an unconditional rule.

A low-volatility stock. Under about 25% volatility the hurdle drops below a point a year, which a reasonable person could expect from a company they know well. The concentration is still uncompensated, but the cost is small enough to trade against other considerations.

You will never sell. If the position passes to heirs and the basis is adjusted at death, the deferral is worth 2.55 points a year rather than 1.18, and the hurdle falls to about 1.8. This is the strongest argument for holding, and it only works if you genuinely never need the money. Holding a concentrated position to preserve a tax benefit for heirs while it funds your own retirement is two incompatible plans.

You are giving it away. Donating appreciated shares avoids the gain entirely rather than deferring it, which makes the lowest-basis lots the natural ones to give. Our guide on donating stock versus cash covers the mechanics and the 2026 rule changes that shrank the benefit.

The shares might qualify under Section 1202. Stock in a qualifying small business, held long enough, can receive a substantial federal exclusion on the gain. The rules are technical and depend on the company, the issuance, and the holding period. If there is any chance your shares came from an early-stage employer, find out before you sell, because the sale is irreversible and the exclusion is not.

Gradual or all at once

The numbers above settle this more cleanly than the usual debate does. Spreading sales across tax years is worth doing when it changes the rate you pay: filling a lower bracket, staying under the net investment income tax threshold, using a low-income year, or absorbing capital losses you already have. Those are specific, and you can put a dollar figure on each one.

Spreading sales because the tax feels large is a different thing. Every year you stay concentrated costs roughly the hurdle above. On the median case that is about three points a year on the un-sold portion, against a tax saving that is usually a fraction of that. A schedule measured in a few tax years can clear that bar. A schedule measured in a decade generally cannot.

Two practical rules follow. Put a completion date on the plan, because gradual diversification without a deadline reliably becomes no diversification. And do not make the schedule depend on the price, since a plan that says “sell when it recovers” is a forecast wearing a schedule’s clothing, and it has no completion date at all.

For which shares to sell first, our guide on specific-lot identification covers the mechanics, and tax-loss harvesting covers pairing sales with losses elsewhere.

Where the money goes

Briefly, because this part is not controversial and is covered elsewhere. The reinvestment decision has nothing to do with the stock you sold. Pick the allocation that fits the household, then buy it.

One trap is worth naming: selling a technology employer’s stock and buying a technology or large-cap growth fund reduces single-name risk while keeping most of the same economic exposure. If your salary depends on the same sector, that matters more than it looks. See human capital risk for tech workers and the tech bro portfolio.

The other trap is delay. Once you have decided to sell and the money is earmarked for long-term investment, holding it in cash while you wait for a better entry is a second market-timing decision that the tax argument does not cover.

Where to stop and get help

Some fact patterns are worth professional advice before you place a trade, because the mistakes are expensive and irreversible:

• Shares that might qualify under Section 1202, where selling forfeits the exclusion.
• Officer, director, or insider status, where trading windows and a 10b5-1 plan come before any tax question. Our RSU guide covers the basics of those plans.
• Employer stock inside a 401(k), where net unrealized appreciation is a separate and one-shot decision that a rollover can destroy.
• Private company shares, unusual option exercises, or shares held in trusts and partnerships.
• Exchange funds, collars, prepaid variable forwards, and other structures whose costs and lock-up terms are the whole story.
• A planned move between states, which can change the arithmetic more than any of the tactics above.

How Summitward helps

The calculator above answers the hurdle question with assumptions you type in. Summitward’s Concentration tab, part of Summit, runs the position-level version on your actual holdings: your Herfindahl index and effective number of positions, your largest position and what share of the portfolio it is, a sell-versus-hold analysis using your real cost basis and holding period, and a five-year Monte Carlo projection comparing holding against diversifying.

The free Portfolio X-Ray needs no login and will tell you whether the funds you plan to buy with the proceeds occupy the same slice of the market as the stock you are selling.

Frequently asked questions

How much company stock is too much?

The house rule of thumb is that any single position above 10% of the portfolio warrants a deliberate decision rather than passive holding, and below 5% is a reasonable target. When your salary depends on the same company, both numbers should be lower, because the position and the paycheck fail together. The more useful test is a dollar one: work out what a 60% decline in the position would do to your funded goals, and if the answer changes your retirement or your housing plans, the position is too large regardless of the percentage.

Doesn't waiting for long-term treatment save a lot?

For shares held under a year, yes, and that is usually worth waiting for: the gap between ordinary rates and long-term rates is far larger than any hurdle in this guide. For shares already long-term, there is nothing further to wait for.

My company is genuinely excellent. Doesn't that change it?

It changes the expected-return term, which is exactly what the hurdle is denominated in. The question is not whether the company is good but whether it is expected to beat a diversified portfolio by three points a year for a decade. Note also that a good company can be a bad position size, and that knowing a company well is different from knowing that its price is wrong.

Why does the deferral benefit shrink over a longer horizon?

Deferring is a one-time increase in how much capital is invested. Its value in annualized terms is that head start spread over the holding period, so more years means a smaller annual figure. The total benefit still grows; the per-year advantage it buys you does not.

What if tax rates rise?

Higher future rates strengthen the case for selling now, and they are the one scenario where holding through a decline gets less bad rather than more. The model here applies one rate at both dates, which is neutral on this question rather than favorable to either side.

Does direct indexing solve this?

It can help fund a diversification schedule by generating losses to absorb the gains, and a low-basis concentrated position is one of the few fact patterns where the fees plausibly pay for themselves. It does not let you keep the stock and cancel the tax. See do you need direct indexing and tax-aware long-short, which cover where those strategies are oversold.

Key takeaways

• Deferring the tax is worth about 1.2 points a year on a low-basis position over ten years. It shrinks with a longer horizon and with a higher basis, which is the opposite of how it is usually described.
• Waiting through a decline costs $3.20 per $1 of tax it saves at a 23.8% rate. The ratio is (1 − t) / t and depends on nothing else.
• Between 41% and 75% of a single stock’s variance is uncompensated. For the median large-cap employer stock that unpaid risk costs about 4.1 points a year of the typical outcome.
• Net, the hurdle is around 3 points a year for a typical position, and it is set mostly by your stock’s volatility: under a point at 25%, about ten points at 50%.
• Sell on a schedule with a completion date, and let taxes shape the path rather than the destination. Price-contingent plans do not finish.
• Holding wins in four specific cases: a low-volatility stock, a position you will genuinely never sell, shares you are donating, and shares that may qualify under Section 1202.

Related guides

• Concentration Risk for HHI, effective positions, and the portfolio-level framework.
• When to Sell a Winning Stock for the one-time tax cost and a decision framework.
• Tax-Efficient Selling and SpecID on choosing which lots to sell.
• Should You Sell RSUs at Vest on stopping the position from growing back.
• Most Stocks Lose to T-Bills on what happens to individual stocks over long horizons.
• Compensated vs. Uncompensated Risk on which risks carry an expected premium.
• Donate Stock or Use a DAF on the one exit that avoids the gain rather than deferring it.

Sources and method

1. Break-even and hurdle figures are the author’s own calculations, reproducible from web/src/lib/concentrated-stock-math.ts, which carries the reference values used here. The comparison equates after-tax terminal wealth from selling today and reinvesting against holding and paying tax at the horizon, applying one tax rate at both dates and ignoring interim dividend taxes. The stock is assumed to have the same expected return as the diversified portfolio, which is the most favorable assumption for holding, so the hurdles are conservative.
2. Volatility, beta, and idiosyncratic risk are the author’s calculations from daily adjusted closing prices, January 2016 through July 2026, roughly 2,650 trading days per stock. Beta and residual volatility come from a regression of each stock’s daily log returns on SPY’s. Median of the eleven stocks: 34.4% total volatility, 28.5% idiosyncratic.
3. The variance-drag relation, that volatility reduces the median outcome by approximately half the variance, is the standard result for lognormally distributed returns and is the same relation used in our leverage material.
4. These are measurements and models over stated windows, not forecasts. Tax rules change and vary by state and by investor. Nothing here is tax, legal, or investment advice, and the fact patterns listed under “where to stop and get help” genuinely warrant a professional.