A business owner has died. How does the estate avoid paying the tax twice?

Where the second layer of tax comes from

The tax gets paid twice because the system treats death and withdrawal as two separate taxable events, and both land on the same value. When someone dies owning shares of a private company, they are deemed to have disposed of every share at fair market value immediately before death, whether or not anything was actually sold. The resulting capital gain goes on the final personal return, the terminal return, and at Ontario's top rates the tax works out to just over a quarter of the gain. Where the deceased built the company from nothing, nearly the entire share value is that gain.

The estate then owns the shares with a cost base stepped up to fair market value, which sounds like the problem is solved. It is not, because the value the family actually wants, the cash, the investments, the business itself, still sits inside the corporation. To reach it, the estate has to take dividends or wind the company up, and either route is taxed as dividend income in the estate's or the beneficiaries' hands. No credit is given for the capital gains tax already paid, so the same dollars are taxed a second time, at dividend rates that run higher than capital gains rates.

It can get worse than two layers. If the corporation itself holds appreciated assets, a building, a portfolio, goodwill in an operating business, the corporate tax on those gains is a third layer waiting inside the company when it sells or winds up. We separate the layers, with the mechanics of each, on our page about why private-company shares are taxed twice at death; the short version is that an unplanned estate can lose well over half of the company's value between the terminal return and the final distribution.

If you are the executor or a family member rather than an accountant, hold on to one orienting fact: nothing about this is a penalty, and nothing about it is final yet. The double tax is simply what happens when two ordinary rules, tax on gains at death and tax on dividends when paid, run into each other with no plan in between. The Act itself contains the tools that stop the overlap, and CRA processes them routinely. What the tools cannot survive is delay, because the cheapest of them is tied to a clock that starts at death.

One exception postpones everything. Shares left to a surviving spouse, or to a qualifying spousal trust, roll over at cost with no gain at all, deferring the entire problem to the second death. That is a genuine reprieve, not a solution: the planning question does not disappear, it moves, and the couple's wills, corporate structure and insurance should be reviewed while there is still time to fix things cheaply.

Everything below is about cancelling one of the layers. Canadian practice has settled on two main plans plus a set of corporate tax accounts, the capital dividend account chief among them, that move money out tax-free under either plan. Which combination fits a given estate depends on deadlines that start running at death, which is why the order of operations matters as much as the plans themselves.

Protect the deadlines first: the graduated rate estate and its first year

Before any plan is chosen, the executor has to protect two administrative statuses, because the cheapest fix dies with them. The first is the estate's designation as a graduated rate estate, a GRE, made in the estate's first T3 trust return. A GRE pays tax at graduated rates rather than the top rate for up to 36 months, can choose an off-calendar year-end, and, critically, is the only kind of estate that can use the subsection 164(6) loss carryback. Only one estate per person can be a GRE, which matters where multiple wills or existing trusts are in play.

The second is the choice of the estate's first taxation year-end, which the executor can set at any date up to twelve months after death. The 164(6) plan must be completed within that first taxation year, so the year-end choice literally sets the planning deadline. Estates that pick a date for administrative convenience, or let one happen by default, routinely shorten their own runway without anyone noticing until it is too late.

GRE status is also easier to lose than most families expect. The designation must be made in the first T3 return, the estate must arise on and as a consequence of the death, and property contributed to the estate from anyone other than the deceased can taint the status entirely. Where the deceased used multiple wills, common in Ontario to manage probate on private-company shares, the wills together still make up one estate for this purpose, but the returns and the designation need to be coordinated so the GRE claim is clean.

The terminal return runs on its own clock. It is due April 30 of the year following death, June 15 where the deceased or their spouse carried on a business, and never earlier than six months after the date of death for deaths late in the year. The tax on the deemed gain is payable with it, and where the estate is asset-rich but cash-poor an election exists to pay the tax attributable to the deemed disposition in instalments over several years, with security posted to CRA. A separate optional return for rights or things, amounts earned but unpaid at death, can shave the bill further by using a second set of graduated brackets.

Underneath all of it sits valuation. Every plan runs off the fair market value of the shares at death, and the executor files that number before CRA ever looks at it. A defensible valuation, built from the company's earnings, assets and history rather than guessed, is the foundation of the terminal return, the estate's cost base and every post-mortem step that follows. It is usually the first piece of work commissioned on a file, because nothing downstream can be modelled without it.

Fix one: the subsection 164(6) loss carryback cancels the capital gain

The loss carryback erases the gain from the terminal return, and it is the plan with the hard deadline. The company redeems the estate's shares, or is wound up, within the estate's first taxation year. The redemption is taxed as a deemed dividend, and because the deemed-dividend portion is carved out of the proceeds for capital gains purposes, the estate also realizes a capital loss roughly equal to the gain that arose at death. Subsection 164(6) lets a graduated rate estate carry that loss back to the terminal return, cancelling the deemed gain and recovering the tax on it.

The arithmetic underneath is worth seeing once. On a redemption, the amount received above the shares' paid-up capital, a tax number that is usually nominal in a founder's company, is a deemed dividend. The same amount is excluded from proceeds for capital gains purposes, so the estate's proceeds fall below its stepped-up cost base, and the difference is the capital loss. One transaction, two characters: dividend in, loss out, and the loss travels back one year to meet the gain it mirrors.

The result is that one layer survives: dividend tax, in the estate, on the value coming out. Whether that is a good trade depends almost entirely on the corporation's tax accounts. Where the company has a capital dividend account balance, part of the redemption can be paid as a tax-free capital dividend. Where refundable tax has accumulated on past investment income, the taxable dividend triggers a corporate refund that offsets a real part of the cost. Where GRIP exists, the dividend can be designated eligible and taxed at lower personal rates. With enough of those attributes stacked, the 164(6) route becomes far cheaper than the headline dividend rate suggests.

The plan has sharp edges. It must be completed, not merely started, within the first taxation year, and the estate must be a GRE. Stop-loss rules can reduce the usable loss where corporate-owned life insurance funds the redemption, subject to grandfathering for some long-standing arrangements, so an insurance-funded plan has to be modelled before any resolution is signed. The deadline arithmetic, the election paperwork and the traps are set out on our page on the subsection 164(6) loss carryback.

Because the window is absolute, 164(6) is the default plan for estates that need value out quickly, and the fallback that stays alive while something better is considered. A well-run file often books the redemption capacity early in the first year precisely so the choice remains open until the analysis is done. It also tends to win outright in smaller companies, where the value at stake does not justify a pipeline's professional cost, and in estates whose beneficiaries want a clean, fast wind-up rather than years of note repayments.

Fix two: the pipeline keeps the capital gain as the only tax

Pipeline planning cancels the second layer instead: value leaves the company as loan repayments rather than dividends, so the capital gain on the terminal return remains the only tax paid. The estate transfers its shares, carrying the high cost base created at death, to a new corporation in exchange for a promissory note. Over time the operating company's assets move up into the new corporation, typically through an amalgamation or wind-up after a reasonable interval, and the note is repaid to the estate in stages. Repaying a note is not income, so the family receives the company's value with no dividend tax at all.

The trade-offs are time and formality. CRA expects the underlying business to carry on for a period after the transfer and the note to be repaid gradually; a pipeline executed too fast, with the company collapsed for cash immediately, risks being recharacterized as a deemed dividend, the very tax it was built to avoid. The plan needs a supportable valuation, careful sequencing of corporate steps, tax and legal documents that match, and in larger or unusual files an advance income tax ruling. The step-by-step version, including timing conventions, lives on our page on pipeline planning after the death of a business owner.

Choosing between the plans is rarely all-or-nothing. Capital gains rates in Ontario sit meaningfully below top dividend rates, which is why the pipeline usually wins on pure arithmetic, but the estate's cash timeline, the corporate accounts and any stop-loss exposure all pull on the answer. Most well-advised estates land on a hybrid: enough shares redeemed under 164(6) to use the capital dividend account and trigger the refundable tax, with the balance pipelined so the remaining value comes out at capital gains cost.

A pipeline sits comfortably alongside a sale of the underlying business, and the combination is common. Where the operating assets are sold to a third party after death, the corporate-level tax on that sale gets paid either way; the pipeline's job is only to make sure the after-tax proceeds reach the family as note repayments rather than dividends. The sequencing between the asset sale, the CDA additions it creates and the pipeline steps is exactly the kind of detail that gets modelled before anything is signed.

A pipeline also assumes there is somewhere for the value to go slowly. An estate with beneficiaries scattered across tax situations, or with a business about to be sold to a third party, may find the loss carryback's speed worth its rate. The plans are tools, not ideologies, and the file's facts pick the tool.

Take inventory before choosing: the corporate accounts change the price of every plan

Before anyone commits to a plan, the corporation's tax accounts need to be computed, because they change what each option costs. None of these balances appears on the financial statements; each is a running computation across the company's whole history, and death itself moves several of them, most dramatically when corporate-owned life insurance pays out. The inventory looks like this:

Timing runs through the inventory too. Insurance proceeds credit the CDA only when received, refundable tax is recovered only in a year a taxable dividend is paid, and a capital loss elsewhere in the company can shrink the CDA before the estate gets to it. The accounts are therefore not just computed once; they are scheduled, so each balance is used in the year, and in the order, that wastes none of it.

The inventory is what makes hybrid design possible. Tax-free and low-cost capacity is used first: shareholder loans repaid, capital dividends elected and paid, taxable dividends sized to recover the refundable tax. Only the value left after that needs a pipeline. An estate that runs a full pipeline while a large CDA balance sits unused has not avoided double tax; it has overpaid single tax.

Computing the accounts takes real work on an older company: decades of gains and losses, insurance policies with adjusted cost basis nobody tracked, old capital dividend elections. Where balances are material we verify them with CRA before relying on them, because paying a capital dividend the account cannot support brings its own penalty tax, and an estate is the wrong place to discover that.

The facts that change the plan, and how the first year should run

Six facts swing nearly every post-mortem file, and they are worth checking against your own situation before the first advisor meeting:

• Whether a spouse survives. A spousal rollover defers everything and reframes the work around the second death, including whether to elect out of the rollover on some shares to use the accounts now.
• Insurance. Corporate-owned proceeds change the CDA, the liquidity and the stop-loss analysis all at once; the policy details matter, not just the amount.
• The corporate accounts. CDA, refundable tax and GRIP decide how cheap the 164(6) route can be, and therefore how much of the estate should be pipelined.
• The family's intentions. Keeping the business, selling it to a third party or winding it down each favours a different plan and a different clock.
• The estate's cash needs. A pipeline pays out over years; an estate that must fund tax, probate costs or bequests soon needs the faster tools first.
• Who inherits. Non-resident beneficiaries complicate both plans, and a family farm corporation passing to children brings the intergenerational rollover into play, which can change the strategy entirely.

For the first meeting with an advisor, gather the file the planning actually runs on. The pieces that matter most are:

• The corporate minute book, share register and any shareholder agreement, because they say who owns what and what must happen on death.
• The last several corporate tax returns and financial statements, which are the starting point for computing the tax accounts.
• Every life insurance policy touching the company, with its ownership, beneficiary and history, not just the face amount.
• The will or wills, and anything already filed for probate.
• A list of what the company owns, especially real estate, investments and anything carrying an unrealized gain.

Around the plans runs ordinary estate administration, and it cannot be ignored: probate, the valuation, the terminal return and any rights-or-things return, the estate's T3 filings, and eventually a clearance certificate before the final distribution. The corporation does not pause while the estate grieves. Corporate filings, payroll and banking continue, and directors need to be replaced quickly so someone can lawfully sign resolutions, including the ones the tax plan requires.

A well-run first year has a recognizable rhythm. The opening months secure the valuation, the GRE designation and the year-end choice while the lawyer moves probate along. The middle months are modelling: the accounts computed and verified, the plans priced side by side, the hybrid designed and documented. The closing months are execution, redemptions, elections, transfers, each dated inside the deadlines chosen at the start. Estates that follow the rhythm rarely pay the second layer; estates that treat the first year as a grieving pause and the second as a filing exercise usually do.

Post-mortem work is genuinely multi-disciplinary. The executor, the estate lawyer and the accountant each hold pieces, and because nearly every immovable deadline is a tax deadline, the accountant usually ends up holding the schedule. For a post-mortem tax planning accountant in Ontario, this is defined-scope work with a clear arc: value the shares, compute the accounts, model 164(6), the pipeline and the hybrid side by side, then execute the elections, returns and corporate steps with counsel. We run it as a Strategic Projects engagement under our post-mortem planning service, with a written scope and fee after a free 15-minute discovery call.

If you are reading this in the months after a death, two things are urgent and neither is a tax form: confirm the estate will qualify and be designated as a GRE, and choose the first year-end deliberately, because that date is the deadline everything else on this page must fit inside, and it only gets chosen once.