The pitch is memorable: fund a permanent policy through your working years, borrow against its cash value in retirement, spend the borrowed money, and let the death benefit repay the lender. Marketed as the insured retirement plan, the insured retirement program, or in its corporate form the corporate insured retirement strategy, it is the mirror image of the immediate financing arrangement: that structure borrows during the funding years to put capital back to work, while this one borrows at the end to spend. Underneath the slide are two very different ways to borrow, an interest deduction that usually does not exist, and a shareholder-benefit question most presentations skip.
What the strategy actually is
The direct answer: an ordinary exempt policy plus a borrowing plan, where the policy does the tax-sheltered compounding and the borrowing converts that value into spendable cash, on terms that differ sharply depending on who does the lending.
The first half is unremarkable and is covered in the corporate life insurance guide: a participating whole life or funded universal life policy grows inside the exemption test without annual tax, which is the entire reason it can serve as a retirement asset at all. The second half is the borrowing plan. Rather than surrendering the policy and paying tax on the gain, the policyholder borrows against the accumulated value, takes the loan proceeds as cash flow, and lets interest accrue or pays it currently. Borrowing from a bank against the policy avoids a disposition entirely; borrowing from the insurer does not, which the next section takes apart, and the two routes are not interchangeable. At death the loan is repaid and whatever remains goes to the estate or, in the corporate version, stays in the corporation.
Nothing in that is exotic. The judgment is in the details, and each detail below is a place where the strategy either holds together or quietly does not.
Policy loan or bank loan: the difference that decides the tax
The direct answer: a loan from a third-party lender secured by the policy is not a disposition and produces no income; a policy loan from the insurer is a disposition and produces income to the extent it exceeds the adjusted cost basis.
This distinction does more work than any other in the strategy, and the two arrangements are routinely discussed as if they were one thing.
A collateral loan or line of credit from a bank, secured by a collateral assignment of the policy, is simply borrowing. There is no disposition, nothing enters income, and the lender relies on the policy's value and its own credit assessment. The policyholder deals with margin requirements, review cycles, and the lender's continued appetite, the same terms surveyed for the leveraged market generally.
A policy loan from the insurer is a different instrument. Subsection 148(9) treats a policy loan as a disposition: the proceeds are the lesser of the loan amount and the policy's cash surrender value net of outstanding loans, and income arises to the extent those proceeds exceed the policy's adjusted cost basis. In the early years, while ACB remains high, a policy loan may produce little or no income. Later, once the ACB has been worn down by the net cost of pure insurance, the same loan is largely taxable. There is symmetry on the way back: paragraph 60(s) allows a deduction for repayments of a policy loan, to the extent the loan was previously included in income and not already deducted.
The practical reading: a bank loan generally gives cleaner tax results for a spending strategy, and a policy loan is simpler to arrange and needs no lender relationship. Which one a plan uses should be a deliberate decision made in advance, not a default discovered at the insurer's service desk in retirement.
The interest deduction that usually is not there
The direct answer: money borrowed to fund retirement spending is not used to earn income, so the interest is generally not deductible, and the collateral insurance deduction fails with it.
This is the honest centre of the strategy and the point that separates it from an IFA. As the interest deductibility post works through, paragraph 20(1)(c) allows an interest deduction only where the borrowed money is used for the purpose of earning income from a business or property. Retirement spending is consumption. No income-earning use, no deduction, and because the collateral insurance deduction under paragraph 20(1)(e.2) requires the interest on the borrowing to be deductible, that capped deduction fails too.
So the interest in an insured retirement strategy is generally a pure after-tax cost, compounding against the policy for as long as the plan runs. That does not condemn the strategy; the tax advantage is in the policy's exempt growth and in the fact that borrowing is not income. It does mean any presentation showing deductible interest on a retirement-spending loan is describing a different structure, or an error.
The corporate version and the guarantee fee
The direct answer: when the corporation owns the policy and the shareholder borrows personally against it, the corporation is providing security for a personal debt, which raises a shareholder benefit question under subsection 15(1) that a reasonable fee is the usual answer to.
The corporate wrinkle is common in practice. The corporation, which had the surplus, owns and pays for the policy. The shareholder, who wants the retirement income personally, borrows from a lender that requires the corporate policy as collateral. The corporation has now provided something of value for the shareholder's private benefit.
The guidance usually quoted here is a CRA technical interpretation from 2006, which treats the question as one of fact to be answered on all the circumstances. It indicates that no benefit is assessed where the shareholder deals at arm's length with the corporation and there is no evidence of an inability to repay when the security is given, and that where the shareholder pays a reasonable fee to the corporation as consideration, the arrangement does not give rise to a benefit. If the corporation is ever called on to honour the security, the amount it pays is itself a benefit, reduced by anything recovered from the shareholder. Two caveats belong with that summary. It is a non-binding interpretation, now two decades old, given on specific facts and carrying CRA's standard warning that it may not reflect its current position. And the arm's-length branch is of little use to an owner-manager, who by definition does not deal at arm's length with a corporation they control, so the reasonable fee is the branch that matters in practice.
What CRA has not done is publish a formula. There is no endorsed rate, no prescribed-rate shortcut, and no safe-harbour percentage, which is where a good deal of casual advice goes wrong. Valuation approaches used in practice, such as comparing the borrowing cost with and without the corporate security, are professional methods rather than CRA's answer. The practical discipline is to price the fee with your accountant, document the reasoning, actually pay it, and record it as income in the corporation.
There is a further trap in the personal-borrowing version worth naming, because it turns a benefit question into an income inclusion. The back-to-back shareholder loan rules in subsections 15(2.16) to 15(2.192) can deem a loan to have been made by the corporation directly to the shareholder where the corporation provides a specified right over property to the lender in connection with the shareholder's debt. Whether they apply turns on how the security is documented, which is exactly the sort of detail that gets settled by a lender's standard forms rather than by anyone's tax plan. Have the assignment reviewed before it is signed, not after the return is filed.
Keeping the borrowing inside the corporation that owns the policy sidesteps both questions, and it is worth considering first, but it does not by itself put money in anyone's pocket. Proceeds borrowed by the corporation belong to the corporation. Funding personal retirement spending from them still requires an extraction, a dividend, a salary, or a shareholder loan, each with its own tax cost: dividends at personal rates net of the corporate tax already paid, salary with payroll consequences, and a shareholder loan that becomes income if it is not repaid within the period the rules allow. That is not an argument against the corporate route. It is an argument for pricing both routes properly, because the honest comparison is the shareholder benefit and back-to-back exposure on one side against the extraction cost on the other, and neither side is free.
The 10/8 boundary
The direct answer: a policy whose investment account is tied to the borrowing is a 10/8 policy, and the consequences reach past the interest deduction into the collateral insurance deduction and the size of the capital dividend account addition.
The 2013 rules ended a specific breed of leveraged insurance design. The statutory term is a 10/8 policy, defined in subsection 248(1), and it captures arrangements where the return credited to the policy's investment account is determined by reference to the borrowing rate, or the maximum account value is determined by reference to the loan amount. Three consequences follow. Interest on the borrowing is not deductible, because subsection 20(2.01) removes it from the interest deductions in paragraphs 20(1)(c) and (d). The collateral insurance deduction is effectively lost as well, because paragraph 20(1)(e.2) excludes the net cost of pure insurance for any period after 2013 during which the policy is a 10/8 policy, and that figure is one of the three amounts the deduction is the least of. And on a death after 2013, the capital dividend account addition is reduced, not wiped out, by the amount of the borrowing outstanding immediately before death. An ordinary loan from an unrelated lender at market rates generally falls outside the definition, which is why mainstream third-party structures continued after 2013. It is still a question to ask directly of any arrangement where the lender and the insurer are closely connected or the rates look suspiciously matched.
Where these plans break
The direct answer: the loan compounds while the policy's performance is not guaranteed, and the failure mode is a lapse or forced surrender that turns a spending plan into a fully taxable gain with no cash left to pay it.
Every risk in this strategy comes from the same place: a growing loan against a policy whose future values are projections, not promises. Dividend scales and credited rates change, the loan balance compounds if interest is capitalized, and lenders re-check their margins. Borrow too aggressively in early retirement and the balance can approach the policy's value, at which point the choices are repayment from other assets, reduced income, or collapse.
The collapse is the case to understand before starting. A lapse or surrender with a large loan outstanding is a disposition, and the resulting policy gain is fully taxable ordinary income, at Ontario's 50.17 percent passive rate inside a corporation rather than the half-inclusion of a capital gain, entering adjusted aggregate investment income as well. The cash that would have paid that tax has already been spent, and the coverage is gone. Conservative borrowing limits, regular in-force illustrations at reduced dividend scales, and an agreed maximum loan-to-value are what keep the plan from finding that edge.
The estate arithmetic, stated honestly
The direct answer: the two borrowing routes end differently, because a bank loan repaid out of proceeds the corporation actually receives generally leaves the capital dividend account credit computed on the full death benefit, while an insurer policy loan is netted from the payout and shrinks the credit with it.
Start with the bank route. The corporation is the policyholder and beneficiary, the assignment is a collateral assignment, and CRA's capital dividends folio treats the corporation as having received the full proceeds even though part of the money goes straight to the lender. The capital dividend account credit is therefore the full death benefit minus the policy's adjusted cost basis immediately before death. Two things are true at once here, and presentations tend to show only the first: the CDA room survives the borrowing intact, while the repayment still reduces the cash actually available to distribute.
The insurer route ends differently. A policy loan is settled out of the death benefit by the insurer, so the proceeds the corporation receives are the net amount, and the capital dividend account addition is computed on what was received. Same spending, same policy, smaller tax-free exit. That difference compounds quietly over a long retirement, which is why the choice between the two borrowing routes deserves to be made at the start rather than discovered by the executor. The CDA estimator prices the room side on specific numbers; the cash side is the proceeds less whatever the loan has grown to.
That last figure is the one to project at several borrowing levels before the strategy starts, because it is what the family actually receives. (Disclosure: I am a licensed insurance and investment advisor, and placing corporately owned policies is part of how I am compensated. This strategy is one I would rather explain fully than sell quickly.) Whether a policy should carry this job at all, alongside the corporation's investments and the rest of the retirement plan, is the standing conversation described in financial planning for physicians.
