We lost money last year. How do we use that loss?

First, name the loss: the Act runs two separate systems

Before anything can be planned, the loss has to be sorted into its legal category, because the two kinds never mix. A non-capital loss is the operating kind: expenses and deductions exceeded revenue for the year, after adjustments. It is the flexible one, usable against almost any income the corporation earns, business profit, interest, taxable capital gains, in other years. A capital loss arises only when the corporation disposes of capital property, investments, real estate, equipment sold below its tax cost, for less than it paid. Its allowable half can only ever offset the taxable half of capital gains, never operating profit, and no amount of waiting changes that.

The sorting is not always obvious from the financial statements. A year can show an accounting loss but no tax loss, or the reverse, because depreciation, reserves and capital cost allowance move independently, and CCA itself is a choice: in a loss year it can make sense to claim less of it, preserving the deduction room for profitable years instead of deepening a loss that may expire. A single bad year can also produce both kinds of loss at once, an operating shortfall plus a losing investment sale, and each piece follows its own rules from that point on. Getting the classification and the continuity schedule right in the loss year itself is the unglamorous foundation for everything below.

Discretionary deductions make the loss itself a chosen number, within limits. Capital cost allowance never has to be claimed in full, and skipping it in a loss year wastes nothing: the undepreciated balance simply carries forward, ready to deduct against profitable years at full value, with no expiry clock running. A corporation heading into a loss it cannot use soon will often claim little or no CCA, reporting a smaller loss now and keeping the deduction room where it is worth more. The same logic runs through reserves and the timing of discretionary bonuses. Sizing the loss deliberately is the first planning act of a bad year.

Carrying back: the fastest way to turn a loss into cash

The carryback is usually the first option to check, because it is the only one that produces money now. The corporation applies the loss against taxable income it reported in any of the three previous years, and CRA refunds the corporate tax that income paid. The request is made on Schedule 4 of the T2 return for the loss year, no amended returns needed, and for a company that just lived through a losing year, the refund often matters more as working capital than as tax planning: it is the cheapest financing available, your own tax coming home.

The decision inside the carryback is which year and which income to hit. Tax is recovered at the rate the original income actually paid, so a loss applied against profit that was taxed at the Ontario combined small business rate of 12.2% recovers about twelve cents per dollar, while the same loss applied against income that was taxed at the general corporate rate recovers roughly twice that. A corporation whose past three years included profit above the $500,000 limit for the small business deduction, or investment income taxed at the high rate, should aim the loss there first. The window is unforgiving: each year, the oldest carryback year drops off. A loss from last year can reach back three years today, but wait two years to file the request properly and the richest target years may be gone.

Expect the refund to be looked at. A carryback invites CRA to check both ends, the loss year that generated it and the prior year it lands on, and loss years attract more attention than profitable ones on their own. None of that argues against claiming what the law provides; it argues for a loss year with clean books: expenses documented, inventory and receivable write-downs supportable, and the line between capital and operating amounts drawn correctly, since misclassifying a capital loss as an operating one is both common and correctable at your expense.

Carrying forward: twenty years, and the rate you shelter decides the value

Carrying forward wins whenever the future income the loss will shelter is taxed at a higher rate than the past income a carryback would recover, and the corporation can afford to wait for the money. That is the whole trade in one sentence, and it comes up constantly in practice. A company whose past profits all sat under the small business limit, but which expects to grow past $500,000, or to realize investment income taxed at roughly 50%, is holding a loss worth more tomorrow than yesterday. Twenty years is long enough that expiry rarely threatens an operating business, but losses have a way of being forgotten: they sit on the continuity schedule, new accountants inherit them without questioning them, and they quietly expire or get applied against low-rate income by default rather than by decision.

Two cautions belong here. First, the value of a carried-forward loss depends on the company still earning the kind of income it can offset: non-capital losses are flexible, but a net capital loss is only worth something to a corporation that will realize capital gains, which argues for timing gains, selling appreciated investments while losses stand ready, rather than letting the two miss each other by a year. Second, losses do not always survive a change of hands. When control of the corporation is acquired by a new person or group, net capital losses die outright, and non-capital losses survive only against future profit from the same or a similar business. A loss company is worth less to a buyer than its loss balance suggests, and a seller counting the losses in the price is often counting money that will not transfer.

One boundary matters for groups: Canada has no consolidated filing. A loss in one corporation cannot simply be netted against profit in a sister corporation on a return; moving losses around a group takes deliberate restructuring, amalgamation, wind-up, or internal financing arrangements, which is defined-scope work through restructuring rather than a filing position.

A loss year flips the owner-pay logic

The salary-versus-dividend decision inverts when the corporation is losing money, and this is where loss years quietly get expensive for owners who run on habit. Salary is deductible to the corporation, so paying yourself salary in a loss year deepens the non-capital loss, that can be fine, even deliberate, if the bigger loss will be carried back against high-rate income, but it is waste if the loss will sit for years while you paid full personal tax on the salary now. Dividends are not deductible, they come out of the corporate pool without touching the loss, but a dividend requires the corporation to actually have the retained earnings and the cash, and in a bad year it may have neither. Salary also does quieter jobs even in a down year: it creates RRSP room and continues CPP participation, and for owners within sight of retirement those can be worth funding through the cycle deliberately.

Three other pieces of the owner-level picture move at the same time. Shareholder loans often reverse direction: instead of drawing money out, the owner lends money in to cover payroll or rent, and that balance, properly documented, becomes tax-free repayment capacity for later, often the first money the owner takes out in the recovery year. Refundable tax balances do not vanish in a loss year, if past investment income built up refundable dividend tax on hand, a taxable dividend can still trigger the corporation's dividend refund, which occasionally makes a dividend the source of cash rather than the use of it. And the CCPC's passive income keeps counting: a loss on operations does not switch off the investment-income grind computation for the group, so a bad business year with a good portfolio year still needs the full analysis. The moving parts are the same ones that drive every year's compensation plan, laid out in corporate tax planning for owner-managed businesses; a loss year just reverses most of the usual answers.

What a loss touches elsewhere in the corporate system

A loss year leaves fingerprints on accounts that only matter later, and they are easy to miss while everyone is watching cash. Capital losses reduce the capital dividend account: the untaxed half of a capital loss shrinks the balance that could otherwise come out tax-free, which is why a corporation holding both a CDA balance and losing investment positions should usually pay the capital dividend before crystallizing the losses. The character of future income shifts too: profit sheltered by loss carryforwards pays no tax, so those years build no new pools, and the corporation's capacity to pay eligible dividends can thin out. Even the small business limit interacts, since applying losses reduces taxable income, income that would have been active business income taxed at 12.2% costs little to shelter, which is exactly why aiming losses at low-rate income is usually the weakest use of them.

None of these are reasons to hoard losses superstitiously. They are reasons the loss decision belongs inside the whole corporate picture, rates, pools, payout plans, rather than being made line by line on a filing deadline.

The facts that change the answer, and how we handle it

Six facts decide what your loss is worth and where it should go. What kind of loss it is, non-capital, capital or both. What rates your last three years of income actually paid, which prices the carryback. What rate your next few years are likely to pay, which prices the carryforward. How badly the company needs cash now, because a refund in hand can beat a better rate in three years. Whether a sale of the company, or any change of control, is plausible, since that can kill the losses outright. And how you plan to pay yourself through the down cycle, salary, dividends or shareholder loan repayments, because the loss changes the price of each. For corporate groups there is a seventh: which entity actually carries the loss, because a loss stranded in the wrong company may need a reorganization before it can meet the profit it should shelter.

We treat a loss year as a planning year, not a lost year. The work is concrete: classify the loss correctly, model carryback against carryforward at your actual rates, file the Schedule 4 request while the best target years are still open, reset the owner-pay mix for the down cycle, and keep the continuity schedule alive so nothing expires unused. That happens inside our corporate tax work, or as part of a broader plan if the loss coincides with a restructuring or a financing. If last year's return already went in without this analysis, the carryback window is still open for a while; a free 15-minute discovery call is enough to tell whether it is worth reopening.