Borrowing to invest — leverage, honestly assessed

Borrowing to invest gets described as a way to earn a higher return. It is not. The loan does not improve any investment you make with it — the same fund returns the same percentage whether you bought it with your own money or somebody else's. What a loan changes is the size of your position, and position size does not raise the expected outcome so much as it stretches the range of possible ones in both directions while subtracting a fixed cost from the middle.

That is the whole subject in one sentence, and everything worth arguing about is downstream of it: whether the fixed cost is low enough, whether the tax treatment lowers it further, and — most decisively — whether you will still be holding the position when the range stretches downward. The mechanics are straightforward. The reason most people should not do this is not.

The arithmetic, in one table

Suppose you have $100,000 of your own to invest and you borrow another $100,000 at 6%, giving a $200,000 position. Interest costs $6,000 a year, paid from cash flow; if it is deductible and your marginal rate is 40%, the net cost is $3,600 — an effective 3.6% on the borrowed amount. Here is what one year does to your own capital:

Three things fall out of that table. The break-even is the after-tax borrowing rate, not zero: below a 3.6% return the leverage subtracts, and a flat year — perfectly ordinary, historically common — costs you money you would not otherwise have lost. The upside is real but bounded by the gap between the return and the cost. And the downside is not bounded by anything: a 30% market decline, which happens, takes 63.6% of your capital with it, and the remaining $36,400 now has to earn 175% just to get back to where it started.

UnleveragedLeveraged

Ending capital after one year, by market return

	Unleveraged	Leveraged
−30%	70000	36400
−10%	90000	76400
0%	100000	96400
+7%	107000	110400
+15%	115000	126400

$100,000 of your own capital, $100,000 borrowed at 6%, interest deductible at a 40% marginal rate. The lines cross at a 3.6% return — the after-tax cost of the loan. Market return over one year on the horizontal axis.

The two lines cross once, at the after-tax cost of borrowing, and the leveraged line is steeper everywhere. That steeper slope is the entire product. You are not buying a better return; you are buying a higher sensitivity to whatever return arrives, at a price of 3.6% a year.

What the deduction actually requires

The Canadian rule is genuinely favourable and genuinely narrow. Interest on money borrowed to earn income from a business or property is deductible against your total income; interest on money borrowed for anything else is not. Four conditions do the work, and failing any one of them turns 6% back into 6%.

Purpose. The borrowed funds must be used to acquire something with a reasonable expectation of producing income — dividends, interest, rent. A holding bought purely for capital appreciation is the weaker case.

Current use. Deductibility follows where the money is now, not where it started. Sell the investment and spend the proceeds and the interest stops being deductible from that point, even though the loan is unchanged.

Tracing. The money must travel from the lender to the investment without mixing with anything else. Park it in a chequing account that also holds your pay, even briefly, and the borrowed dollar becomes unidentifiable — and an unidentifiable dollar is not a deductible one.

Non-registered only. Borrowing to contribute to an RRSP, a TFSA, or any other registered plan earns no deduction whatever. This is the single most common misunderstanding, and it inverts the arithmetic of every "borrow to top up your contribution" pitch.

The deduction is also worth the most to the households that need it least. At a 45% marginal rate it cuts a 6% loan to 3.3%; at 20% it cuts it to 4.8%, which is a far less interesting number to be paying for the privilege of a wider outcome range. If you want to see what the same money does compounding without a loan attached, the compound growth calculator is the honest comparison to run first.

Run your numbersCompound growth→

Three ways people borrow, and how they fail

Margin. Borrowing directly from your broker against the securities you hold. Rates are usually the lowest available, the setup takes minutes, and that convenience is the problem. Brokers lend a percentage of a position's market value — commonly around 70% for widely held listed equities — so the loan can only stay within that cushion while prices hold. On the $200,000 position above, a call triggers once the portfolio falls below roughly $142,857, a decline of about 29%. At that point you deposit cash or the broker sells, and the broker decides what and when. Margin converts a paper loss into a realized one at precisely the moment you would most want to hold.

Home-secured credit. A line of credit against home equity carries no margin call, which removes the forced-selling problem and replaces it with a larger one: the collateral is the house. Rates are variable, tied to prime, so the cost of the leveraged half of the plan can rise while the portfolio is falling — the two are entirely capable of happening in the same quarter. How home equity lines of credit work covers the structure, and the HELOC calculator shows what carrying a balance costs. The disciplined version of this arrangement, where every mortgage principal payment is reborrowed and invested, is the Smith Manoeuvre — a long commitment with its own documentation burden.

Investment loans. A dedicated lump-sum loan, frequently sold with a fund purchase attached and sometimes on interest-only terms with no principal repayment at all. The structure is clean for tracing purposes; the sales channel is the concern. A loan recommended by the person earning a commission on what it buys deserves a level of scepticism that "the interest is tax-deductible" does not answer.

The part that actually decides it

None of the above is what usually goes wrong. What goes wrong is that a leveraged portfolio in a bad market produces a monthly interest payment on an asset that is visibly shrinking, and that combination is much harder to hold than the same drawdown without a loan.

Somebody down 20% unleveraged has a bad year and a decision they can defer. Somebody down 40% with a loan still owing has a bad year, a payment due, and a lender relationship — and if the drawdown coincides with a job loss, which is exactly when broad markets tend to fall, they have all three at once. The sale at the bottom is not a failure of nerve so much as the predictable result of a structure that gives you a payment to make and no reason to expect it to stop. Leverage does not merely widen the distribution of outcomes; it shortens the horizon over which you are permitted to wait, which is the one advantage a private investor genuinely has.

Who has any business doing this

The honest list of candidates is short. A horizon of at least ten to fifteen years, so a full market cycle can pass without forcing your hand. Income secure enough that the payments continue through a job loss, backed by an emergency fund held entirely outside the structure. A borrowing rate low and secured, not a promotional rate that reprices. A marginal tax rate high enough for the deduction to matter. No consumer debt of any kind — a credit card balance is a guaranteed cost that dominates any uncertain return, and the distinction between debt that builds and debt that drains is the prior question. And a demonstrated history of holding through a real decline, not a stated intention to.

Everyone else is better served by the boring version: invest more of what you earn, unleveraged, automatically. That approach has a worse maximum outcome and a far better median one, and it never puts you in a position where the market decides how long you get to be patient. If the leveraged plan does not beat it by a margin that clearly compensates for the risk on your own numbers, the risk is not paying you — and the margin has to be large, because you are the one absorbing every unit of it.