Investor Clinic

The RRSP &lsquo;refund&rsquo; fallacy Add to ...

A couple of weeks ago, I debunked the math that a financial adviser used to justify borrowing to invest. The adviser claimed that investors could make money by borrowing at one rate and earning the same after-tax rate on the investments, which isn’t true. The adviser even argued that investors could “break even” if they earned a lower return than the rate on the loan. Also untrue.

Well, I thought I had put the matter to bed. Then I got an e-mail from Corey.

Corey, as I discovered with some expert journalistic sleuthing (Google), works for a company that provides investment loans to advisers. And he was not pleased with my analysis. Here’s his e-mail, which I’ve edited and condensed for clarity, followed by my response.

****

Hi John:

I was just reading your article on borrowing to invest. I see that you refer to after-tax costs of the loan. Well, depending on whether this is an RRSP loan or investment loan would make a large difference on your apples-to-apples comparison.

How can you not factor in the deductibility of RRSP lump sum contributions or the 100-per-cent deductibility of interest on investment loans for non-registered accounts?

For example, if John Smith makes a \$25,000 RRSP contribution he would receive a \$10,000 refund (assuming a marginal tax rate of 40 per cent), leaving only \$15,000 to pay off. At an 8-per-cent annual interest rate, compounded monthly, he could retire the \$15,000 debt over 10 years with monthly payments of just \$181.99 (or \$21,838.80 in total).

Please ensure you do your homework on borrowing to invest before you inform the public. You and the weatherman are the only ones who can get paid to be wrong.

Corey

***

Hi Corey:

Hope you are having a swell day. I’ll deal with the question of interest deductibility first.

True, if you borrow and invest in a non-registered account, you can usually deduct interest if you’re seeking to earn income. You’ll also pay a preferential tax rate on dividends and capital gains. But the fact remains: You need to earn a higher after-tax return than the after-tax interest rate on the loan for leverage to be beneficial.

Can some people make the strategy work? Sure. Is it a slam-dunk? No. It can be especially challenging when, in addition to paying loan interest, an unsophisticated retail investor is forking over fees of 2.5 per cent or more on mutual funds his adviser bought with the borrowed cash – a common scenario.

Not all financial advisers play this game. I heard from several who said they refuse to use leverage with clients because, although it’s the fastest way to build their book, they consider the practice “predatory” and “despicable” – their words, not mine.

Now to clarify, I didn’t mention interest deductibility in my column for a reason: It didn’t apply to the scenario in question. To keep things simple I assumed that interest costs and investment gains both compound tax-free. This is essentially what happens when you borrow and invest in a registered account, and you can’t deduct interest in that case, so I didn’t.

I also assumed all amounts were after-tax dollars. The analysis would have gotten unwieldy and confusing if I’d tried to account for tax refunds and taxes on withdrawals. But, as you’ll see in a moment, the tax refund is a red herring, anyway. It’s irrelevant.

This may come as a surprise. After all, advisers love to point to the refund as a reason to borrow and invest in an RRSP, because it looks like money is being created out of thin air, which isn’t the case.

You state, correctly, that an investor could borrow \$25,000 and apply the tax refund of \$10,000 against the principal. Doing so would reduce the loan to \$15,000. The investor would then have a loan outstanding of \$15,000, and an RRSP worth \$25,000. Sounds good!

What you’re forgetting is that, in your example, the \$25,000 inside the RRSP is pre-tax dollars. The tax liability on the RRSP – which is what the \$10,000 refund represents – is only being deferred to another day, not forgiven. At the investor’s current tax rate of 40 per cent, his RRSP is theoretically worth only \$15,000 – which is the same as what he owes on the loan.

In other words, the investor is no further ahead, even after getting the refund. Nor will he be further ahead over time, assuming his marginal tax rate doesn’t change. Why? Because as the RRSP grows, his tax liability will grow at the same rate.

To illustrate, let’s fast-forward a decade and see what happens. If the \$25,000 in the RRSP grows by 8 per cent annually (all figures compounded monthly), after 10 years it will be worth \$55,491. But now if he withdraws the money he’ll have to pay tax of \$22,196 (40 per cent of \$55,491). That leaves him with \$33,295 after-tax.

To recap: He spends \$181.99 a month over 10 years to retire the loan – or \$21,838.80 in total – and he now has \$33,295 after-tax, so he’s still done very well.

But here’s the key thing: He didn’t need to borrow and contribute to an RRSP or get a tax refund to achieve the same result. The proof? Instead of taking out a loan, he could have contributed the same monthly payment of \$181.99 directly into a tax-free savings account. After 10 years, assuming the same growth rate as the first scenario, he’d end up with an identical amount of \$33,295.

There are a lot of moving parts here, and one can make borrowing look good or bad by playing with assumptions about marginal tax rates, market returns and interest rates.

But the bottom line is this: All else being equal, an investor must earn a higher after-tax return than the after-tax interest rate on the loan for borrowing to be advantageous. The tax refund may help advisers sell clients on the merits of borrowing, but it can’t alter that inviolable law of investing.

John