A few weeks ago, a reader in Ottawa wrote to me with a fascinating question. He explained that he was investing for his retirement and wasn’t going to touch his savings for at least three decades. Since stocks have always done better than bonds over long periods of time, why should he hold any bonds at all? Why not go with an all-stock portfolio?

My first reaction was to scoff. No one has an all-stock portfolio. It’s too much of a good thing, like eating nothing but ice cream for dinner, or spending eight hours on a roller coaster. But it turns out that our reader was asking a good question, one that still causes heated debate among top investing gurus.

Why do we need bonds?

Conventional portfolios almost always contain a healthy dollop of bonds, because investing can be a wild and scary ride, and bonds can cushion the bumps without significantly hurting your returns.

For example, if you went all-stock and invested your life savings in the major players on the Toronto Stock Exchange 30 years ago, you would have enjoyed an average annual return of 12.1%. That sounds fantastic, but it glosses over the horrifying years you would have endured along the way. Like 1990, when the markets went into a funk that would have cost you 15% of your money in three years, or 2001, when you would have lost 24%.

On the other hand, if you put 60% of your money into stocks and 40% into bonds, your worst drop would have been a minor 8% tumble in 2001. Not only that, but your average return of 11.5% a year would have been almost as high as your return with the all-stock portfolio.

So why would you ditch bonds?

Bonds make investing much less risky, so I was surprised when I called up William Bernstein, author of The Four Pillars of Investing, for a second opinion and he told me that there are indeed cases where you might want to ditch them altogether.

“I’m not entirely unsympathetic to your reader’s point of view,” Bernstein told me from his Portland, Ore., office. “Because he’s right, over long periods of time stock returns probably are going to be higher than bond returns. But it only works if he’s investing his money in a taxable account.”

Bernstein, a well-known author and money manager who has devoted much of his life to analyzing securities performance data, took me through the numbers and showed me that if you’re investing inside an RRSP, where you don’t have to worry about taxes until you take your money out, an all-stock portfolio just isn’t worth it. Part of the reason is that you’ll get an extra boost every time you rebalance a portfolio that contains bonds. For instance, if you have a portfolio that’s 60% stocks and 40% bonds, rebalancing it back to its original proportions once a year forces you to sell high and buy low, delivering an extra 0.3 to 0.5 of a percentage point per year. That helps elevate the performance of a mixed portfolio to the point where it’s getting pretty darn close to that of an all-stock portfolio.

But here’s where the debate starts to heat up: Though your financial adviser would have kittens at the thought of it, Bernstein and others, such as Stephen Jarislowsky, the billionaire Canadian money manager, say that if you plan to hold a large sum of money outside of an RRSP for a long period of time, you may indeed want to ditch the bonds altogether and go 100% stocks. Why? Because the interest you get from bonds is taxed at a much higher rate than the capital gains and dividends you get from stocks, and those extra taxes drag down your returns. Bernstein’s opinion is that if you’re exposed to taxes, an all-stock portfolio boosts your performance enough to make the extra bumps along the way worth it.

But can you handle the ride?

Still, Bernstein had some final words of caution. “The problem with your reader’s point of view is that this is the sort of reasoning you always hear at the top of a bull market. There are a lot of people out there who think they can tolerate an all-stock portfolio, but when the ottoman hits the fan, they’re not quite as disciplined as they thought.”

In other words, if you think that you might start questioning your strategy after three years of negative returns and losing a quarter of your life savings, throw some bonds in there. You’ll sleep much better, even if you make a bit less.