Nobody raised rates.
So why did bonds lose money?

If you hold a mix of stocks and bonds, there's a good chance your last statement showed the stock side up and the bond side down, and you wondered what went wrong. Neither the Bank of Canada nor the US Federal Reserve raised rates this year. So why did bonds lose money?

Short answer: the Bank of Canada and the Federal Reserve each set one rate, the one banks charge each other overnight. Every other interest rate, including the ones on government bonds, is set by buyers and sellers in the market. This year those market rates climbed on their own.

What actually happened

What US government bonds pay, January vs August 2026

Every US government bond pays more today than it did in January. A bond maturing in two years went from 3.47% to 4.19%. Ten years, from 4.19% to 4.65%. Thirty years, from 4.86% to 5.19%.

Canada moved the same way, just less. Our ten-year bond pays 3.74%, up about a quarter of a percentage point over the past year and the highest since May 2024. Our thirty-year pays 4.16%.

Here's the part that trips people up. When new bonds start paying more, the older ones you already own are worth less, because nobody wants a bond paying 3% when they can buy one paying 4.5%.

Why your bond is worth less when rates go up

“Bonds are broken. Try private credit instead.”

Whenever bonds have a rough stretch, a certain argument starts making the rounds. It goes: bonds aren't working anymore, the old mix of 60% stocks and 40% bonds is finished, and you should move some of that money into private loans, infrastructure, private real estate or hedge funds. These get called alternative investments, or just alternatives.

The pitch works because the charts look great. Alternatives usually show solid returns, small ups and downs, and a tendency not to fall at the same time as stocks. On paper it's exactly what you'd want instead of a bond fund that just lost money.

Four reasons I don't buy the pitch

The calm is mostly in the measuring. A regular bond gets a price every day, set by people actually buying and selling it. A private loan or a private building usually gets valued four times a year, by an appraiser using a model. Anything checked less often looks steadier than something checked constantly.

You already own something like this. Canadian house prices peaked in March 2022 and have fallen for sixteen months straight, leaving the typical home about 21% below that peak by June 2026, and closer to 28% down once you adjust for inflation. It's the largest correction since the early 1990s.

Almost no homeowner has felt a single day of it. Nobody sends you a monthly statement for your house. Meanwhile a bond fund down 3% feels alarming, because the number arrives in front of you every month.

MLS Home Price Index Aggregate Composite Benchmark, $841,100 in March 2022 against $665,600 in June 2026. Source: Canadian Real Estate Association. Declines varied widely by region, with Ontario and British Columbia falling further than Alberta and Atlantic Canada.

The same investment, measured two different ways

The "doesn't fall with stocks" part tends to fail when you need it. Private credit means lending money to companies. The kind of economy that knocks down your stocks is the same economy where borrowers have trouble paying loans back. The reason it hasn't looked that way in the numbers is partly the pricing issue above, and partly that the past several years have been kind.

They cost a lot more. These products often charge several times what a plain bond fund charges, and some take a cut of the profits on top. Bonds are the part of your portfolio expected to earn the least to begin with. A big fee eats a much larger share of a small return than it does of a big one.

You can't always get your money out. Many of these products lock your money up for a set period, require months of notice, or can simply stop withdrawals if too many people ask at once. One of the reasons you hold bonds is so there's something you can sell in a bad month without blowing up your financial plan. Something you can't sell doesn't do that job.

To be fair: big pension funds that can leave money untouched for decades have made real money doing this, and getting paid extra for tying your money up is a genuine idea rather than a marketing trick. The question isn't whether these investments work. It's whether they replace what bonds actually do, at the prices ordinary investors pay.

What bonds are actually for

This is where the conversation usually goes off the rails. Bonds aren't there to keep up with stocks. They're there to be the steady, easily sold part: something that usually holds up when stocks drop, pays you income, and can be turned into cash quickly at a price you can see.

A year where they slip a few percent because rates rose isn't a failure. It's the normal cost of owning them. Steady doesn't mean the number never moves. It means it moves less, and for different reasons than stocks do.

Higher rates hurt now and help later

Higher yields hurt the bonds you already own. They also mean every dollar you put into bonds from here earns more than it would have in January.

A Canadian ten-year bond paying 3.74%, or a thirty-year paying 4.16%, hands you noticeably more income than either did through most of the past decade. If you won't need the money for years, rising rates eventually help you. The pain comes first and the benefit shows up later, which is exactly the kind of thing people abandon at the worst moment.

Where the two central banks stand

Policy rates and government bond yields as of August 19 and 20, 2026. Inflation figures are for July 2026. Sources: Bank of Canada, US Department of the Treasury, Statistics Canada, US Bureau of Labor Statistics, Trading Economics.

The Bank of Canada has left its rate alone six times running, reading an economy that's been weak but is picking up, based on their July statement. The Fed is somewhere else entirely: it also held in July, but three of the twelve officials voted against it, and all three wanted rates higher. Inflation there has been above target for more than five years.

The inflation numbers explain most of the gap. Canada's headline rate hit 3.0% in July, at the top of the Bank's range, but almost all of that was gasoline at 25.7% year over year on the Strait of Hormuz blockade. Take gasoline out and inflation held at 2.2% for a third straight month, with the Bank's two core measures at 1.9% and 2.0%, essentially at target.

The US number eased to 3.4% from 3.5%, with core at 2.5%. Lower than Canada's headline in one sense, higher in every sense that matters, because the American figure isn't one commodity distorting an otherwise calm picture. It's been above 2% for five years.

Canadian CPI for July 2026, released August 17 by Statistics Canada. US CPI for July 2026, released August 12 by the Bureau of Labor Statistics.

What I'd actually do

Match how long you're lending to when you'll actually need the money. Spread it across many borrowers and several countries. Protect the foreign portion against currency swings. Keep the fees low. None of that is exciting, and none of it depends on guessing where rates go next. That's the whole point.

If someone does put an alternatives proposal in front of you, four questions cover most of it. How often is it priced, and who decides the price? What does it cost in total, including any share of profits? When can't I get my money out? And what is it replacing in my portfolio, and does it do that job as well?