Rate hikes and cuts: the price of money.

012022: from zero to 4.5% in nine months

For two years the Federal Reserve had held its target rate near zero. On March 16, 2022, with inflation running far above its 2% goal, it raised the range to 0.25–0.50%. Then it kept going: seven hikes that year, four of them three-quarters of a point at a time, ending at 4.25–4.50% in December. It peaked at 5.25–5.50% in July 2023. The Bank of Canada moved in step: its first hike came on March 2, 2022, and it peaked at 5% that same July.

Almost everything with a price felt it. Over 2022:

• The 10-year US Treasury yield rose from 1.52% to 3.88%.
• The average 30-year US mortgage rate went from 3.11% to 6.42%, after touching 7.08% in the fall.
• The broad US investment-grade bond market lost about 13%, its worst year in records going back to 1976.
• The S&P 500 fell 19.4% and the tech-heavy Nasdaq Composite 33.1%, before dividends.
• The US dollar rose: at its peak in late September it was about 11% stronger against a broad basket of currencies than at the start of the year.

And in March 2023 the hikes helped bring down a bank (section 06). This lesson is about how one rate, on overnight loans between banks, reaches that far.

02What the central bank actually sets

A central bank sets one short-term rate, its policy rate: the rate at which banks lend to each other overnight. In the US it is the Federal Reserve's target range for the federal funds rate; in Canada, the Bank of Canada's overnight rate. It steers that rate mainly by setting what it pays banks on the money they keep with it; what it charges banks to borrow from it acts as a backstop.

Other short-term rates follow closely: Treasury bills, money market funds, credit cards, lines of credit, variable-rate mortgages. Ordinary savings accounts follow slowly and only partly: from early 2022 to the end of 2023 the average US savings rate rose from 0.06% to 0.46%, while the Fed's rate rose more than five points. Our Fed target rate and Bank of Canada rate pages track both, with their history.

Long-term rates, like the 10-year government bond yield and the US fixed mortgage rates tied to it, are set by the market, not the central bank. They reflect where investors expect the policy rate to go over the years ahead, plus, usually, a little extra for locking money up that long. The central bank moves them mostly by changing those expectations. Since 2008 the Fed, and since 2020 the Bank of Canada, have also bought long-term bonds directly, called quantitative easing, and later let them mature without replacing them, called quantitative tightening; both move long rates without touching the policy rate. And sometimes long rates fall after a hike, when the market reads it as inflation beaten and expects cuts later. When long rates end up below short ones, the yield curve is said to be inverted, a story for another lesson.

03Why it raises or cuts

Central banks are balancing two things: prices and jobs. The Federal Reserve's mandate from Congress is maximum employment and stable prices, which it reads as 2% inflation over the longer run, measured by the PCE price index. The Bank of Canada targets 2% inflation, the middle of a 1% to 3% range.

• Hike when inflation runs too hot. Costlier borrowing and better rewards for saving cool spending, hiring and price rises.
• Cut when the economy is weakening and inflation is under control. Cheaper money encourages borrowing, spending and investment.

The effect is slow. A hike today shows up in spending and prices over the following year or two, so central banks are steering by where they think the economy is heading, not where it is.

04How one rate travels through everything else

The same arithmetic runs through the rest of the economy. A business deciding whether to borrow for a project that returns 6% says yes when borrowing costs 3% and no at 7%. A saver earning 4.5% on $10,000 gets $450 a year instead of $50 at 0.5%, and is a little more inclined to save than spend. Higher rates also tend to strengthen the currency, as foreign money comes in to earn them, which makes imports cheaper and exports harder to sell. What counts is a country's rates relative to everyone else's. In 2022 the Bank of Canada raised its rate four points, nearly as much as the Fed, and the Canadian dollar still fell about 6% against a US dollar that was rising against most currencies.

05Bonds: the seesaw

A bond pays a fixed coupon. When rates rise, new bonds pay more, so an old bond's price must fall until its yield matches theirs. When rates fall, old bonds with higher coupons become more valuable. Price and yield sit on opposite ends of a seesaw, as the bonds lesson explains.

The longer the bond, the further its end of the seesaw swings. Start with bonds paying 4% and let rates rise one point: a 2-year bond falls 1.9%, a 10-year bond 7.8%, a 30-year bond 15.5%. The short bond hands your money back soon, to reinvest at the new rate; the long one locks you into the old coupon for decades. That sensitivity is called duration.

06When rates find the weak hand: Silicon Valley Bank

Silicon Valley Bank was the bank of the tech industry. Between 2019 and 2021 it tripled in size as startups flush with cash deposited it there, and it put much of the money into long-term bonds while rates were near zero. By the end of 2022, securities made up 55% of its assets; at similar banks the average was 25%.

Then rates rose and those bonds fell, just as long bonds do on the seesaw. By the end of 2022 the losses on the bonds it had labeled "held to maturity" came to $15.2 billion, close to the bank's entire $16.0 billion of shareholders' equity. The accounting rules let it carry those bonds at what it paid, as long as it never had to sell them; the losses sat in the notes to its accounts, not in its reported capital. And as rates rose during 2022, it removed hedges that would have protected it against rising rates.

Meanwhile the money was leaving. Startups were burning cash with less new funding coming in, and they drew down their deposits to pay the bills: SVB's deposits fell from $198 billion in March 2022 to $173 billion at the end of the year. On March 8, 2023, to raise cash, it sold almost all the bonds it was free to sell, about $21 billion, at a $1.8 billion after-tax loss, and said it would raise $2.25 billion by selling new shares, the kind of emergency share sale the dilution lesson describes. About 94% of its deposits were above the insurance limit, and those depositors took fright: on March 9 more than $40 billion left in a single day, and more than $100 billion more was lined up to leave the next morning. Regulators closed the bank on March 10.

Rates were the weather, not the whole story. US banks together were sitting on $620 billion of unrealized losses on their securities at the end of 2022, and nearly all of them stayed open. SVB had piled into long bonds, dropped its protection against rising rates, and leaned on one kind of depositor, most of them uninsured and quick to move together. The lesson: a loss on a bond is only on paper until you have to sell, and the people who decide when you have to sell aren't always you.

07Stocks: the value of later

A stock is worth the cash the business will pay out over its life, each dollar discounted back to today. Higher rates make a dollar received later worth less today, and the further away the dollar, the bigger the hit.

That is why growth stocks, whose profits lie mostly in the future, tend to fall harder when rates rise than value stocks, which pay out more today. One simple model makes it concrete: price = next year's payout ÷ (the return investors require − the payout's growth rate). The return investors require is roughly what a safe bond pays plus a few points for the risk of owning a business. With bonds at 4% and a 5-point premium, call it 9%. At 9%, a value stock that will pay $1 next year, growing 1% a year after that, is worth $12.50; a growth stock that will pay $1 next year, growing 4% a year, is worth $20. Raise bond yields one point and, if the premium holds, the required return goes to 10%: the value stock falls 11.1%, the growth stock 16.7%.

The model is a sketch, and the premium rarely holds still. Real stocks also move on what rates say about the economy: a cut because inflation is beaten is good news for profits; a cut because a recession is coming may not be.

08Surprises move markets, not decisions

Most hiking cycles are quieter than 2022. The S&P 500 rose about 12% from the first hike of the 2004–06 cycle to the last, and about 21% across the 2015–18 cycle: the economy was strong, and the hikes came slowly and were expected. 2022 hurt because the start was zero and the pace was the fastest since the 1980s. In early February 2022, futures priced in about four quarter-point hikes for the year. The Fed delivered the equivalent of seventeen.

Markets trade on what they expect, and they expect a lot. Futures on the federal funds rate show what traders think the Fed will do at each meeting, and by decision day the move is usually fully priced in. A widely expected hike may barely move markets on the day; what moves prices is a surprise in the decision, or in what the central bank says about the meetings ahead. That is why investors parse every word of the statement and the press conference.

09Who wins and who loses

"Tends to" is doing real work in that table. Each row is a pull, not a law; a war, an earnings miss or a change in what the market expects can outweigh it.

10Try it: move the rate

Slide the rate change and watch four prices respond: two bonds, and the two stocks from section 07. The gold lines are the bonds (solid, 30-year; dashed, 2-year), the cream lines the stocks (solid, growth; dashed, value). Below the chart: what the same move does to a mortgage payment and to savings.

The model moves every rate by the same amount, which real markets rarely do: short rates usually move more than long ones. The savings line assumes an account that tracks the bond yield you set. The growth stock can fall further than the 30-year bond because, in this model, its payouts sit about as far out as a 20-year bond's. It is there to show who is most sensitive, not to forecast a price.

11Common misreads

• "The Fed sets mortgage rates." It sets an overnight rate. US fixed mortgage rates follow the 10-year Treasury yield, which the market sets. In the last months of 2024 the Fed cut by a full point, and the average 30-year mortgage rate went up, from 6.09% in mid-September to 6.85% by late December, because the 10-year Treasury yield rose almost a point.
• "Rates are going up, so stocks must fall." Stocks rose through the hikes of 2004–06 and 2015–18. What hurt in 2022 was the speed, and the surprise.
• "Rate cuts are good for stocks." Depends why. Cuts into a recession have come with falling stock prices.
• "Rising rates are good for banks." For their lending margins, at first. Not for the bonds they already own.
• "Bond losses are only on paper." Only if you never have to sell before the bond matures.
• "Rates went up, so my bond fund is ruined." The fall in price comes with a higher yield from then on. Hold for about the fund's duration in years and the extra income roughly makes up the loss.