* Index returns are total returns, including dividends.
WIDE AWAKE
In Washington Irving’s 1819 tale, Rip Van Winkle wanders into New York’s Catskill Mountains, shares a drink with a band of strange men playing ninepins, and lies down for a nap that ends up lasting twenty years. He returns to a village that no longer knows him, where he causes an uproar by declaring his loyalty to King George, unaware that a revolution has been fought and won while he slept. Even the sign above the village inn has changed: the king’s portrait has been repainted as “General Washington,” though the old face still shows through.
An investor who similarly dozed off in 2002 and wakened again this fall would find one number strangely familiar: the yield on the 10-year US Treasury, which has climbed past its 2007 peak to its highest level in roughly two decades. Like the king’s face beneath the paint, the number is recognizable, but the world around it is not, reshaped by a financial crisis, a global pandemic and over a decade in which borrowing cost almost nothing. Historian Edward Chancellor observed, “in the second decade of the twenty-first century, interest rates fell to their lowest level in five thousand years of recorded history.” Viewed through that lens, today’s rates are less a shock than simply a return to levels we have seen before, in a world that has been rebuilt around cheap money.
Much of that change happened this quarter. The US 10-year yield rose from 4.2% to 5.3%, and stubborn inflation and higher oil prices led the Federal Reserve to raise rates in September for the first time since 2023. Stock markets told two different stories: the S&P 500 rose about 2%, carried by a small group of very large companies, while the Dow Jones Industrial Average slipped – more than half of all US stocks were at least 20% below their June highs by late September, a drop that would normally be considered a bear market. For anyone who didn’t rely on the market’s biggest contributors in Q3, the quarter felt rougher than the headlines suggested.
When money is practically free, borrowing becomes the easiest way to grow. Over the past fifteen years, families took on larger mortgages, companies borrowed to buy back shares and fund acquisitions, and governments ran deficits that cost very little. The portion of the US government debt held by the public has climbed from 35% in 2007 to roughly 100% today. As that debt comes due, it must be refinanced at today’s prices. US mortgage rates are back above 7%, Canadian homeowners renewing loans taken out during the low-rate years face noticeably larger payments, and the US government now spends more than $1 trillion a year on interest. Borrowed money works like a lever: cheap debt lifts everything it touches; as rates rise, that same lever grows heavier. None of this happens overnight, but over time it leaves less money for spending, hiring and investment.
“Give me a place to stand, and with a lever I will move the whole world.” – Archimedes
For our clients, the crucial question is what higher rates mean for the stocks and bonds they own. Bonds work like a seesaw: when yields rise, the prices of existing bonds fall, because newer bonds pay more and are thus more attractive. The longer a bond has until it matures, the bigger that swing, which is why we have kept our bond portfolio’s duration shorter than the benchmark. That choice softened the impact of this quarter’s rise in yields, and it leaves us with maturing bonds we can reinvest at today’s higher rates. We have begun doing so with the Oracle bond discussed later.
Stocks feel the change, too. When a government bond pays over 5%, a utility or real estate stock paying a 3% or 4% dividend loses some of its appeal. Those sectors were among the quarter’s weakest, while banks and insurers also lagged as investors weighed what higher rates mean for their balance sheets. In the long run, though, higher yields are good news for savers, because every dollar we reinvest now earns more than it did a few years ago.
History shows us that when rates rise quickly, they can expose risks that remained hidden while money was cheap. In 1994, the Federal Reserve doubled its key rate from 3% to 6% in about a year. That spring, a little-known accountant named John Moorlach ran for Treasurer of Orange County, California, warning that the long-serving incumbent, Robert Citron, had borrowed heavily, betting that rates would stay low. Citron had delivered years of impressive returns, and so voters re-elected him. By December, however, the county’s investment pool had lost about $1.6 billion, and Orange County filed what was at the time the largest municipal bankruptcy in American history.
We share this story not as a dire prediction, but as a reminder of why we pay attention. So far this cycle, nothing comparable has happened, and the Federal Reserve still describes an economy expanding at a solid pace. Higher yields may simply be the cost of a stronger economy with stickier inflation. Rip Van Winkle, for his part, made peace with the new world he found himself in and spent his later years telling his story to anyone who would listen at the village inn. We intend to adapt as well by owning durable businesses with sound balance sheets and by being ready to act when fear creates bargains. We have seen these interest rates before; what matters is meeting them with clear eyes in a world that has changed around them.
CROSSING THE BORDER
For many years, the income portion of our clients’ portfolios was built largely from Canadian bonds. The reason was simple: a Canadian bond pays Canadian dollars, so clients relying on that income never had to worry about the exchange rate. That strategy has served us well, but the Canadian bond market is small and today it often pays meaningfully less than its US counterpart. On October 5, for example, a 10-year Government of Canada bond yielded 3.96%, while a 10-year US Treasury yielded 5.31%, a gap of more than 1.3 percentage points. We do not hedge our US dollar exposure, so if the US dollar weakens against the loonie, part of the extra income is lost when converted back, and if it strengthens, clients gain. However, we believe a gap this wide more than compensates for that risk, and so we have begun, selectively, to look south of the border.
Our first such foray is Oracle Corporation’s 4.125% bond1 maturing in 2045. Oracle was founded in 1977 by Larry Ellison, Bob Miner and Ed Oates, who named their company after a project for their first customer, the CIA. Its software became a standard way for organizations to store and manage their data, and today it remains embedded in the systems of banks, hospitals and governments around the world. Over the decades, Oracle added business software through acquisitions such as PeopleSoft, NetSuite and the health records company Cerner. Its newest business arm rents computing power to artificial intelligence companies from its own data centres: that segment grew 121% last quarter to US$7.4 billion. Customers have signed contracts worth US$664 billion that Oracle has yet to deliver, more than seven times the revenue it expects this year.
That growth, though, has come at a price. Oracle has borrowed heavily to build its data centres, with US$125 billion of borrowings on its balance sheet, and last quarter it spent US$5.4 billion more on construction and equipment than its operations brought in. In July, S&P lowered its credit rating to BBB-, the lowest rung of investment grade, citing the spending and the large share of the backlog tied to a single customer, OpenAI. Those worries are why the bond trades at roughly 62 cents on the dollar and yields about 8.1% to maturity, nearly double the 4.3% the Government of Canada pays to borrow over the long term. We take comfort, however, in several things: Oracle’s core database and software business remains highly profitable; management has stated its intention to keep an investment-grade rating; and it raised US$20 billion by issuing new shares last quarter, money that ranks behind bondholders and cushions our claim. As lenders, we do not need Oracle to win the AI race. We only need it to pay its interest and return our principal in 2045.
We believe their long-term business means that they will. Companies rarely replace the database at the heart of their operations because doing so is costly, slow and risky. This makes Oracle customers extremely loyal. If its AI contracts are delivered as signed, cash flow should recover, and S&P itself expects Oracle’s free cash flow to turn positive by 2028. Should AI demand cool, Oracle could slow its spending and lean on a software franchise that has generated profits for decades. With the 10-year US Treasury yield at its highest level since 2002, bonds like this let clients lock in a rate of return that has been hard to find for most of the past two decades. The margin of safety rests not on Oracle’s AI ambitions, but on the durable business beneath them.
HAILING A BARGAIN
In the equity market, the robotaxi scare created a rare bargain in one of the high-quality businesses we now own, a company many investors had started avoiding for the wrong reason: Uber Technologies2 (NYSE: UBER). Uber has gradually grown from merely a ride-hailing app into an indispensable tool for getting around, eating and shopping. It owns no cars, employs no drivers and runs no restaurants. Instead, it operates a marketplace that matches demand with supply in real time. More riders attract more drivers, more drivers shorten wait times, and shorter waits attract more riders. With roughly 300 million trips taken each week and about three-quarters of US rideshare spending, Uber runs a network no newcomer could ever afford to rebuild city by city.
Under Dara Khosrowshahi, CEO since 2017, a company once famous for burning cash now generates more than US$10 billion of free cash flow per year. He recently cut 10% of the workforce, saying Uber’s results let it keep investing “from a position of strength”, telling us exactly the kind of operator running this business. Management has also taken a stake: The CEO bought US$10 million of stock, the COO US$5.3 million, and the CFO US$1.6 million. Uber’s balance sheet is investment grade, and share buybacks are the default use of spare cash. The runway also remains long: fewer than one in ten consumers in Uber’s US suburban and rural markets used the service last year compared with more than half in big cities, and those sparser markets are growing 1.5 times faster. Uber’s grocery division alone has become a US$15 billion business, expanding about 40% per year.
The company’s share price dipped primarily over concerns surrounding autonomous vehicles. As Waymo (Alphabet’s autonomous driving technology company) expanded, it ended its exclusive arrangement with the Uber app in Atlanta and Austin and left the platform entirely in Phoenix. The market began to fear that robotaxi operators would bypass Uber and sell rides directly. But the hotel industry ran this experiment a decade ago. Tired of paying commissions of 15% or more to Expedia and Booking.com, Hilton launched “Stop Clicking Around” in 2016, the largest marketing campaign in its history at the time. Marriott, Hyatt and Starwood soon followed with direct-booking discounts of their own. Strangely, even though the hotels owned the rooms, the brands and the loyalty programs, the travel sites kept growing, because a room left empty tonight is revenue lost forever. In 2025, Booking Holdings alone sold more than 1.2 billion room nights, up 8% from the year before. Owning the asset, it turns out, is not the same as owning the customer.
We believe the market has drawn the wrong conclusion once again. A driverless car is an expensive asset that earns its keep only when carrying passengers, so the network that keeps it busiest wins, and Uber reports roughly 30% higher utilization for vehicles on its platform. Even at the 500,000 weekly rides Waymo reported this spring, its entire service handled less than 0.2% of Uber’s weekly trips, and robotaxis alone will struggle to meet demand on the busiest nights in the densest cities. Uber is also willing to own more of its supply when it makes sense, committing about US$10 billion to bring self-driving cars onto its network through partners. Autonomous rides are already live on the app in seven cities, with as many as fifteen planned by year-end. In a driverless world, we believe owning the demand matters more than owning the car.
A dominant, well-run business temporarily rendered cheap by a misunderstood threat is precisely the kind of opportunity we look for, so we bought while others were heading for the exit. At the time, Uber traded at 15.3 times 2027 earnings, despite 20%-plus bookings growth, an investment-grade balance sheet and years of runway ahead. We do not try to predict who will win the race to autonomy or when regulators will approve driverless cars in every city. We simply buy a solid business when fear makes it cheap and then hold it for as long as it keeps earning our confidence.
Thank you for your continued trust and support.
The Evans Team