October 14, 2025
I hope you all had a nice Thanksgiving weekend. I recognize that we have our share of problems here in Canada but when we look around the world, I'm certainly thankful that my parents made the courageous decision to move here from Malawi 50 years ago this month. I'm sure I'm not alone in giving thanks this weekend for being Canadian.
In this edition, I wanted to touch on two main topics: the US government shut down; and the opportunity and concern around artificial intelligence.
US Government Shut Down
In the US, a government shutdown occurs when the funding legislation required to finance the government is not passed before September 30th, the end of the fiscal year. It should be noted that most shutdowns happen because of budget disagreements, but the current one is different. Republicans now control a trifecta of the House, the Senate, and the Presidency—so their goal is to stay aligned with President Trump and what he wants. The Democrats are drawing a line to protect some core programs that were enacted under Democratic leadership, including the Affordable Care Act and Medicaid.
A key grievance for Democrats relates to an impasse last March when the government nearly shut down. The shutdown didn’t occur because Republicans promised to seek compromise on several key disagreements in exchange for Democrats supporting a short-term continuing resolution. But these disagreements were never resolved or even seriously considered.
A shutdown does not affect the government's ability to pay its debt to bondholders and therefore, does not have a direct impact on the government’s borrowing costs or creditworthiness. Treasury interest payments and Social Security continue to be paid, but millions of federal employees go unpaid during the shutdown. They are, however, guaranteed back pay once funding resumes.
There have been 21 government shutdowns since 1976, the longest being 35 days during President Trump's first term. It seems like both sides have dug their heels in this time so it’s possible that this one could be longer. The stock market for the most part has ignored shutdowns. Going back to 1976, 50% of the time the stock market is up during a shutdown, 60% of the time it’s up three months after the shut down, and 70% of the time it's up six months after the shutdown. It's possible that this is the reason that the markets have so far shrugged it off. However, the longer this goes, it may weigh on consumer confidence and ultimately earnings from companies.
During the shutdown government agencies that report employment and inflation statistics are not open. Going into the shutdown, unemployment rates were worsening, and inflation was slightly trending upwards. The hope is that those two numbers don't get materially worse by the time they do report and surprise the market. I can't help but think that the economy will weaken, forcing the FED to cut rates when inflation could be rising. We continue to own short duration bonds and are prepared for increased volatility.
Growth of AI
Artificial Intelligence has rapidly evolved into a transformative force impacting our lives. From personal recommendations to autonomous vehicles, to advanced healthcare diagnostics, it's an integral part of our daily lives. Tools such as Chat GPT and Google Gemini are not just sites to access data. These highly complex models can recognize patterns, help make decisions based on a data set, and more importantly, keep learning.
The computational power that is required for AI to function is housed in data centers. Some that are owned by Google or Meta can be over 1,000,000 square feet and house servers, storage systems, networking equipment, power infrastructure, and cooling systems that operate 24 hours a day, 7 days a week. The size and scale of the investment that is being made in data centers globally is unprecedented. According to the International Energy Agency (IEA), $500B was spent in 2024 in building data centers.
An immense amount of power is required to operate data centers. In 2024, Data Center energy consumption amounted to 1.5% of global energy consumption. This is expected to increase to 3% by 2030. Three percent is equivalent to what Japan consumes in an entire year. To look at it differently, one data center by Google consumes enough energy to power 100,000 households. Currently there is a data center under construction that would need 20 times more or the equivalent of powering 2,000,000 homes. Most of these Data Centers are in various clusters around the world with 45% located in the US, 25% in China, and 15% in Europe.
Electricity grids are already under tremendous strain in many countries. It's very possible that the growth of data centers is going to be limited by their access to energy. The International Energy Agency estimates that 20% of planned data centers may be delayed for this reason. In addition, 50% of planned data centers are near clusters of other data centers which can put further pressure on the electric grid in that region.
I'm convinced AI is here to stay. We will see companies increasingly use it to improve productivity and efficiency and I think it will support incredible advancements in healthcare. Will everyone that is investing in data centers and developing their own AI make money? I'm not sure. The dot com era created a lot of casualties but from that we got Google, Amazon, and Meta. I fear that AI companies may face the same fate. Regardless of which AI company, how to power it will be the most important question in the coming years.
Rather than try and guess which AI company will succeed, I think investing in energy infrastructure, traditional and alternative energy will be a sustainable opportunity. Making investment in companies that make the electrical grid stable and more efficient will be beneficial to any company engaged in AI. Nuclear energy has great promise but there's a long lead time in making that a less viable option today but cannot be discounted. I'm sure novel ideas will emerge in battery and storage that also might help solve the problem.
In September my wife Kris and I spent time in Tuscany, Italy and Croatia. It's quite liberating to go on vacation outside of the "school schedule." Both places were fantastic, and I would certainly go back to both. It's interesting when you travel now and say that you are Canadian….you tend to get a lot of apologies - Europeans apologizing for calling me an American and Americans apologizing to me for their country’s behavior. Once again, a reminder of the gratitude I have to my parents for their bravery in coming here and to the country I call home. We aren’t perfect but when we look at the alternatives, I feel fortunate. At this time of thanksgiving, I hope you feel the same.
Ashit