Data Center Insurance, AI Targets Seniors, Housing Rate Problem, Apple Under Ternus, Fed Rate Hike, Military Tax Exemption & More

Another big cost for data centers: insurance!

You always hear about the obvious expenses involved in building data centers: incredibly expensive chips, massive energy requirements, cooling systems, and the cost of constructing the facility itself. But there’s another major expense that doesn’t get nearly as much attention: insurance.

Once completed, a large data center can be worth anywhere from $20 billion to $50 billion or more. Meta’s data center project in Northeast Louisiana, for example, is expected to cost more than $50 billion and will house thousands of servers along with extremely valuable power and cooling infrastructure.

In the rush to build data centers, many hyperscalers and other companies have located facilities in areas where real estate is relatively inexpensive. The problem? An estimated 40% of U.S. data centers are located in tornado-prone areas. I guess some companies didn’t think quite that far ahead.

For insurance companies, this could become a massive market. The potential insurance premiums associated with data centers could eventually be 4–5 times the current global aviation insurance market.

Data centers could become some of the most valuable insured assets in the world.

And insuring them won’t necessarily be cheap. An average cost of around $20 billion for a data center will exceed the insurance cost for bridges, tunnels, and skyscrapers.

There’s very little historical data showing how these massive facilities will perform during extreme weather events such as tornadoes and hurricanes. Companies also want coverage for risks like equipment failures, business interruption and even terrorism, all of which can add significantly to premiums.

So, there’s yet another enormous cost that hyperscalers must absorb on top of the hundreds of billions of dollars they’re already spending to build out AI infrastructure.

The big question I still have is: How are companies ultimately going to recover all of these costs?

The AI infrastructure buildout is becoming more expensive by the day. I’m sure glad we own an insurance company in our portfolio.

AI has taken scamming seniors to a whole new level

The Federal Trade Commission estimates that roughly $196 billion was lost to fraud, with nearly half of that or about $82 billion coming from people over the age of 60. It’s now estimated that roughly 37% of fraud involves artificial intelligence.

The cases are frightening because scammers can now use AI to identify targets, make their pitches more convincing, execute scams faster, and target thousands of people at a scale we’ve never seen before.

With AI, scammers can create what appears to be legitimate proof of identity, including realistic passports, driver’s licenses, websites, and other documents. Gone are the days when scams are easy to spot because of grammatical errors, thick accents on phone calls, or obviously fake videos.

And it goes far beyond dating scams.

Imagine getting a phone call saying that your daughter witnessed a major drug deal and has been taken by traffickers unless you come up with $20,000. That sounds unbelievable, until they put your daughter on the phone, and you hear her voice, panicked and pleading for help. The problem? It isn’t actually your daughter. It’s an AI-generated voice that sounds just like her.

So how do you protect yourself from AI fraud?

One major red flag is a demand for cryptocurrency. Another is the pressure to keep the situation secret. Scammers often tell you not to contact anyone because they want to isolate you and prevent someone else from recognizing the scam.

Scammers are becoming so bold that some will even show up in person pretending to be federal agents with fake badges and credentials. Remember this: The federal government will not demand that you hand over cash, cryptocurrency, gold bars, or other assets simply because someone claims to be a federal agent. If someone claims to be a federal agent and wants to meet at your home, you can always ask to meet at their office instead. A legitimate agent should have no problem with that.

If someone is pressuring you to act immediately, keep it secret, and hand over money or valuables, stop and verify who you are dealing with before doing anything. AI is making scams more convincing than ever. The best defense may be simply slowing down long enough to question what you're being told.

The Housing Market Has a Rate Problem

Existing-home sales fell 2% in August to an annualized pace of 3.98 million homes, the lowest level since June 2025. Sales are now down 1.2% from a year ago and have remained stuck around the 4 million annualized level for much of the past few years. Historically, a more normal housing market has been closer to 5.2 million existing home sales annually.

At the same time, we are finally seeing more inventory. There were 1.62 million existing homes for sale in August, up 5.9% from a year ago and the highest level since 2019. That represents about 4.9 months of supply, which is the highest level in over a decade. This gives buyers more choices and more negotiating power than they've had in quite some time.

But here's the problem: home prices aren't falling. The median existing home price rose 1.6% from a year ago to $429,100. So, buyers are dealing with a combination of high home prices and high borrowing costs.

And I think one of the most important things to watch is the 10-year Treasury. A lot of people assume mortgage rates are primarily determined by the Federal Reserve. That's not really the case. The 30-year fixed mortgage is heavily influenced by longer-term Treasury yields, particularly the 10-year Treasury.

The 10-year Treasury recently approached 5%, reaching its highest level since 2023. As the 10-year moves higher, mortgage rates generally move higher as well. The 30-year fixed mortgage climbed over 7%. The last time the 30-year fixed mortgage was at 7% or higher was May 26, 2025, when it reached 7.02%.

That's a big deal for affordability. A buyer who could comfortably afford a $500,000 mortgage at 5.5% has a substantially different purchasing power at 7%. Higher rates can push buyers to either purchase a less expensive home, put more money down, or simply stay on the sidelines.

I don't think the current data points to a traditional housing crash. We don't have the same combination of excessive speculation, massive overbuilding, and widespread distressed selling that characterized 2008.

Instead, we're seeing a housing market that is essentially frozen by affordability. The big question going forward isn't just what the Fed does. Watch the 10-year Treasury. If the 10-year moves meaningfully lower, mortgage rates could follow, and housing activity could improve. But if the 10-year stays near 5% or moves above it, don't be surprised if 30-year mortgage rates remain around 7% and housing continues to struggle.

For buyers, the good news is that inventory is improving, and negotiating power is coming back. The bad news is that the cost of financing remains extremely high. Housing may finally be shifting from a seller's market toward a buyer's market, but that doesn't necessarily mean homes are getting cheaper.

How will Apple stock do under new CEO John Ternus?

Apple has turned a new page in its history books with Tim Cook stepping down after 15 years at the helm. Cook did a tremendous job managing the business and growing the stock mainly by improving the supply chain to overseeing the launch of products like the Apple Watch and AirPods. During his 15-year tenure, Apple stock increased roughly 2,680%, which is a phenomenal return for investors.

This reminds me of another major business story that was incredibly successful: General Electric under the leadership of Jack Welch. Welch ran GE for 20 years, from 1981 to 2001. During his tenure, GE stock increased roughly 4,000%. Yes, that was over five more years than Cook's tenure at Apple, but I doubt Apple would have produced another 1,300% return if Cook had stayed for five additional years.

At the end of both CEOs' tenures, however, their companies had one major similarity: very high valuations. The comparison is a little scary. When Jack Welch left GE in 2001, the stock was trading at roughly 30 times earnings. This was during the dot-com bubble, when the S&P 500 was trading at about 36 times earnings.

When Tim Cook left Apple, Apple was trading at roughly 37 times earnings, while the S&P 500 was around 29 times earnings. Those are very high valuations, especially considering we're also experiencing what could be an AI bubble.

When Welch stepped down in 2001, his hand-picked successor, Jeffrey Immelt, took over. From 2001 through 2017, GE stock generated a total return of roughly -27%.

Yes, investors actually lost 27% over 16 years.

Apple is an incredible business with an enormous ecosystem, tremendous cash flow, and some very valuable brands. But Ternus is taking over at a time when expectations are extremely high.

Apple does have some interesting things coming. The company announced a new foldable iPhone on September 9, along with an updated Siri and other AI improvements. But I personally didn't think the event was a big deal.

The bigger question is whether consumers are really willing to spend around $2,000 on a new, foldable iPhone. And perhaps even more importantly, will AI actually create the upgrade cycle that Apple investors are expecting?

There are also potential headwinds from significantly higher chip costs, which could push up the prices of iPhones and other Apple devices.

I remember talking to people years ago about potentially selling some GE stock and diversifying. The response I often heard was, "Don't touch my GE stock." I'm hearing very similar stories about Apple today. The big question is whether history repeats itself with Apple like it did with General Electric.

I don't know if it will. But I am pretty confident about one thing: Apple shareholders are very unlikely to see anything close to the returns they experienced under Tim Cook over the next 15 years.

A Fed Rate Hike Now Looks Inevitable

We received two important inflation reports this past week, and my takeaway is pretty straightforward: inflation itself still doesn't look overly problematic, but there are some areas that have me concerned.

The Producer Price Index showed wholesale prices increased 0.4% in August. Producer prices are now up 5.4% from a year ago, which was slightly above expectations. Then on Friday, we got the Consumer Price Index, which showed consumer prices increased 0.4% in August and are up 3.4% over the last year.

Core CPI, which excludes food and energy, increased 0.3% for the month and 2.4% over the past year. Those numbers aren't exactly what the Fed wants to see, but I also don't think we're looking at runaway inflation.

What concerns me more are two specific areas: diesel prices and the enormous AI buildout. Diesel is a major problem because it works its way through virtually every part of the economy. The national average price of diesel has now climbed above $6 a gallon for the first time ever and is up from around $3.70 this time last year. In California, diesel is now around $8 a gallon.

That's not just a problem for people who drive diesel trucks. Higher diesel prices increase the cost of transporting goods, operating farm equipment, shipping products, and ultimately getting those products onto store shelves.

The other issue I'm watching is the AI infrastructure buildout. We're seeing an extraordinary amount of capital being committed to data centers, chips, power generation and related infrastructure. I'm concerned that if companies continue spending at this pace, it could create further upward pressure on costs, particularly for energy, construction, memory, and financing.

So, while I don't think the current inflation numbers are particularly alarming by themselves, I don't think we should ignore the inflationary pressures building underneath the surface.

The market has already pushed the probability of a hike extremely high following these inflation reports and on Friday odds were close to 90%. At this point, if the Fed doesn't hike, it risks surprising the market and potentially damaging its credibility.

The bigger question isn't whether they hike next week. It's what they do after that.

If diesel prices continue to rise, and the AI buildout continues to put pressure on energy and other resources, the Fed may have a much harder time getting inflation back to its 2% target than the headline numbers currently suggest.

Inflation isn't my biggest concern right now. It's whether inflation is stuck and how do we get it lower from here.

Financial Planning: Tax Exemption for Retired Military

California military retirees may now be eligible for a valuable new state tax break. Through tax year 2029, California allows qualified taxpayers to exclude up to $20,000 of military retirement pay from California for taxable income. The provision applies to retirement pay received from the federal government for service in the uniformed services, as well as qualifying annuity payments from the Department of Defense Survivor Benefit Plan.

To qualify, the taxpayer’s federal adjusted gross income must generally be $125,000 or less for single filers or $250,000 or less for married couples filing jointly. This is an income exclusion, not a $20,000 tax credit, meaning it reduces the amount of income subject to California tax. For someone in the 9.3% California tax bracket, the maximum potential savings is $1,860. Eligible taxpayers claim the adjustment when filing their California return, generally through Schedule CA. Income planning is especially important for military retirees because there is no phaseout, if income exceeds the applicable limit, the entire exclusion is lost.

Non-alcoholic beer may actually contain alcohol

Non-alcoholic beer sales in the United States were up 14% year-over-year for the period ending July 18. For the quarter ending June 27, non-alcoholic beer sales reached nearly $235 million.

But if you’re trying to completely avoid alcohol, there’s an important distinction to understand. “Alcohol-free” means no alcohol, while “non-alcoholic” beer can contain up to 0.5% alcohol by volume.

Before you get too upset with the beer companies, keep in mind that some everyday foods can contain similar amounts of naturally occurring alcohol. A ripe banana can contain around 0.2%, and very ripe bananas with brown spots can reach roughly 0.4% to 0.5%. Bread can also contain small amounts of alcohol.

Making completely alcohol-free beer can be more complicated and expensive. Some brands avoid fermentation altogether, which is the process that naturally produces alcohol. Others start with a traditional beer and then remove the alcohol through processes such as vacuum distillation or filtration. Another option is diluting the original formulation to reduce the alcohol content.

Some brands marketed as 0.0% alcohol include Budweiser Zero, Michelob Ultra Zero and Coors 0.0%. So, if you want to get a little buzz without anyone knowing… apparently, a couple of ripe bananas and some bread might be the way to go.

Have you heard the term “everywhere millionaires”?

You often hear about a tech company that was sold for $100 million or $200 million, but you don’t hear nearly as much about the people who worked hard and built a small business worth $5 million, $10 million, or even more.

These people are what as is known as the “everywhere millionaires.” They could be the coach on your child’s soccer team, the person sitting next to you at your favorite restaurant, or the owner of the local plumbing, HVAC, construction, landscaping, or manufacturing business.

In the United States, there are roughly 30,000 businesses generating between $10 million and $50 million in profits. These aren’t necessarily technology companies. They could be businesses that have to move something, fix something, install something, manufacture something, or provide a service.

They may not be exciting businesses that make headlines, but they can create tremendous wealth. And for someone willing to work hard sometimes 60 or 70 hours a week and stay patient, building a business can provide financial stability and potentially generational wealth.

You don’t necessarily need a degree from Harvard or even a college degree. You do, however, need to understand the fundamentals of business. Taking courses in marketing, finance, accounting, and management can be incredibly valuable.

Interestingly, research has found that the relationship between academic ability and becoming a successful business owner isn’t nearly as strong as you might think. People with very high SAT scores became successful business owners only about 1.3 times as often as those with average scores.

In some cases, joining the military may even provide a more useful foundation for entrepreneurship than a traditional four-year college degree. You can learn teamwork, organization, planning, leadership, training, discipline, and perhaps most importantly how to persevere when things get difficult.

So, what’s the best way to start a business? Find something that isn’t done well.

Find something you believe can be done better. Find a problem customers have that nobody is solving or a service people would gladly pay more for if it were done properly.

I want to reiterate that I’m not saying this is easy. Building a successful business takes years of hard work, patience, persistence, and plenty of failure along the way. But you don’t have to invent the next Facebook or build the next AI company to become wealthy. Sometimes, the best opportunities are hiding in the most ordinary businesses right in front of us.