A few headlines caught our attention recently that you should be aware of too.

Mortgage rates topped 7% for the first time since January 2025 and a swatch of corporate bonds are now trading at levels not seen since bank foreclosures last dominated headlines. Project Jupiter, a Silicon Valley AI darling hit a wall and invoked force majeure. And a new Citi survey found inflation is now the top concern for the wealthiest American families ... many of whom are tight on cash too.

On the surface, these stories may not appear to have much in common. To us, they all point to the same question:

"What happens to financial assets if inflation and borrowing costs stay higher for longer than most investors expect?"

The cost of borrowing is showing up in housing

According to Freddie Mac, the average 30-year fixed mortgage rate reached 7.03% on September 24, compared with 6.30% a year earlier, the highest reading since January 2025.

For a family trying to buy a home, that difference is immediate. A higher rate can change the monthly payment enough to put a house out of reach. It can also discourage current homeowners from moving or refinancing.

But this is bigger than housing. Borrowing costs influence construction projects, consumer spending and the investments businesses are willing to make. When money remains expensive, plans that make sense at lower rates may need to be reconsidered.

That is why we watch mortgage rates as more than a real estate statistic. They are one visible measure of how the cost of capital is affecting the broader economy.

Stress may be building in commercial real estate and banking

Commercial real estate is where that pressure may show up next. Trillions of dollars in commercial property debt will have to be refinanced at today’s higher rates, and some properties are now worth less than the loans against them. Some analysts argue that lenders have been slow to recognize the resulting losses; if they are right, recognizing them could strain some banks’ capital.

Polymarket, which allows individuals to bet on almost any outcome, is now offering markets on bank failures unfolding in the near future. Who bears the losses when banks fail is not a hypothetical question. Former New York Fed President William Dudley addressed it in a recent article, arguing that an effective resolution regime should require a failing institution’s own investors to absorb losses (a bail-in) rather than government picking up the tab. Dudley’s words revisited how $20 billion in losses from the collapse of Silicon Valley Bank’s parent were born by the FDIC. The infuriating reality of this lies in how the vast majority of that bailout was beyond the scope of the FDIC. Designed to protect small savers, government chose to make mega-investors in FDIC whole, where the average large account was in excess of $4 million. This policy error protected the richest and drained 12% of the nation’s entire reserve account for just 0.19% of American deposits.

Remember: when rates rise, reality sets in and when bank stress rears its head, government protection is a subjective decision.

Even the AI boom has to contend with these costs

We often hear about AI infrastructure as if demand alone will carry every project forward. Oracle’s Project Jupiter data center in New Mexico is a reminder of what it takes to turn that demand into a working facility: substantial financing, reliable power, permitting and years of construction.

The project has drawn attention over its financing and infrastructure challenges. Oracle issued a force majeure notice to STACK Infrastructure, the Blue Owl Capital-owned developer of Project Jupiter, reportedly to protect itself from potential payments if the facility misses its planned 2028 in-service date. Reuters reported that the notice cited potential delays in securing power.   That does not mean the AI buildout is ending, or that this particular project will fail. It does show how sensitive large projects can be to delays and the cost of getting them built.

The scale of the obligations behind the AI buildout was also on display this week, when Reuters reported on Anthropic’ s confidential IPO prospectus, which it reviewed but which has not been publicly released. According to that report, Anthropic posted a net loss of about $42 billion. It also has about $518 billion in tech spending commitments ahead with partners including Google, Amazon and Microsoft. Apparently almost 25% of revenues come from two clients who can walk away from Anthropic at any time. These commitments stand against “only” $20 billion in cash. Commitments on that scale underscore how changes in rates and inflation expectations can grind any indebted project to a halt… and imagine what happens to Nvidia and other tech names who are counting on Anthropic buying $518 billion in product from them. After all, hyperscalers like Anthropic are the “rock-solid” future earnings the tech sector is levitating upon.

Inflation is often discussed in terms of groceries or gasoline. Investors also need to consider the rising costs of energy, construction and financing. Those costs can affect the returns on even the most promising investments.

Wealthy investors are watching inflation, too

Citi’s 2026 Global Family Office Report identified inflation as the top concern among the family offices it surveyed (63% ranked it first, up from 37% in 2025), followed by interest rates, financial-system stability, and market volatility.

We find that telling. Inflation affects far more than a household budget. It changes what future income is worth, what companies pay to operate, what bond investors demand in yield and what stock investors may be willing to pay for earnings. Citi’s survey also found that family offices are more likely to cut private credit than to add to it, with the report pointing to record default rates.

This insight will weigh heavily on future marks in private credit books and negatively impact private equity valuations in the future. Further, it suggests current private credit valuations may be fantasies.

If inflation proves persistent, investors may have to adjust assumptions they have made about rates and valuations. We do not know that this will happen. We do know that it is a scenario worth preparing for.

What history can—and cannot—tell us

The 1970s offer a useful example of how persistent inflation can change markets. Financial media often say that rising rates and inflation are bad for gold, which pays no interest. Less often discussed is that rising rates and inflation tend to compress valuations of financial assets. Consider the data: over the 1970’s, the S&P 500’s price-to-earnings multiple fell by nearly half, while gold rose roughly 15-fold, from about $35 to about $512 an ounce between the end of 1969 and the end of 1979.

We can’t treat history as a forecast. The economy, monetary system, and investment markets are different today. And comparing gold’s former official price of roughly $35 an ounce with its January 1980 peak of about $850 does not represent a straightforward investment return.

The lesson is broader: when inflation and monetary expectations change, the relationship between financial assets and physical assets can change substantially, too. Be clear with yourself and your advisors – If inflation persists, are you positioned properly?

Where does gold fit?

Gold is not a guarantee against losses.

Still, physical gold has a distinct place in a diversification conversation.

The question does not even need to be, “Are we about to repeat the 1970s?” It is just simply: “How would my portfolio hold up if inflation remains stubborn, and borrowing costs stay elevated?”

That is a conversation to have while you have time to weigh your options thoughtfully.

As we pen this article, sentiment is very poor in the metals, which contrarian investors sometimes treat as a potential buying signal. Consumer confidence fell 6.7 points in September to 81.9, its lowest level since 2014, according to the Conference Board, with higher prices and fuel costs cited.