Blueprint for $10,000 Gold

Why the recent drop in gold is a gift for the taking...

If your head is spinning, you’re not alone. We can’t recall another time when so much was happening in such a short span. The implications for the U.S., and, indeed the world, are profound. For now, the most we can do is to point to the most important of these events, indicate how they are interrelated, and suggest what the potential implications are. We will publish additional blogs as greater clarity is provided.

First up, the Fed. About a week ago—though it seems a lot longer—the Fed began a new era under the leadership of Kevin. In our opinion, he followed a predictable script. It didn’t surprise us that Warsh—appointed by Trump, who has pushed hard for lower interest rates and a lower dollar—nonetheless took a seemingly hawkish stance. Warsh could not simply have jumped in and immediately lowered interest rates with no rationale. He would have looked like a puppet and lost all credibility. He had to set the stage.

And it seems that Warsh succeeded in coming across as hawkish, as evidenced by the shift in interest rate futures. Expectations now project one or two Fed rate hikes by yearend. Prior to the Fed meeting, the expectations were for no hikes. The seemingly hawkish turn is at least partially responsible for the recent weakness in precious metals and strength in the dollar. But we think that ultimately the opposite will occur. We expect the next interest rate move will be down, not up, despite what Warsh said.

Driving the about-face will be a combination of three factors.

First will be unexpected economic weakness stemming from a near-record-low savings rate of 3%. Behind this low savings rate is consumer spending that has significantly exceeded consumer income. Consumer spending represents about 70% of economic activity. And just today, the release of May new home sales data showed a stunning 10% shortfall relative to expectations. Housing, of course, is a major component of consumer spending.

The second factor relates to Warsh’s announcement of five task forces, including ones dedicated to finding an alternative way of measuring inflation along with other measures of other economic data. Task force members will be appointed by Warsh. Currrently, most of the data that the Fed uses comes from government statistical agencies like the Bureau of Labor Statistics and Bureau of Economic Analysis. The task forces are charged with finding new ways to compile the data. If you control the data, you control the narrative and can point to the data to back up your story.

If Warsh is to march in accordance with Trump’s wishes, it is nearly a sure bet that one of the alternative measures of inflation will resemble what’s known as the “true inflation measure.” This is a real-time number composed of 15 million components—vs. the 80,000 that make up the consumer price index (CPI). It is also measured in real time vs. the month-long lag in current inflation indicators. This likely new measure will show inflation to be significantly less than the CPI. As we write this, the true inflation indicator is 1.85% compared to a CPI reading of 4.20%. So, “true inflation” would indicate a need to cut rates, while the only question the Fed would ask if going by the CPI would be, “how much should rates be raised?”

Finally, if you wondered if we’re forgetting about AI, we have not. The third and perhaps most important factor is that the current level of spending on AI capital investment—spending on data centers and everything else necessary for the AI buildout—is unsustainable. According to David Rosenberg, the esteemed former chief economist at Merrill Lynch, these expenditures as a percentage of economic activity are the highest in the country’s history—higher than the dot com bubble at the beginning of the century. And if the highly credible Rosenberg is to be believed, higher than expenditures on the railroads in the 19th century. There are multiple compelling reasons, even if you buy into all the AI hype, to forecast a decline, and very likely a dramatic decline, in these expenditures.

The major reason is that there is a shortage of resources to sustain them. Data centers are voracious for electricity, water, and even land. Add to these “brick wall” constraints the fact that large language models (LLMs) are foundational to these massive expenditures. About 80% of LLM revenues are generated by two private companies— Open AI and Anthropic—each of which is losing billions of dollars. While both plan to go public this year, there is work cut out for them, to say the least. No matter how you spin it, they are being tasked with convincing investors to pony up about a trillion dollars for each of these money-losing companies.

Moreover, even assuming the IPOs go off without a hitch, both are going to be left with a mountain of debt. It is not an exaggeration to say that the current combined projected revenues of both—about $200 billion, which is an extremely liberal estimate—will be asked to support both past and future capital expenditures of trillions of dollars. Even a small disappointment could be devastating, while a major disappointment would be catastrophic.

The question that will soon face the Fed won’t be whether or not to raise interest rates—it will be how fast can they cut them. It is hardly far-fetched to expect a reemergence of quantitative easing in the not-too-distant future.

So how will this play out?

The best model we have is 2008. One implication from this model relates to gold, and it suggests that as soon as there is an interest rate cut, gold will blast off. Prior to that, levels as low as $3,500 can’t be ruled out, but following that same model, the high over the next two to three years should be into five digits, taking the metal above $10,000.

---