A Note on Format
This is MCI's first macro analysis. Every prior piece has been a company. This week is different on purpose. The week of August 17 was dominated not by earnings, not by a Fed decision, not by a product launch, but by a violent selloff in long-dated U.S. Treasury bonds that dragged equities lower, sent gold and crypto sharply higher, and raised a question that every investor in every company in this blog needs to answer: what happens to stock valuations when the risk-free rate moves violently against them?
This is not a bond fund pitch. It is an explanation of why a move in 30-year Treasury yields from 4.5% to 5.28% in a single week matters more to the stocks in this blog than most quarterly earnings reports. Understanding that mechanism is the most important macro concept MCI has not yet covered.
What Happened This Week
The week of August 17 produced one of the most disorderly moves in the Treasury market in years. The 30-year yield finished the week at approximately 5.28%, its highest level in decades. The 10-year yield ended near 4.74%, at the top of its 52-year range. The move was not driven by a single catalyst. It was driven by a confluence of four forces arriving simultaneously.
First: supply. The U.S. federal deficit is running at approximately 7% of GDP. The Treasury Department must issue enormous quantities of new debt each quarter to fund that deficit. When the market must absorb more supply than it expects, buyers demand higher yields to compensate for the risk of holding long-duration government bonds. This week saw a particularly heavy corporate issuance calendar running alongside Treasury auctions, creating a supply wall that pushed prices down and yields up.
Second: term premium. After years of suppression through quantitative easing, the term premium, the extra yield investors demand for holding longer-duration bonds rather than rolling over short-term bills, has been reasserting itself. A term premium returning to its historical average of 1 to 1.5% on the 10-year note, on top of an expected short rate of 3.5 to 3.75%, implies a 10-year yield of 4.5 to 5.25%. That is exactly where the market went.
Third: AI capex inflation. This is the connection that most Treasury market analyses miss. Hyperscalers are spending $800 billion in aggregate AI capex in 2026. That spending requires copper, steel, aluminum, power infrastructure, and land. It is adding genuine demand to the real economy that is partially inflationary. The same AI buildout that makes Nvidia's revenue double and Microsoft's Azure accelerate is also putting upward pressure on construction costs, equipment prices, and energy infrastructure that flows through the economy into the price indices the Fed monitors.
Fourth: Warsh uncertainty. Before Jackson Hole clarified his position, the market was genuinely uncertain about what new Fed Chair Kevin Warsh's reaction function was. That uncertainty adds a risk premium to long-dated bonds when the inflation picture is already uncomfortable.
Treasury Secretary Scott Bessent announced on Wednesday of this week that the Treasury would at least double its long-dated buyback operations from $2 billion to at least $4 billion per issue starting in early September. Yields crashed on the news. Then snapped back immediately when strategists correctly noted that buybacks reshuffle the maturity profile of government debt without reducing the total supply the market must absorb. The relief lasted one session. That is not a policy solution. It is a signal that the Treasury is aware of the problem and does not yet have a structural answer.
| Metric | Value |
|---|---|
| 30-year Treasury yield (week close) | ~5.28% (multi-decade high) |
| 10-year Treasury yield (week close) | ~4.74% (top of 52-year range) |
| U.S. federal deficit | ~7% of GDP |
| Historical 10-year term premium | 1.0 to 1.5% |
| 2026 hyperscaler AI capex (aggregate) | ~$800B |
| Bessent buyback increase | $2B → at least $4B per issue, starting early September |
Why This Matters for Every MCI Portfolio Stock
Every valuation in this blog rests on a discount rate. The discount rate determines how much a dollar of future earnings is worth today. When the risk-free rate rises, the discount rate rises, and the present value of future cash flows falls. The math is not complicated. The implications are large.
Consider what a move from 4.0% to 5.0% in the 10-year yield does to a stock trading at 30x forward earnings. At 4.0% risk-free, a 30x multiple implies investors are willing to accept a roughly 3.3% earnings yield (1/30) for a business with a modest risk premium. At 5.0% risk-free, that same 30x multiple implies investors are accepting a negative risk premium relative to a risk-free Treasury. The math requires the multiple to compress toward 20 to 22x to restore a reasonable risk premium. A compression from 30x to 22x is a 27% decline in the stock price with no change in earnings.
This is not a theoretical exercise this week. It is live market pricing. The S&P 500 fell across the week as yields rose. Technology stocks with the highest multiples fell the most, because the duration of their cash flows is the longest. A company like Palantir at 75x forward earnings, or Apple at 36x, or Nvidia at 25x on next-year estimates, carries more interest rate sensitivity than a company like ExxonMobil at 12x, because more of the value is embedded in cash flows years into the future. When the discount rate rises, those distant cash flows are worth less today.
The companies least exposed to this dynamic in the MCI portfolio are the ones generating the most current free cash flow relative to their market cap. Apple at 104% ROIC generating $129 billion in annual free cash flow is more insulated than Palantir, which is still in the phase where most of the value resides in future earnings growth. ExxonMobil at 12x forward earnings with a 3% dividend yield and commodity-linked revenues is actually partially benefiting from the same inflation that is driving yields higher, because oil prices tend to rise with inflation expectations.
The Bessent Buyback Plan: What It Can and Cannot Do
Treasury Secretary Bessent's announcement of doubled buyback operations generated enormous market attention. The mechanism is that the Treasury buys back its own older, long-dated debt using proceeds from issuing new short-term bills, flattening the yield curve by reducing long-dated supply. In theory, this reduces the term premium by removing supply from the part of the curve investors least want to own.
In practice, the problem is math. The Treasury needs to issue several trillion dollars in new debt over the next 12 months to fund the deficit and roll over maturing obligations. Buying back $4 to $8 billion per auction cycle is not meaningfully changing the supply equation. It is a signal, not a solution. Markets understood that within hours, which is why yields snapped back to previous levels the following day. The policy is better than nothing. It is not large enough to arrest the structural repricing of long-dated bonds.
The Investment Implication
Rising long-term yields are not uniformly bad for every stock in the portfolio. They are bad for high-multiple growth stocks and good for certain value stocks, commodity producers, and financial companies that benefit from a steeper yield curve.
JPMorgan at Week 13 benefits from a steeper yield curve because it widens the net interest margin between what JPMorgan earns on loans and pays on deposits. ExxonMobil benefits because oil prices tend to rise alongside inflation and growth expectations that also drive yields higher. Palantir at 75x forward earnings is the most exposed stock in the entire MCI portfolio to sustained elevated yields, because its valuation requires a long stretch of compounding growth that gets discounted at a higher rate when yields stay elevated.
The analytical lesson is this: you cannot evaluate a stock's valuation in isolation from the risk-free rate environment. A stock at 25x earnings in a 2% yield world is priced very differently from a stock at 25x earnings in a 5% yield world. The multiple looks the same on a spreadsheet. The implied risk premium is completely different. Every MCI analysis has evaluated valuation relative to the company's history and sector peers. This week's Treasury market is a reminder to evaluate valuation relative to the risk-free alternative as well.
What to Watch
The September 15-16 Fed meeting is now the most important near-term event in markets. The CME FedWatch tool is pricing an elevated probability of a rate hike. If Warsh's Jackson Hole speech, landing next week, clarifies that the hiking bar is high and the hold is the base case, yields likely retrace and multiples recover. If Warsh signals genuine willingness to hike in September, the bond market selloff accelerates and growth stocks face another leg down.
The August payrolls report, due September 5, is the critical data point before the Fed meeting. A strong jobs number alongside sticky PCE inflation closes the door on any dovish pivot and validates the hawkish yield move. A weak jobs number reopens the hold debate.
This analysis introduced the concept of duration risk, the idea that the longer into the future a cash flow arrives, the more sensitive its present value is to changes in the discount rate.
Every bond has a duration, measured in years, that tells you how much its price changes for a 1% move in interest rates. A 10-year bond has roughly 8 to 9 years of duration. A 30-year bond has roughly 18 to 20 years of duration. When yields rise 1%, a 30-year bond falls approximately 18 to 20% in price.
Stocks work the same way, even though most investors do not think of them in these terms. A company where most of the value resides in near-term earnings, like ExxonMobil at 12x with a 3% dividend, has short duration because investors are not betting heavily on distant future growth. A company where most of the value resides in compounding growth over the next decade, like Palantir or Nvidia, has long duration because investors are discounting cash flows that arrive far in the future.
The 30-year Treasury yield going from 4.5% to 5.28% in a week is a violent reminder that duration is not a concept that lives only in bond portfolios. It lives in every valuation multiple in this blog. The companies most exposed are the ones with the longest duration: the highest multiples, the most growth-dependent valuations, and the smallest current free cash flow relative to market cap. Owning a portfolio that accounts for duration risk across both bonds and equities is not a bond manager's job. It is every investor's job.