← Back to The Signal
Oikos · Global Macro · Cost of Capital

The 5% Hurdle: When Capital Stops Being Cheap

Higher Treasury yields raise the bar for stocks, AI investment, gold and Bitcoin. The cause of the move matters more than a single number.

Evidence reviewed after the October 5 U.S. equity close. Official Treasury daily estimates, not intraday highs. Source observation periods are stated below.

Conceptual brass balance weighing government debt against AI infrastructure, with a gold bar in the foreground
AI-generated As Above editorial illustration. The balance is a metaphor for competing uses of capital, not a measurement of borrowing or investment flows.

A higher Treasury yield is not just a warning on a trading screen. It is a new price against which a business expansion, a stock valuation and a speculative investment must compete.

On October 5, the U.S. Treasury's daily 10-year par yield was 5.31%; its 30-year reading was 5.66%. Those are dated official curve estimates, not the day's intraday peaks. Meanwhile, the Nasdaq Composite closed at a record and the S&P 500 rose 0.7%. More expensive money and rising stocks were happening together. Treasury data; AP closing report.

That apparent contradiction is the useful signal. Higher yields can reflect stronger expected growth, persistent inflation, greater compensation for holding long bonds, or a mixture. Each explanation implies a different risk. The question is not simply how high the yield goes. It is what investors are being paid to endure, and whether businesses can earn more than that cost.

The readout in four lines

  • The observation: the official 10-year yield is above 5%, while stocks can still rise.
  • The constraint: future cash flows must justify a higher required return.
  • The distinction: gold, Bitcoin and dollar tokens do not respond through the same mechanism.
  • The next test: watch earnings, inflation, funding conditions and credit together. One yield is not a diagnosis.

Who this is for: investors and business owners who want to understand what higher rates change. No bond-market background is required. This is a decision framework, not a recommended allocation.

First, separate three prices people call “the interest rate”

The policy rate is the short-term rate the Federal Reserve targets. The 10-year Treasury yield is a market benchmark for lending to the government over a much longer horizon. A business borrowing rate adds compensation for that borrower's risk and the loan's terms. They influence one another, but they are not interchangeable.

The Fed raised its target range to 3.75% to 4.00% on September 16. A company refinancing a loan does not automatically borrow at that rate, or at the 10-year Treasury rate. Its creditworthiness, collateral, maturity and lender's appetite matter. A profitable business with fixed-rate debt may feel little immediate change; an otherwise similar business facing a refinancing next month may feel a great deal. Federal Reserve statement.

The price changes with the commitment

U.S. Treasury nominal par yields · October 5, 2026

3 months4.22%
2 years4.84%
10 years5.31%
30 years5.66%

Bar lengths share a 0% to 6% scale. Selected maturities, not a time series.

Source: Treasury daily par curve. These are annualized benchmark yields, not deposit offers or guaranteed one-year total returns. Short maturities carry reinvestment risk; long maturities carry greater price sensitivity.

What the yield does, and does not, tell us

A long bond reflects expectations for future short-term rates plus a term premium: compensation for bearing the uncertainty of a long commitment instead of repeatedly buying short bills. That premium is estimated, not read directly from a screen. Different models can produce different answers. The New York Fed makes this identification problem explicit. New York Fed explanation.

Another useful comparison is with inflation-protected Treasuries, or TIPS. Treasury's October 5 10-year real par yield was 2.95%. Subtracting it from the nominal 5.31% gives an approximate nominal-real yield gap of 2.36 percentage points. That gap is not a pure inflation forecast: risk premiums, liquidity and differences between the instruments also affect it. Nor does the TIPS yield isolate the real term premium. Treasury real-yield data.

These observations establish a demanding inflation-adjusted benchmark. They do not, by themselves, prove that fiscal fear caused the entire move or that inflation expectations suddenly broke loose. Without a dated decomposition and a consistent comparison window, assigning exact shares to those stories would be false precision.

The financing pressure is tangible. Treasury's August forecast called for $739 billion of privately held net marketable borrowing in the third quarter and $628 billion in the fourth. Those are borrowing estimates, not completed issuance. More supply can require better prices for buyers, but the effect depends on demand, maturity mix, monetary policy and global saving. This is a pricing question, not automatic evidence of a buyers' strike. Treasury borrowing estimates.

The counter-signal: this is not a one-way inflation story

The freshest economic releases pull in different directions. September payrolls increased by only 29,000, unemployment was 4.2%, and July and August payroll gains were revised down by a combined 60,000. That is a reason to take labor weakness seriously, not to assume an overheating economy indefinitely. BLS, released October 2.

Yet the revised second-quarter GDP estimate showed 2.2% annualized growth, with private domestic final sales up 4.6%. August real consumer spending rose 0.6% from July. August PCE inflation, the broad consumer-price measure behind the Fed's target, was still 3.4% year over year; excluding food and energy, it was 3.0%. These releases cover different periods. None is an October growth reading. BEA GDP; BEA income and prices.

Our interpretation is positive but uneven activity, persistent inflation friction and a higher financing hurdle. It is not yet a clean recession call, a clean productivity boom or proof of fiscal dominance. The latter would mean fiscal financing needs overpower monetary discipline. A central bank that just raised rates is an inconvenient fact for any claim that this surrender is already complete.

Why stocks can rise while money gets more expensive

A stock price depends on both the cash investors expect a business to produce and the return they demand for waiting and taking risk. Higher expected profits can offset a higher required return. So can a temporary reduction in the extra return investors demand for owning stocks rather than government bonds.

That is why “yields up, stocks must fall” is not a trading rule. But it also explains why expensive stocks become vulnerable when the growth assumption disappoints. The cash-flow forecast and the discount rate can move against them together.

The same future $100 can be worth less today

Illustration: one payment of $100 in 10 years, unchanged in both cases.

Required return: 4%$67.56Present value today
Required return: 6%$55.84Present value today
Calculation: $100 ÷ (1 + annual discount rate)10. The present value falls approximately 17.3%. This isolates the mathematics of waiting. It is not a stock-market forecast, a Treasury-price calculation, or a claim that a stock's discount rate equals the Treasury yield.

There is no universal cliff at a 5.5% or 6% Treasury yield. What matters is the speed and cause of the change, the starting valuation, the debt that needs refinancing and the earnings delivered. A record capitalization-weighted index also cannot establish that the typical company is thriving. For that, compare equal-weight performance, smaller companies and earnings revisions over the same dates.

Nor are banks and insurers automatic winners. Higher reinvestment yields can help, while deposit costs, credit losses and unrealized bond losses can hurt. “Higher rates benefit financials” is a starting hypothesis, not sufficient company analysis.

AI connects the growth story to the financing story

AI infrastructure can support both sides of this tension. Building data centers creates investment demand today. Useful AI can, in principle, lift future productivity and cash flows. BEA's latest revision specifically identified data-center construction among the contributors to stronger fixed investment. That is evidence of construction activity, not proof of AI's eventual return on capital. BEA revision detail.

The important investor question is not whether the infrastructure is impressive. It is whether revenue, margins and useful output can justify the full cost of servers, power, buildings, financing and replacement. Internally funded investment and bond-funded investment also have different balance-sheet consequences. An announced capital budget is not equivalent to a completed bond sale.

This is the connection to our Physical AI and Robotics Desk: a demonstration is not a deployment, and a deployment is not an attractive return. A higher cost of capital raises the standard at every step. A provider can solve an engineering bottleneck and still destroy shareholder value if its customer economics never cover its funding and upkeep.

Gold, Bitcoin and dollar tokens are three different exposures

Gold: no income, but a different reason to own it

Gold pays no coupon. Higher real yields therefore make an income-producing alternative more attractive, all else equal. But all else rarely stays equal. Demand for monetary diversification or protection against policy mistakes may offset that headwind.

Gold's case should be tested against real yields, currencies and observed demand, not declared victorious whenever deficits rise. It can suffer drawdowns. It is not a substitute for money needed to pay a near-term bill.

Bitcoin: scarcity does not remove the need for buyers

Bitcoin is not a business with contractual cash flows, so assigning it an ordinary discounted-cash-flow value would be misleading. The relevant channels are opportunity cost, risk appetite, financing and the willingness of new holders to absorb supply.

A fiscal-confidence shock might strengthen its non-sovereign scarcity narrative. A funding squeeze might simultaneously force leveraged holders to sell. Both mechanisms can operate in the same week. Even crypto perpetual-futures funding is shaped by positioning and market design, not mechanically set by the Fed or the 10-year yield.

We are not publishing a current Bitcoin-yield correlation or ETF-flow estimate here: a matched, reproducible dataset was not established for this review. A few episodes of decoupling do not establish a reliable hedge. Our earlier distinction still helps: gold insurance and Bitcoin optionality are different portfolio ideas, not interchangeable claims of safety.

Stablecoins and tokenized Treasuries: follow who receives the interest

A dollar token can benefit its issuer without paying its holder a yield. Circle says USDC reserves include cash and short-dated Treasury exposures. That makes short-term reserve rates, not a rise in the 10-year yield alone, central to reserve earnings. Circle reserve disclosure.

In its second-quarter report, Circle disclosed $668 million of reserve income and $412 million of distribution, transaction and other costs. Larger average USDC balances supported income even as its reserve return rate declined year over year. The lesson is not “rates up, profits up.” It is that balances, reserve yields and distribution economics all matter. These are company-reported results. Circle Q2 2026 results.

A tokenized Treasury fund that passes income to investors is a different product from an ordinary payment stablecoin. Read the legal claim, eligibility, fees, redemption terms and custody arrangements. A token wrapper does not remove bond risk and can add operational or smart-contract risk. It distributes the bond market; it does not transcend it.

The global transmission: your currency matters

For an investor outside the United States, a dollar yield is only the beginning. Currency movement, hedging cost and local inflation can change the result. For a borrower with dollar debt but local-currency revenue, a stronger dollar can compound a higher financing rate.

Simultaneous increases in several countries' yields would be consistent with common global pressures, but they would not rule out local inflation, issuance or central-bank changes. This edition does not establish a synchronized UK, French and Japanese attribution dataset. The prudent global question is narrower: what currency funds the asset, when does the debt reset, and what currency pays it back?

Four paths, and the evidence that would distinguish them

1. Productive growth absorbs the cost

Yields stay elevated, but cash flows and investment returns improve. Look for broader earnings growth and useful output, not merely larger spending budgets. This would weaken a simple bearish-equity thesis.

2. Inflation keeps policy restrictive

Repeated firm core-price readings keep the expected short-rate path high. Long bonds and expensive equities can both struggle. A sequence of softer core readings would challenge this path.

3. Long-term financing risk gets repriced

Long yields rise without corresponding upgrades to growth or near-term policy expectations. Seek model estimates, repeated auction evidence and funding-market confirmation. Deficits alone are insufficient proof.

4. Growth weakens enough to change policy

Labor and demand deteriorate while inflation eases. Long Treasuries may regain their defensive role, even as equities or Bitcoin initially suffer. Persistent inflation would make this path harder.

These are conditional paths, not probability forecasts. The central thesis would weaken if financing costs fell sustainably, inflation eased and cash-flow growth broadened. It would strengthen if debt repriced upward while earnings and labor deteriorated. The evidence must be allowed to change the view.

A practical review you can do this week

  1. Separate spending money from investment money. On a simple list, put each obligation beside its date. Money needed soon has a different job from a long-term growth holding. Do not treat a long bond, gold or Bitcoin as equivalent to cash.
  2. Find the refinancing date. For your business or a company you own, identify which debt is fixed, which floats and which matures next. Look in the loan agreement or the financial statement's debt note. If the rate or maturity is unclear, ask the lender or adviser rather than estimate it.
  3. Stress the payment, not just the story. On an illustrative $100,000 interest-only floating balance, a two-percentage-point increase adds $2,000 a year before fees. Actual amortizing loans differ. Ask whether cash generation can absorb that change without relying on another refinancing.
  4. Name the job of each holding. Is it liquidity, income, protection against weaker growth, or a long-term growth wager? “It is a good asset” is not a sufficient answer.
  5. Write one observation that would change your mind. Falling inflation, broader earnings, worsening credit or confirmed funding stress should matter more than another confident prediction.

A Treasury yield above 5% is not a guaranteed annual profit if you sell before maturity. Price changes, reinvestment, inflation and taxes still matter. The promise to repay dollars and the stability of a bond's market value are different things. SEC investor education on bonds.

The As Above conclusion

In The Bond Hedge Is Changing, we argued that safety depends on the shock an asset is supposed to survive. This week's extension is that opportunity depends on the cost an asset must overcome.

That is the useful correspondence between macroeconomics and everyday decisions. At the national level, governments and businesses must finance their ambitions. At the household level, we must match commitments to resources. At the company level, technological possibility must become cash flow before financing runs out. The analogy is a way to ask better questions, not evidence for a market forecast.

More expensive capital does not require a crash to change the world. It can simply make weak projects harder to refinance and strong projects more valuable to distinguish. The discipline is not to abandon the future. It is to stop treating the future as if it were free.

Sources and evidence boundaries

Reviewed October 5, 2026, after the U.S. equity close. Publication dates and observation periods differ. Primary evidence is listed first; the AP report supplies the equity close. Scenario analysis and practical examples are As Above synthesis.

  1. U.S. Treasury: daily nominal par yield curve. October 5 observations; not intraday highs.
  2. U.S. Treasury: daily real par yield curve. October 5 TIPS observations.
  3. Federal Reserve: September 16 FOMC decision. Policy action, not a measurement of current market liquidity.
  4. New York Fed: Disentangling Messages from the Treasury Market. November 16, 2023; methodological background, not a 2026 term-premium estimate.
  5. Treasury: August 3 marketable borrowing estimates. Q3 and Q4 forecasts, not realized issuance.
  6. BLS: September Employment Situation. Released October 2; includes prior-month revisions.
  7. BEA: Q2 GDP, third estimate. Released September 30; annualized growth and revised investment evidence.
  8. BEA: August Personal Income and Outlays. Released September 30; monthly spending and year-over-year prices.
  9. Circle: reserve transparency. Issuer disclosure checked October 5; not an endorsement or an audit by As Above.
  10. Circle: second-quarter 2026 results. Historical quarterly issuer economics, not an October earnings forecast.
  11. Investor.gov: bonds and their risks. SEC investor education.
  12. Associated Press: October 5 U.S. market close. Nasdaq record and S&P 500 daily return.

We omit unverified correction targets, current crypto correlations, synchronized foreign-yield attribution and claims that auction demand has either failed or been definitively cleared. No present-day liquidity or positioning estimate is inferred from the headline yield.

Educational market commentary, not individualized financial, tax or investment advice. All investments involve risk. Hypothetical calculations omit taxes and transaction costs. AI assisted research organization, drafting and production; Marc Theiler retains editorial responsibility. The hero is AI-generated conceptual artwork, not market evidence.

The weekly Signal

Follow the evidence, not just the story.

One free weekly letter connecting AI, robotics and global macro, with sources, counter-evidence and the next test.

Share this Signal