본문 바로가기
Foresight Signals
Economics × Management Theory Weekly

Jul 26 – Aug 1, 2026

English translation of the Korean original, prepared with AI assistance. Korean original

Not investment advice — these are resources for learning and forming your own view.

This week the market moved along a single axis: the tension between the future profits AI has promised and the bills piling up today. On the macro side, the Fed held rates despite inflation worries, and long-term Treasury yields jumped to a 19-year high. That pushed up the discount rate on growth stocks, whose profits arrive only in the distant future. On the corporate earnings side, Microsoft posted record Azure results and justified its higher capital expenditure. Meta, by contrast, fell 10% as attention fixed on the cost of the same kind of AI investment, so the same spending was read in opposite ways at different companies. On top of both sits credit. Big Tech is sharply increasing bond issuance and borrowing to build data centers, so AI’s bills are starting to pile up first in the bond market, not the stock market. In short, the AI story remains strong, but the price of the money needed to turn it into physical assets is rising fast. Next week, watch which way Big Tech free cash flow and corporate bond spreads resolve this tension.

Deep dive — one theory Finance

When Rates Discount the Future: Why a Surge in the 30-Year Yield Hits Growth Stocks

Theory
Discounted Cash Flow (DCF)
Origin
Irving Fisher's The Theory of Interest (1930) laid the foundation for present value, and John Burr Williams (1938) applied it to stock valuation.
In short
An asset's value is the sum of the money it will earn in the future, discounted for waiting and risk and converted into today's money, because ten thousand won in the future is worth less than ten thousand won now. This is where the power of interest rates comes from. When the discount rate rises, assets that earn cash only in the distant future lose more value, which is why growth stocks are especially vulnerable to rates.
Limits
Results depend too heavily on assumptions. In particular, the lumped-together terminal value for the distant future and the discount rate can overturn the conclusion with a small change, so it is easy to work backward to a desired answer. It is also hard to use at all for assets that are not yet profitable.

Present value and discounted cash flow are working with unusual sharpness now because this week's market replayed the theory's central idea: an asset is worth the sum of its future earnings, discounted back to today's money. The Fed held rates despite inflation worries, and long-term yields rebelled, sending the 30-year Treasury yield sharply higher. A higher discount rate cuts the value of assets that earn cash only in the distant future by more, and that showed up immediately as a sell-off in growth stocks.

Seen through this lens, story 82 is explained in a new way. Instead of treating the post-Fed drop in stocks as simple disappointment selling, the lens explains why tech stocks were hit hardest. The value of large AI stocks rests heavily on recovering data center investment years from now. With their cash concentrated in the distant future, they are the assets most sensitive to a rise in the discount rate. The chip sector plunge in story 65, which put the Nasdaq 100 into correction, follows the same principle: as rates rose, the weight of future cash fell across the board.

The general lesson for alumni is that the real risk in growth investing lies not in a company's business story but in the discount rate used to convert that story into today's price. Even with identical earnings and the same vision, a move of a few tens of basis points in long-term rates can overturn a valuation. Alumni should therefore recheck how far into the future a portfolio's cash flows are concentrated in each rate regime. The theory's limits are worth remembering too: it is hard to apply to assets that do not yet earn profits, which is also a root cause of the volatility in pure-play AI stocks.

Source story: Stocks Fall as 30-Year Bond Yields Surge After Fed | The Close 7/29/2026

Cases by layer 6

Macro environment · Monetary Policy Transmission Mechanism

US borrowing costs hit 19-year high as Fed defies inflation fears

US Treasury yields hit their highest level in 19 years, yet the Fed held rates and kept a hawkish stance. It did so despite fears of an inflation surge from geopolitical risk tied to Iran. The market expressed its skepticism that the Fed would respond enough to inflation by selling bonds.

Monetary policy transmission theory holds that changes in policy rates, or in expectations about them, reach the real economy only through intermediate channels: market rates, lending and asset prices. This week those very channels gave an abnormal signal. The Fed held its short-term policy rate, yet long-term Treasury yields jumped to a 19-year high. That means the central bank's hand and the price the market puts on money have started to point in different directions.

This news is commonly read as criticism of policy: that the Fed is complacent about inflation. Through this lens, however, the point is not the Fed's attitude but which link in the transmission chain has loosened. Even if the policy rate is held, if the long-term rate channel opens on its own, corporate funding costs and stock discount rates tighten regardless of what the Fed intends. What decides the outcome is not where the central bank sets rates but how the market reinterprets that signal and passes it on to the real economy.

Through this lens, the hold is a signal of 'loss of control' rather than either easing or tightening. Even while the Fed holds short-term rates, the long-term rate channel is carrying out tightening on its own. We have entered a phase in which the bond market, not the Fed, decides corporate debt burdens and growth stock discount rates. In story 80, Gundlach said that 'the bond market is demanding that the Fed act on inflation', which points to the same revolt in the channel.

Takeaway — This is a phase in which judging monetary conditions from the policy rate headline alone is easy to get wrong. Alumni should treat how the long-term yield curve actually moves, not the Fed's decision, as the true signal of funding costs.

Financial Times

Industry · M&A Synergy / Winner's Curse

Big Tech credit risks rise sharply as AI spending soars

As AI investment expands, Big Tech's data center capital expenditure is surging, and borrowing to fund it is rising fast. Credit risk in the sector is clearly increasing. Concerns include wider corporate bond spreads, downgrade pressure, and a lending squeeze if the economy slows.

Real options theory applies here because data center investment is a typical case. Once built, it carries irreversible sunk costs, and the uncertainty over whether AI profits will materialize clears only with time. Story 75 points out that AI revenue is growing fast but the real benefit is uncertain compared with trillions of dollars in spending. In such a situation, the flexibility of the investment itself has value, and that is the world of real options.

Big Tech's capex surge is commonly praised as an aggressive bet to claim the future or, conversely, criticized as reckless waste. Through the real options lens, the real question is whether the spending leaves the choice to scale up or wind down in stages, or hardens into concrete all at once. The theory holds that the more uncertain things are, the more valuable the right to wait becomes. It shows that companies that pour money into large, irreversible up-front investments are burning that option value themselves.

Through this lens, the rise in credit risk from data center capex is not just a worsening financial indicator. It is the result of Big Tech giving up the costly option of flexibility and switching to an irreversible bet. In a market where rivals fill the space first, waiting becomes a loss, so up-front investment may be rational. But if AI revenue falls short of expectations, the hardened facilities remain as debt that can be neither wound down nor scaled back.

Takeaway — When looking at AI infrastructure stocks, ask whether the spending can be adjusted in stages, not just how large the capex is. The more a company concentrates on large irreversible up-front investment, the sooner it will pay a price in the bond market if revenue arrives late.

Financial Times

Industry · Moral Hazard / Adverse Selection

Nvidia and OpenAI in talks for up to $250 billion backstop to fund AI infrastructure plans

There is a possibility that OpenAI could secure a debt backstop of up to $250 billion, based on Nvidia's credit, to fund data center construction. In this structure, Nvidia, which sells the chips, also stands behind the borrowing of the customer that will buy them.

Moral hazard theory fits here because the moment Nvidia backstops a customer's debt, a structure forms in which part of the losses shifts to a third party rather than the actual borrower. If OpenAI can raise debt at a scale it could not carry on its own credit, by leaning on Nvidia's guarantee, the theory's premise holds exactly: a party that knows someone else will cover the risk takes on more of it.

This backstop is commonly welcomed as strong momentum that reconfirms demand for AI infrastructure. Through the moral hazard lens, however, it may not be evidence of demand but a circular structure that props demand up artificially. When a chip supplier also funds the chip buyer, revenue looks larger, but it becomes unclear whether that revenue comes from the market's real ability to pay or was created by the supplier itself. The lens reveals the channel through which risk is passed on behind a glossy deal.

Through this lens, Nvidia's backstop reads as circular financing that props up its own chip sales. A structure in which the supplier takes on the customer's risk looks good to both sides in the short run. But if AI revenue falls short of expectations, Nvidia will hold both revenue it could not actually book and guarantees it must pay out. Seen alongside story 25, in which BlackRock arranged a $12.5 billion corporate bond for Meta's data center, AI infrastructure financing is increasingly tied into a chain of mutual guarantees.

Takeaway — When looking at revenue growth across the AI value chain, distinguish whether the demand is self-generated or created by a supplier's credit guarantee. A structure in which risk has been passed somewhere along the chain stays invisible in an upswing and is exposed all at once in a downturn.

CNBC Markets

Company · Signaling Theory

Meta Stock Drops 10% on Steeper AI Costs, Missed Forecast

Meta's shares fell 10% after it announced higher AI infrastructure costs and a sharp drop in free cash flow. The same week, Microsoft rose 7% as explosive growth in Azure cloud revenue justified higher capital expenditure. Both companies raised huge AI investments, yet the market reacted in opposite ways.

Signaling theory explains this contrast because the two companies' results show its principle exactly: only costly actions carry information, and a signal is credible only when it is more expensive for others to imitate. Microsoft's capex increase was backed by proven cash generation, a 32% surge in Azure revenue, and so became a costly signal that 'we can recover this spending'. Meta's spending at a similar scale left only costs, because free cash flow shrank to under $1 billion and it did not prove it could recover the outlay.

The divergence in the two share prices is commonly summed up in hindsight as good or bad results. Through the signaling theory lens, however, what the market actually priced was not the size of the spending but how credible a signal that spending was. Even for the same AI investment, only the side that passed the test of proven recovery capacity was accepted as a signal. The lens shows why the same action was read as opposite messages at different companies.

Through this lens, this week's Big Tech earnings season was a process in which the market sorted signals. It was not about AI investment itself but about whether companies had proven the cash generation to support it. Microsoft secured credibility for its signal with Azure as costly collateral. Meta voiced an agent vision but could not show the cash to back it, so it was not accepted as a signal. In story 67, Alphabet's negative free cash flow spread doubt across cloud competitors generally, an extension of the same sorting.

Takeaway — When evaluating large AI stocks, a rise in capex is by itself neither good nor bad news. Alumni should use proof of the cash generation to fund the spending as the test of a signal's credibility. The more a company's words and collateral diverge, the greater the risk of a correction.

WSJ Markets

Company · Network Effect

Apple tops $5tn valuation for first time

Apple topped a $5 trillion market capitalization for the first time and retook the title of the world's most valuable company from Nvidia. The main drivers cited were safe-haven demand from investors looking to avoid the burden of excessive AI spending and to spread out semiconductor sector risk.

Network effects theory applies to Apple because its conditions fit the Apple ecosystem exactly: the number of users itself raises the product's value, complementary goods are drawn in, and switching costs block movement. Billions of devices, along with a chain of complements such as the App Store, iMessage and app developers, reinforce one another. Users who enter find it hard to switch to another camp, and that structure has produced a solid base of recurring revenue.

Apple's $5 trillion breakout is commonly read as a belated rebound for a stock left out of the AI boom, or as a flight to safety. Through the network effects lens, the reason the market again gave Apple a premium becomes clearer. If Nvidia led on speculative overheating in AI infrastructure, Apple's value comes from the fact that people are already inside its ecosystem and find it hard to leave. The lens shows why, as themes rotate, money returned to Apple in particular.

Through this lens, story 33, where Apple overtook Nvidia, and story 64, on its $5 trillion breakout, read as money moving from speculative overheating around AI infrastructure to the stable cash flows of an ecosystem where people are already locked in. Apple's strength is not slightly better performance but the lock-in created by switching costs and a chain of complements. That leaves it relatively free of the volatility of the AI spending race.

Takeaway — When a theme overheats and then cools, money ends up returning to companies with lock-in structures that users cannot easily leave. Alumni should judge switching costs and ecosystem strength as sources of a defensive premium, rather than flashy growth stories.

Financial Times

Company · M&A Synergy / Winner's Curse

Sazerac Wants Another Shot at Jack Daniel’s Maker Brown-Forman

Liquor company Sazerac is approaching the founding family of Jack Daniel's parent Brown-Forman, renewing a takeover proposal that was rejected earlier. Observers say expectations of M&A in the premium spirits sector may rise, and share prices of related companies may become more volatile.

M&A synergy and winner's curse theory apply to this deal because the situation itself, reviving an offer already rejected once to persuade the founding family, exposes the tension between the premium paid and the expected synergy. Persuading a family to sell a company it controls will require a higher premium than before. Once that premium exceeds the real synergy from combining the two, the excess passes entirely to the selling family and the buyer's shareholders lose.

A renewed takeover offer is commonly taken as good news, on hopes of sector consolidation. Through this lens, however, the point is not whether the deal closes: the very fact of a second attempt already raises the risk of the winner's curse. To raise an offer that has been turned down once, the buyer must add to the price. The auction logic that the highest bidder is usually the one who has misjudged the most applies fully to this repeated courtship. The lens reveals the risk to the acquirer's shareholders behind the romance of consolidation.

Through this lens, Sazerac's second attempt is good news for Brown-Forman shareholders as a premium prospect, but a warning of winner's curse risk for Sazerac shareholders, who must pay a higher price. The Jack Daniel's brand is an attractive asset, but what matters is whether synergies possible only through combination justify the premium to be paid. Under a founding family's governance, the difficulty of executing the integration is added on top.

Takeaway — Focusing only on the rise in the target's share price in M&A news makes it easy to miss the risk to the acquirer's shareholders. Alumni should also weigh, from the acquirer's side, that a renewed offer carries a greater risk that the premium paid exceeds the synergy.

WSJ Markets

Other signals this week 18

Macro environment
Industry
Company

Based on 156 items over 6 days

Past issues