본문 바로가기
Economics × Management Theory Weekly

Oct 4 – Oct 10, 2026

Korean original

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

The single tension running through this week is a contradiction: the huge expectations AI is generating for the future are rising at the same time as the discount rate used to convert those expectations into today’s money. On the macro side, oil and energy costs are pushing inflation expectations up again, and talk of a further rate hike before year-end has emerged. Gold has slipped to near a nine-week low while the dollar and long-term yields hold firm, which means expensive money will not ease quickly. On the growth side, Samsung Electronics is forecasting quarterly operating profit in the 100 trillion won range, the S&P 500 is eyeing a record high, and SpaceX is seeking to raise $40 billion to buy Nvidia chips: AI capital spending is lifting earnings and share prices together. The two forces meet in the credit market. Hyperscaler borrowing is forcing a repricing of risk in the $10 trillion US corporate bond market, high-yield spreads are flashing a warning, Asian issuance has stalled and a large IPO in Australia has been withdrawn. In short, AI is no longer an equity story but a story of capital spending funded by debt. Next week’s results from banks and chipmakers will show how fast that investment comes back as cash flow.

Deep dive — one theory Finance

Why tech stocks hit records as rates rise: the tug-of-war between discount rates and cash flows

Theory
Discounted Cash Flow (DCF)
Origin
Irving Fisher's *The Theory of Interest* (1930) laid the foundation of present value, and John Burr Williams (1938) applied it to valuing stocks.
In short
An asset is worth the money it will bring in, discounted for waiting and risk, converted into today's money and added up, because 10,000 won in the future is worth less than 10,000 won now. This is where the power of interest rates comes from. When the discount rate rises, assets that earn their money further in the future are cut more heavily, which is why growth stocks are especially weak against rates.
Limits
The results depend too heavily on assumptions. In particular, nudging the lumped-together value of the distant future or the discount rate even slightly can flip the conclusion, which also makes it easy to reverse-engineer the answer you want. It is hard to use at all for assets that do not yet earn money.

The discounted cash flow lens is especially sharp right now because the market is in an unusual phase: it is raising the discount rate on one side and enlarging the numerator on the other. While Treasury yields sat at their highest in 20 years, dividend stocks and long bonds were hit together, and TLT recorded a historic drawdown, tech stocks in the same market set new all-time highs. This flatly contradicts the textbook intuition that growth stocks should fall first when discount rates rise, because the further away the money, the more it is cut. The contradiction arises when you look only at the denominator and not the numerator. An asset's price is the sum of the cash it will earn in the future, discounted for waiting and risk. If expectations for the numerator rise faster than the denominator, the price can keep going up.

Put this week's news into this frame and the picture becomes clear. A handful of very large companies such as Nvidia, Microsoft and Meta show every quarter that AI demand is turning into real orders and profit, so their numerators have visibly thickened. Samsung Electronics' forecast of quarterly operating profit in the 100 trillion won range was evidence that the numerator is not one company's story but is spreading across memory and infrastructure. Dividend stocks and long bonds, by contrast, have fixed numerators and must absorb the full rise in discount rates, which is why those relying on dividends for retirement income were hit especially hard. Winners and losers split this sharply in the same rate environment not because the market is irrational, but simply because assets whose numerators can move are treated differently from those whose numerators cannot. Still, Samsung investors did not bid the stock up despite record results and demanded evidence that 'these profits will last'. That reads as a signal that the market has started to judge the durability of the numerator rather than its height.

Alumni can take away two generalizations. First, when looking at a portfolio in a rate cycle, sort it not by 'growth versus value' but by 'can the numerator move with rates?' Companies whose profits come from the same causes as rising rates (a strong economy, the AI investment boom) offset higher discount rates on their own, but fixed-coupon bonds and dividend stocks with stagnant earnings have no way to do so. Second, do not forget this framework's weakness: it is heavily driven by assumptions. Nudge the lumped-together value of the distant future or the discount rate even slightly and the conclusion flips, so today's record highs rest on the assumption that 'AI cash flows will last for decades'. That is why next week's results from banks and chipmakers are not just a quarterly report card but the first piece of evidence for that assumption. It is time to watch for this: the moment the numerator wobbles, the higher denominator comes back to hurt twice as much.

Source story: Stocks are hitting records despite surging yields. Cramer explains why

Cases by layer 6

Macro environment · Aggregate Demand - Aggregate Supply (AD-AS)

How Gas, Diesel and Electricity Prices Are Squeezing American Voters

Rising costs for essential energy in the US, including gasoline, diesel and electricity, have made the burden on households a political issue. The same week, Asian stocks fell on inflation worries stoked by a spike in oil prices, and high energy prices were cited as putting upward pressure on consumer prices and pushing back expectations for when the Fed will cut rates.

The aggregate demand and aggregate supply framework fits because this is not a cost problem at one company but a shock at the aggregate level, across US prices and output. More decisive is that the source of the shock is clearly on the supply side. Prices rose not because hot consumption made energy demand explode, but because the producing side was blocked: Middle East tensions, risks around the Strait of Hormuz and hurricane-related disruptions to Gulf production. On top of that, electricity rates and transport costs cannot adjust freely in the short run, so prices are sticky. That stickiness is exactly what opens the path for the shock to drag down output as well.

This news is usually read as a political story: 'gas prices rose and voters are angry'. Through this lens, though, it becomes clear why policymakers have so few cards to play against this inflation. If inflation came from too much buying power, raising rates to suppress it would work. If it came from blocked production, raising rates would push output down too. Trump toying with a suspension of the federal gasoline tax and the G7 considering releases of reserve oil and diesel also read as attempts to act directly on supply rather than suppress demand. In effect, policymakers have themselves acknowledged where this pressure came from.

Through this framework, the US economy is at the start of an uncomfortable combination in which prices and growth do not move in the same direction. While energy pushes up costs, consumer sentiment in Australia fell further under rate hikes and oil-price pressure, and Japanese household spending fell for a ninth straight month despite rising wages. This is the classic pattern of prices rising from the supply side while buying power weakens. If the Fed stays tight, demand may break before inflation is tamed. If it turns to easing, energy-driven inflation expectations may spill into wage demands. Only options with costs on both sides remain.

Takeaway — Don't treat 'inflation' in a portfolio as a single lump; first separate whether it is demand-driven or supply-driven. If it is supply-driven, energy and real assets, and companies with the power to pass costs on in prices, are in a relatively stronger position than assets betting on rate-cut expectations.

WSJ Markets

Macro environment · Monetary Policy Transmission Mechanism

Trump reiterates pledge to send $5,000 checks, and hands out smaller payments, as midterm elections loom

Ahead of the midterm elections, former President Trump reaffirmed $5,000 direct payments conditional on Republicans controlling Congress, while also promoting federal payment programs already under way. The same week, the governor of the Bank of France warned that the country risked being 'strangled by interest rates'. A US 10-year auction went well, but French, Italian and British government bonds came under pressure.

The premise of this lens holds because, although the policy rate itself has not moved, expectations about rates are shifting sharply. Large fiscal payment promises raise both future government bond issuance and inflation expectations, so they push long-term market rates up before the central bank even acts. Those rates then reach the real economy through mortgages and corporate borrowing. US mortgage rates topping 7% for the first time in years, leaving homeowners locked in and unable to move or renovate, and Asian dollar-bond issuance stalling because of the highest rates in about two years are evidence of where and how that channel is operating now.

This news is usually consumed as a political fight over 'cash handouts for the election'. But recall that money travels from the central bank to households along several channels, and a picture emerges: the channel fiscal policy has opened is colliding head-on with the channel monetary policy is trying to close. The important question is not 'how much will $5,000 raise consumption?' but whether the promise will push up long-term rates and narrow the housing and corporate-investment channels even further.

Read through this framework, fiscal expansion boosts consumption in the short run while tripping itself up through long-term rates. Households stuck in place by high mortgage rates, with home-equity credit lines also more expensive, are cutting spending on durable goods. In effect, even as one hand gives them a check, the other lets interest costs eat into disposable income. The same rise in rates turning into warnings about fiscal health in a heavily indebted country like France means this channel works across borders. The asymmetry, in which peripheral spreads widen the more stable US auctions become, is explained in the same way.

Takeaway — When reading fiscal news, check the mark a promise leaves on long-term rates and the term premium before its payment size. Stimulus delivered when the channel has narrowed tends to be absorbed by rates first rather than reaching the real economy as hoped.

CNBC Markets

Industry · Moral Hazard / Adverse Selection

AI Debt Spree Hammers Tech Debt as Traders Rush to Reprice Risk

As massive fundraising for AI infrastructure continues, the risk of tech-sector debt is being rapidly repriced in the US corporate bond market, which exceeds $10 trillion. Credit insurance prices are jumping and volatility is rising, high-yield spreads are flashing a warning and issuance in Asia has slowed.

The premise that lenders and borrowers see different things holds clearly here. Hyperscalers and data-center operators know far more about how much cash the money they raised will actually generate in a few years and how short chip depreciation cycles will be. Bond investors price the debt on a summarized signal: 'upgraded earnings estimates'. On top of that, much AI capital spending is funded through special-purpose structures or private credit, which blurs where the risk ultimately sits. The less clear it is where losses will flow, the greater the incentive to take on more risk.

This news is usually read as a growth narrative: 'the AI investment frenzy has spread to the bond market'. Seen as what happens in a market with asymmetric information, however, the widening spreads are not a spread of pessimism. They are closer to investors being unable to tell good borrowers from bad ones and starting to raise the average price. When that happens, borrowers with solid cash flows step back rather than issue at higher prices, and only the urgent ones stay in the market. SpaceX seeking its $40 billion through private financing led by Apollo rather than public bonds, and Nvidia-backed Firmus shelving its Australian IPO, look like two sides of the same coin.

Through this lens, the AI financing market is not 'getting more expensive' but 'starting to sort'. Credit insurance premiums jumping first means investors are at the stage of insuring the whole sector before scrutinizing individual companies. If this stage drags on, even sound borrowers end up paying high prices and giving up on issuance. The slowdown in Asian issuance and Firmus's withdrawn IPO signal that the sorting is tightening from the periphery, not the center. A contagion path, in which a credit event at one company spreads to funding conditions for the whole sector, as in the Paramount case, is also open.

Takeaway — When considering AI-related investments, look first at the terms on which a company is borrowing rather than at its equity valuation. For companies shifting funding from public to private markets, weigh the possibility that they are avoiding public-market scrutiny, not just growing fast.

Bloomberg Markets

Industry · Porter's Five Forces

SpaceX Makes Big Play to Become a Wireless Carrier

SpaceX announced plans to acquire spectrum licenses to enter the wireless business. A powerful potential competitor backed by a satellite network has appeared in a telecom infrastructure market long dominated by large incumbent carriers.

The five forces framework fits because the unit of analysis is not one company's results but the profit structure of the telecom industry. Telecom is an industry where a few operators have long protected their profits thanks to two huge barriers to entry: spectrum licenses and investment in base stations. If a player that can bypass terrestrial networks by satellite also seeks licenses, the height of those barriers has to be measured again. The condition that the industry's boundaries are still clear and its structure will not melt overnight, so it is worth taking a snapshot, also holds.

This news is usually read as a corporate-strategy story: 'Musk is jumping into another new business'. Through the lens of industry structure, though, the point is not whether SpaceX can make money in telecom. It is that the mere existence of an entry threat erodes incumbents' pricing power in advance. Even without actually entering, the possibility of entry changes rates and investment plans. That is why the real victim of this news is not SpaceX's rivals but the average profit margin of the entire telecom industry.

Through this framework, the telecom industry is entering a phase in which the threats of entry and substitution are rising at the same time. Direct-to-satellite services will first take areas terrestrial networks never reached, such as mountains and the sea, neutralizing the coverage premium incumbents sold most dearly. Carriers must raise capital spending to defend themselves, and borrowing that money now costs more. The article's reading that 'the long-term growth premium has risen' applies only to the newcomer, not the industry as a whole. For incumbents, the same event comes back as margin pressure.

Takeaway — When investing in industries such as telecom that have protected profits with barriers, check what the barriers are made of rather than current market share. Walls built of licenses and facilities become sunk costs, not assets, the moment a technology emerges that bypasses both.

WSJ Markets

Company · Resource-Based View (RBV)

Samsung Investors Want Evidence Profit Boom Has Staying Power

Samsung Electronics forecast record quarterly operating profit in the 100 trillion won range on the back of the AI boom, but its share price did not gain comparable momentum. The market is demanding evidence that these profits are structural, not a one-off cycle.

This lens fits because, even within the same memory industry, each company holds different resources and the gaps are hard to close. High-bandwidth memory design and yield, a track record of passing customer qualification, and accumulated experience running leading-edge process equipment cannot be bought. That is exactly why the question investors ask Samsung has shifted from 'how much did you earn this quarter?' to 'what produced that profit, and can others copy it?' The condition that the unit of analysis is a company's internal capabilities rather than industry conditions also holds.

This news is usually read as 'a strange case of strong results but a flat share price'. Through the lens of resources, however, there is nothing strange about it. Money made from rising memory prices comes from a shared resource, industry conditions: everyone enjoys it on the same terms, and it disappears for everyone when prices turn. The 'evidence of durability' the market demands is, in the end, a demand to separate how much of this profit belongs to Samsung alone and how much is a gift of the cycle. Until that distinction is made, a high multiple will not follow.

Through this framework, Samsung's task is not to sell more but to show that the source of its profit has changed. Profit ridden on commodity memory prices has to be given back when the cycle turns. Profit from long-term supply arrangements with specific AI customers, or from process capabilities that rivals could not match even after years, remains when prices fall. This week's move by investors to award high multiples only to companies with structural cash flow suggests a market has begun in which companies that cannot explain where their numbers come from, rather than the numbers themselves, are systematically discounted.

Takeaway — When evaluating a company, write down the size of its profit and the source of its profit separately. Profit handed over by industry conditions and profit created by what only that company has receive completely different multiples, even as the same operating profit.

Bloomberg Markets

Company · M&A Synergy / Winner's Curse

Schneider Said to Near Deal to Buy PTC for More Than $20 Billion

Schneider Electric was reported to be in the final stages of talks to acquire US software company PTC for more than $20 billion. The deal targets demand for digital transformation in manufacturing and industrial automation, with a hardware powerhouse buying an entire design-software capability.

This framework fits because there is clearly a value path that opens only through combination. Combining Schneider's on-site equipment with PTC's design and lifecycle software would let it sell everything from design to operation of a customer's factory as a single flow, with room to be bigger than the two apart. At the same time, the $20 billion price tag is public, so the premium paid can be weighed against expected synergies. And because this is a hardware company absorbing a software organization, integration execution is the real variable that decides the outcome, which also fits the conditions. Moreover, it is a large deal struck while M&A deal counts are shrinking on rising rates and election risk, so it is being paid for at a higher cost of capital than before.

This news is usually read as confirmation of industry conditions: 'demand for industrial automation is that strong'. Through the lens of acquisition arithmetic, however, the question changes. For shareholders, what matters is not whether synergies exist but who captures them. The old lesson that the highest bidder in a competitive auction is usually the one who misjudged most bites especially hard in a phase like this, when everyone believes digital transformation is the right answer.

Through this lens, the success of this deal will be decided not by the strategic logic on announcement day but by integration execution years later. To spend $20 billion when the cost of capital has risen, the investment must earn more than the company's cost of funding. But hardware sales organizations and software subscription models run on different cycles for pricing and customer relationships, so estimated synergies easily leak away while cultures and systems are joined. The higher rates are, the more heavily the present value of synergies realized far in the future is cut. So today's deal starts with a task it must prove in a much shorter time than the same deal in the low-rate era.

Takeaway — When reading news of a large acquisition, note the premium paid and the timing of realization before the strategic rationale. Whether the integration plan announced from next week onward gives concrete numbers for the timing of cost savings and cross-selling targets will be the first clue to whether the buyer has avoided the winner's curse.

Bloomberg Markets

Other signals this week 18

Based on 176 items over 7 days

Past issues