<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Ai-Infrastructure-Cycle on Korea Invest Insights</title><link>https://koreainvestinsights.com/series/ai-infrastructure-cycle/</link><description>Recent content in Ai-Infrastructure-Cycle on Korea Invest Insights</description><generator>Hugo -- gohugo.io</generator><language>en</language><copyright>koreainvestinsights.com · @korea_invest_insights</copyright><lastBuildDate>Fri, 19 Jun 2026 22:35:53 +0900</lastBuildDate><atom:link href="https://koreainvestinsights.com/series/ai-infrastructure-cycle/feed.xml" rel="self" type="application/rss+xml"/><item><title>June Second-Half Event Map: A Weaker Central-Bank Put and Korea AI Memory Entry Discipline</title><link>https://koreainvestinsights.com/post/june-second-half-event-map-central-bank-put-korea-ai-memory-2026-06-19/</link><pubDate>Fri, 19 Jun 2026 23:10:00 +0900</pubDate><guid>https://koreainvestinsights.com/post/june-second-half-event-map-central-bank-put-korea-ai-memory-2026-06-19/</guid><description>
 &lt;blockquote&gt;
 &lt;p&gt;Context&lt;br&gt;
This follows the &lt;a class="link" href="https://koreainvestinsights.com/post/june-event-cluster-midpoint-review-rates-oil-ai-stress-test-2026-06-13/" &gt;June event cluster midpoint review&lt;/a&gt;, &lt;a class="link" href="https://koreainvestinsights.com/post/us-cpi-boj-fomc-macro-event-cluster-korea-reaction-function-2026-06-06/" &gt;CPI/BOJ/FOMC reaction-function note&lt;/a&gt;, &lt;a class="link" href="https://koreainvestinsights.com/post/ai-1996-vs-1999-fomc-hawkish-hold-productivity-capex-2026-06-17/" &gt;AI 1996 vs 1999&lt;/a&gt;, &lt;a class="link" href="https://koreainvestinsights.com/post/warsh-fed-expensive-money-era-forward-guidance-ai-infra-2026-06-19/" &gt;The Era of Expensive Money&lt;/a&gt; and &lt;a class="link" href="https://koreainvestinsights.com/post/foreign-return-after-24-day-kospi-selling-memory-rebalance-2026-06-12/" &gt;Have Foreign Investors Returned?&lt;/a&gt;.&lt;/p&gt;

 &lt;/blockquote&gt;
&lt;h2 id="tldr"&gt;TL;DR
&lt;/h2&gt;&lt;ul&gt;
&lt;li&gt;June is not simply an “AI rally continues” tape. It is an &lt;strong&gt;energy-led inflation stress → weaker central-bank put → higher volatility in long-duration growth and Korea beta&lt;/strong&gt; tape.&lt;/li&gt;
&lt;li&gt;Jobs, CPI, PPI, ECB, BOJ, FOMC and BoE all pointed in the same direction: central banks are not rushing back to easing.&lt;/li&gt;
&lt;li&gt;Korea AI memory and HBM remain high-quality structural themes. The issue is timing. The better setup is &lt;strong&gt;conditional buying after the June 25 PCE/GDP/durable-goods gate&lt;/strong&gt;, not chasing before it.&lt;/li&gt;
&lt;li&gt;The remaining June calendar: &lt;strong&gt;June 23 flash PMIs, June 24 ET Fed stress-test results, June 25 PCE/GDP/durable goods, and June 26 Russell reconstitution effective after the U.S. close&lt;/strong&gt;.&lt;/li&gt;
&lt;/ul&gt;
&lt;div class="thesis-callout"&gt;
 &lt;div class="thesis-callout__label"&gt;Core View&lt;/div&gt;
 &lt;div class="thesis-callout__body"&gt;
 The question is not whether AI is over. The question is whether Korea AI memory can withstand rates, dollar and oil stress in a world where central banks are less willing to cushion markets quickly.
 &lt;/div&gt;
&lt;/div&gt;
&lt;h2 id="what-june-has-confirmed"&gt;What June Has Confirmed
&lt;/h2&gt;&lt;table&gt;
 &lt;thead&gt;
 &lt;tr&gt;
 &lt;th&gt;Date&lt;/th&gt;
 &lt;th&gt;Event&lt;/th&gt;
 &lt;th style="text-align: right"&gt;Confirmed&lt;/th&gt;
 &lt;th&gt;Market read&lt;/th&gt;
 &lt;/tr&gt;
 &lt;/thead&gt;
 &lt;tbody&gt;
 &lt;tr&gt;
 &lt;td&gt;Jun 5&lt;/td&gt;
 &lt;td&gt;U.S. jobs&lt;/td&gt;
 &lt;td style="text-align: right"&gt;Payrolls +172k, unemployment 4.3%&lt;/td&gt;
 &lt;td&gt;No recession signal, less rate-cut comfort&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Jun 10&lt;/td&gt;
 &lt;td&gt;U.S. CPI&lt;/td&gt;
 &lt;td style="text-align: right"&gt;Headline +0.5% MoM / +4.2% YoY, core +0.2% / +2.9% YoY&lt;/td&gt;
 &lt;td&gt;Core stable, headline energy stress&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Jun 11&lt;/td&gt;
 &lt;td&gt;U.S. PPI&lt;/td&gt;
 &lt;td style="text-align: right"&gt;Final demand +1.1% MoM / +6.5% YoY&lt;/td&gt;
 &lt;td&gt;Cost pressure still uncomfortable&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Jun 11&lt;/td&gt;
 &lt;td&gt;ECB&lt;/td&gt;
 &lt;td style="text-align: right"&gt;Three policy rates +25bp&lt;/td&gt;
 &lt;td&gt;Europe also reads energy as inflation risk&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Jun 16&lt;/td&gt;
 &lt;td&gt;BOJ&lt;/td&gt;
 &lt;td style="text-align: right"&gt;Policy rate around 1.00%&lt;/td&gt;
 &lt;td&gt;Japan normalization continues&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Jun 17&lt;/td&gt;
 &lt;td&gt;FOMC&lt;/td&gt;
 &lt;td style="text-align: right"&gt;3.50-3.75% hold&lt;/td&gt;
 &lt;td&gt;Hold, but not easing&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Jun 18&lt;/td&gt;
 &lt;td&gt;BoE&lt;/td&gt;
 &lt;td style="text-align: right"&gt;3.75% hold, 7-2 vote&lt;/td&gt;
 &lt;td&gt;Two members wanted a hike&lt;/td&gt;
 &lt;/tr&gt;
 &lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;The BLS employment report showed +172k payrolls and a 4.3% unemployment rate. CPI showed stable core but an uncomfortable headline because energy was strong. PPI was more difficult: final demand rose 1.1% MoM and 6.5% YoY. That combination does not create an easy central-bank pivot.&lt;/p&gt;
&lt;h2 id="the-central-bank-put-has-weakened"&gt;The Central-Bank Put Has Weakened
&lt;/h2&gt;&lt;p&gt;The Fed held rates, but its statement still described inflation as elevated. The ECB hiked by 25bp. The BOJ raised its policy rate to around 1.00%. The BoE held, but the 7-2 vote included two hike votes.&lt;/p&gt;
&lt;p&gt;This is the practical meaning of a weaker central-bank put. Money is not disappearing, but it has a higher price. Multiples face stricter tests.&lt;/p&gt;
&lt;h2 id="korea-ai-memory-strong-theme-event-dependent-entry"&gt;Korea AI Memory: Strong Theme, Event-Dependent Entry
&lt;/h2&gt;&lt;p&gt;Samsung Electronics and SK hynix remain the cleanest Korea exposures to the AI memory bottleneck. HBM supply is constrained and customer qualification is hard. But Korean semiconductors are also high-beta assets tied to U.S. rates, the dollar, oil and foreign flows.&lt;/p&gt;
&lt;p&gt;The right sequence is therefore:&lt;/p&gt;
&lt;table&gt;
 &lt;thead&gt;
 &lt;tr&gt;
 &lt;th&gt;Condition&lt;/th&gt;
 &lt;th&gt;Meaning&lt;/th&gt;
 &lt;/tr&gt;
 &lt;/thead&gt;
 &lt;tbody&gt;
 &lt;tr&gt;
 &lt;td&gt;PCE does not reaccelerate&lt;/td&gt;
 &lt;td&gt;Fed hike pricing remains contained&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;U.S. long yields do not jump&lt;/td&gt;
 &lt;td&gt;Duration pressure eases&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Dollar and oil stabilize&lt;/td&gt;
 &lt;td&gt;Foreign flows into Korea improve&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Foreign buying in Samsung/SK hynix persists&lt;/td&gt;
 &lt;td&gt;The industrial thesis becomes actual flow&lt;/td&gt;
 &lt;/tr&gt;
 &lt;/tbody&gt;
&lt;/table&gt;
&lt;h2 id="remaining-june-gates"&gt;Remaining June Gates
&lt;/h2&gt;&lt;table&gt;
 &lt;thead&gt;
 &lt;tr&gt;
 &lt;th&gt;Korea time&lt;/th&gt;
 &lt;th&gt;Event&lt;/th&gt;
 &lt;th&gt;What to watch&lt;/th&gt;
 &lt;/tr&gt;
 &lt;/thead&gt;
 &lt;tbody&gt;
 &lt;tr&gt;
 &lt;td&gt;Jun 23&lt;/td&gt;
 &lt;td&gt;Global flash PMIs&lt;/td&gt;
 &lt;td&gt;Activity and price indexes&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Jun 25 05:00&lt;/td&gt;
 &lt;td&gt;Fed stress-test results&lt;/td&gt;
 &lt;td&gt;Bank capital-return visibility&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Jun 25 21:30&lt;/td&gt;
 &lt;td&gt;U.S. May PCE&lt;/td&gt;
 &lt;td&gt;The most direct Fed input&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Jun 25 21:30&lt;/td&gt;
 &lt;td&gt;U.S. Q1 GDP third estimate / profits&lt;/td&gt;
 &lt;td&gt;Growth and margin reality&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Jun 25 21:30&lt;/td&gt;
 &lt;td&gt;U.S. durable goods&lt;/td&gt;
 &lt;td&gt;CapEx and AI investment strength&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Around Jun 27 05:00&lt;/td&gt;
 &lt;td&gt;Russell reconstitution&lt;/td&gt;
 &lt;td&gt;Forced flows and close-auction liquidity&lt;/td&gt;
 &lt;/tr&gt;
 &lt;/tbody&gt;
&lt;/table&gt;
&lt;h2 id="strategy"&gt;Strategy
&lt;/h2&gt;&lt;p&gt;Do not chase QQQ or EWY before PCE. Keep Korea AI memory and HBM on watch, then buy pullbacks if PCE, rates, the dollar and oil cooperate. Watch large banks after the Fed stress test, but treat them as a financial-conditions signal unless capital-return visibility is clear. Energy-cost beneficiaries are tactical unless oil stability persists.&lt;/p&gt;
&lt;h2 id="final-view"&gt;Final View
&lt;/h2&gt;&lt;p&gt;The best June second-half idea is not “buy everything AI.” It is &lt;strong&gt;wait for the macro gate, then selectively buy Korea AI memory/HBM if the tape confirms&lt;/strong&gt;. In an expensive-money regime, a good theme still needs a good entry.&lt;/p&gt;
&lt;p&gt;Sources: &lt;a class="link" href="https://www.bls.gov/news.release/empsit.nr0.htm" target="_blank" rel="noopener"
 &gt;BLS jobs&lt;/a&gt;, &lt;a class="link" href="https://www.bls.gov/news.release/cpi.nr0.htm" target="_blank" rel="noopener"
 &gt;BLS CPI&lt;/a&gt;, &lt;a class="link" href="https://www.bls.gov/news.release/ppi.nr0.htm" target="_blank" rel="noopener"
 &gt;BLS PPI&lt;/a&gt;, &lt;a class="link" href="https://www.ecb.europa.eu/press/pr/date/2026/html/ecb.mp260611~4d41bd5e83.en.html" target="_blank" rel="noopener"
 &gt;ECB&lt;/a&gt;, &lt;a class="link" href="https://www.boj.or.jp/en/mopo/mpmdeci/mpr_2026/k260616a.pdf" target="_blank" rel="noopener"
 &gt;BOJ&lt;/a&gt;, &lt;a class="link" href="https://www.federalreserve.gov/newsevents/pressreleases/monetary20260617a.htm" target="_blank" rel="noopener"
 &gt;Federal Reserve&lt;/a&gt;, &lt;a class="link" href="https://www.bankofengland.co.uk/monetary-policy-summary-and-minutes/2026/june-2026" target="_blank" rel="noopener"
 &gt;BoE&lt;/a&gt;, &lt;a class="link" href="https://www.bea.gov/news/schedule" target="_blank" rel="noopener"
 &gt;BEA schedule&lt;/a&gt;, &lt;a class="link" href="https://www.lseg.com/en/ftse-russell/russell-reconstitution" target="_blank" rel="noopener"
 &gt;LSEG Russell&lt;/a&gt;.&lt;/p&gt;</description></item><item><title>The Era of Expensive Money: What the Warsh Fed Really Changed</title><link>https://koreainvestinsights.com/post/warsh-fed-expensive-money-era-forward-guidance-ai-infra-2026-06-19/</link><pubDate>Fri, 19 Jun 2026 22:20:00 +0900</pubDate><guid>https://koreainvestinsights.com/post/warsh-fed-expensive-money-era-forward-guidance-ai-infra-2026-06-19/</guid><description>
 &lt;blockquote&gt;
 &lt;p&gt;Context&lt;br&gt;
This note follows &lt;a class="link" href="https://koreainvestinsights.com/post/ai-productivity-real-evidence-generative-ai-at-work-fed-2026-06-17/" &gt;Is AI Productivity Real?&lt;/a&gt;, &lt;a class="link" href="https://koreainvestinsights.com/post/ai-1996-vs-1999-fomc-hawkish-hold-productivity-capex-2026-06-17/" &gt;Is AI 1996 or 1999?&lt;/a&gt;, and our June macro event reviews. The question here is what actually changed after the June FOMC.&lt;/p&gt;

 &lt;/blockquote&gt;
&lt;h2 id="tldr"&gt;TL;DR
&lt;/h2&gt;&lt;ul&gt;
&lt;li&gt;The June FOMC did &lt;strong&gt;not&lt;/strong&gt; announce a rate hike. The target range stayed at &lt;strong&gt;3.50-3.75%&lt;/strong&gt;.&lt;/li&gt;
&lt;li&gt;The more important shift is communication. Chair Warsh did not submit his own SEP projections, the statement became shorter, and forward guidance was deliberately reduced.&lt;/li&gt;
&lt;li&gt;The SEP was hawkish: the 2026 federal funds rate median rose to &lt;strong&gt;3.8%&lt;/strong&gt;, while 2026 PCE and core PCE projections rose to &lt;strong&gt;3.6%&lt;/strong&gt; and &lt;strong&gt;3.3%&lt;/strong&gt;.&lt;/li&gt;
&lt;li&gt;But the longer-run federal funds rate median stayed at &lt;strong&gt;3.1%&lt;/strong&gt;. This is not yet a confirmed 1999-style multi-hike cycle.&lt;/li&gt;
&lt;li&gt;The investment implication is not “no liquidity.” It is &lt;strong&gt;expensive money&lt;/strong&gt;. Capital still flows, but it becomes more selective and prefers AI infrastructure, power, defense, energy, physical bottlenecks, and proven cash flow.&lt;/li&gt;
&lt;/ul&gt;
&lt;div class="thesis-callout"&gt;
 &lt;div class="thesis-callout__label"&gt;Core View&lt;/div&gt;
 &lt;div class="thesis-callout__body"&gt;
 The Warsh Fed changed the market's habit more than the rate path. Investors can no longer rely on the Fed to translate the data for them. Markets need to price inflation, jobs, productivity, CapEx and cash flow more directly.
 &lt;/div&gt;
&lt;/div&gt;
&lt;h2 id="what-changed"&gt;What Changed
&lt;/h2&gt;&lt;p&gt;The Federal Reserve&amp;rsquo;s June 17 materials confirm three facts.&lt;/p&gt;
&lt;table&gt;
 &lt;thead&gt;
 &lt;tr&gt;
 &lt;th&gt;Item&lt;/th&gt;
 &lt;th style="text-align: right"&gt;June FOMC&lt;/th&gt;
 &lt;th&gt;Read-through&lt;/th&gt;
 &lt;/tr&gt;
 &lt;/thead&gt;
 &lt;tbody&gt;
 &lt;tr&gt;
 &lt;td&gt;Policy rate&lt;/td&gt;
 &lt;td style="text-align: right"&gt;3.50-3.75% hold&lt;/td&gt;
 &lt;td&gt;No actual hike yet&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;2026 funds-rate median&lt;/td&gt;
 &lt;td style="text-align: right"&gt;3.8%&lt;/td&gt;
 &lt;td&gt;Higher than March&amp;rsquo;s 3.4%&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Longer-run funds-rate median&lt;/td&gt;
 &lt;td style="text-align: right"&gt;3.1%&lt;/td&gt;
 &lt;td&gt;Structural neutral rate not yet repriced&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;2026 PCE / core PCE&lt;/td&gt;
 &lt;td style="text-align: right"&gt;3.6% / 3.3%&lt;/td&gt;
 &lt;td&gt;Inflation forecast moved up&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Warsh SEP&lt;/td&gt;
 &lt;td style="text-align: right"&gt;Not submitted&lt;/td&gt;
 &lt;td&gt;Dot-plot interpretation is less mechanical&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Statement&lt;/td&gt;
 &lt;td style="text-align: right"&gt;Shorter, less guidance&lt;/td&gt;
 &lt;td&gt;Market must read the data itself&lt;/td&gt;
 &lt;/tr&gt;
 &lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;The key is not only the dot plot. It is the removal of the market&amp;rsquo;s comfort mechanism. For years, investors have asked what the Fed would do with each data point. Warsh is trying to push the market back toward pricing the data itself.&lt;/p&gt;
&lt;p&gt;That weakens the Fed put. It does not eliminate it, but it makes valuation support less automatic.&lt;/p&gt;
&lt;h2 id="expensive-money-is-not-no-money"&gt;Expensive Money Is Not No Money
&lt;/h2&gt;&lt;p&gt;This regime does not mean capital disappears. Governments still spend. Big technology companies still invest. Energy security, defense, reshoring, AI data centers and power grids still pull capital.&lt;/p&gt;
&lt;p&gt;But money has a higher price. That changes the question.&lt;/p&gt;
&lt;div class="highlight"&gt;&lt;pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;-webkit-text-size-adjust:none;"&gt;&lt;code class="language-text" data-lang="text"&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;Old regime: How large can the TAM become?
&lt;/span&gt;&lt;/span&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;New regime: Who actually earns cash flow through the bottleneck?
&lt;/span&gt;&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;&lt;/div&gt;&lt;p&gt;In the old easy-money regime, long-duration growth, software multiples and unproven TAM stories could be priced generously. In the expensive-money regime, the market asks whether a company has pricing power, recurring demand, high ROIC and a direct link to big wallets such as AI hyperscalers, governments, defense budgets and utilities.&lt;/p&gt;
&lt;h2 id="1997-or-1999"&gt;1997 Or 1999?
&lt;/h2&gt;&lt;p&gt;The June FOMC looks closer to a 1997-style insurance-hike risk than a fully confirmed 1999-style tightening cycle.&lt;/p&gt;
&lt;p&gt;The reason is the long-run dot. If the Fed were formally accepting a much higher structural neutral rate, the longer-run federal funds rate median would likely move higher. It did not.&lt;/p&gt;
&lt;p&gt;But the risk is clear. If the September SEP lifts the longer-run dot, if inflation stays sticky, or if AI CapEx keeps easing financial conditions through equity markets, the 1999 analogy becomes stronger.&lt;/p&gt;
&lt;h2 id="korea-market-translation"&gt;Korea Market Translation
&lt;/h2&gt;&lt;p&gt;For Korea, the implication is selective exposure rather than broad beta.&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;Samsung Electronics and SK hynix remain core because HBM and AI memory are direct bottlenecks.&lt;/li&gt;
&lt;li&gt;Second-line semiconductors need stricter proof: orders, process bottlenecks, margins, and customer CapEx links.&lt;/li&gt;
&lt;li&gt;Power equipment, grid, defense, shipbuilding, nuclear and infrastructure remain important because they connect to hard spending needs.&lt;/li&gt;
&lt;li&gt;Pure theme stocks with no cash flow become more vulnerable when the Fed put weakens.&lt;/li&gt;
&lt;/ul&gt;
&lt;h2 id="final-view"&gt;Final View
&lt;/h2&gt;&lt;p&gt;The June FOMC was not mainly a rate-hike shock. It was a change in the market&amp;rsquo;s operating system. The Fed is reducing guidance. Investors must read the data directly. In that world, capital does not vanish, but it moves more narrowly toward real bottlenecks and cash flow.&lt;/p&gt;
&lt;p&gt;Sources: &lt;a class="link" href="https://www.federalreserve.gov/newsevents/pressreleases/monetary20260617a.htm" target="_blank" rel="noopener"
 &gt;Fed statement&lt;/a&gt;, &lt;a class="link" href="https://www.federalreserve.gov/mediacenter/files/FOMCpresconf20260617.pdf" target="_blank" rel="noopener"
 &gt;Warsh press conference transcript&lt;/a&gt;, &lt;a class="link" href="https://www.federalreserve.gov/monetarypolicy/files/fomcprojtabl20260617.pdf" target="_blank" rel="noopener"
 &gt;June 2026 SEP&lt;/a&gt;, &lt;a class="link" href="https://www.youtube.com/watch?v=YuaYWKXQLsw" target="_blank" rel="noopener"
 &gt;Wallga Ajae FOMC review&lt;/a&gt;, &lt;a class="link" href="https://www.youtube.com/watch?v=cN7xWS_LAxI" target="_blank" rel="noopener"
 &gt;Hankyung Global Market&lt;/a&gt;.&lt;/p&gt;</description></item><item><title>Is AI 1996 or 1999? The June FOMC and What Comes Next</title><link>https://koreainvestinsights.com/post/ai-1996-vs-1999-fomc-hawkish-hold-productivity-capex-2026-06-17/</link><pubDate>Wed, 17 Jun 2026 18:00:00 +0900</pubDate><guid>https://koreainvestinsights.com/post/ai-1996-vs-1999-fomc-hawkish-hold-productivity-capex-2026-06-17/</guid><description>
 &lt;blockquote&gt;
 &lt;p&gt;Context&lt;br&gt;
This note follows our work on AI productivity evidence, the mid-cycle AI supercycle, and the June macro event cluster. The question here is narrower: &lt;strong&gt;does AI look more like a 1996-style productivity disinflation story, or a 1999-style expectations and CapEx cycle?&lt;/strong&gt;&lt;/p&gt;

 &lt;/blockquote&gt;
&lt;h2 id="tldr"&gt;TL;DR
&lt;/h2&gt;&lt;ul&gt;
&lt;li&gt;AI today looks less like a clean &lt;strong&gt;1996 productivity-disinflation&lt;/strong&gt; regime and more like a &lt;strong&gt;1999-style expectations and CapEx cycle with a 1996 productivity option embedded inside it&lt;/strong&gt;.&lt;/li&gt;
&lt;li&gt;Workplace productivity evidence is real. Adoption is rising. But broad macro disinflation from AI has not yet been proven.&lt;/li&gt;
&lt;li&gt;AI data centers, chips, power, cooling, land, labor, and credit demand are already visible. These are current macro inputs for the Fed.&lt;/li&gt;
&lt;li&gt;The base case for the June FOMC is a hold, but not a dovish hold. It is more likely to be a &lt;strong&gt;hawkish hold&lt;/strong&gt; or &lt;strong&gt;hawkish neutral&lt;/strong&gt; outcome.&lt;/li&gt;
&lt;li&gt;The statement matters, but the key signals are the SEP dots, the longer-run neutral-rate dot, Chair Warsh&amp;rsquo;s comments on AI and r-star, and the July 8 minutes.&lt;/li&gt;
&lt;/ul&gt;
&lt;div class="thesis-callout"&gt;
 &lt;div class="thesis-callout__label"&gt;Core View&lt;/div&gt;
 &lt;div class="thesis-callout__body"&gt;
 The Fed is unlikely to pre-emptively treat AI as a rate-cutting disinflation force. Productivity is an option. CapEx, power demand, semiconductors, and financial conditions are current facts.
 &lt;/div&gt;
&lt;/div&gt;
&lt;h2 id="1996-vs-1999"&gt;1996 vs 1999
&lt;/h2&gt;&lt;p&gt;The 1996 version of the story is positive supply. Productivity rises, unit labor costs slow, inflation pressure eases, and the Fed can be patient even with strong growth.&lt;/p&gt;
&lt;p&gt;The 1999 version is different. The productivity story may be real, but asset prices, corporate investment, and capital-market expectations run ahead of realized diffusion. The Fed then has to watch demand, credit, valuation, and financial conditions, not only future productivity.&lt;/p&gt;
&lt;p&gt;AI has both elements. The 1996 option is real. Brynjolfsson, Li, and Raymond&amp;rsquo;s &lt;em&gt;Generative AI at Work&lt;/em&gt; finds a roughly &lt;strong&gt;15%&lt;/strong&gt; productivity gain for customer-support agents. Fed FEDS Notes shows AI adoption has moved beyond experimentation, and the San Francisco Fed argues policymakers need micro and sector evidence before the macro aggregates fully show the transformation.&lt;/p&gt;
&lt;p&gt;But the immediate macro shock looks more like 1999. AI infrastructure spending is pulling on chips, power, data centers, cooling, land, skilled labor, and financing. The April FOMC minutes already treated AI as two-sided: possible productivity disinflation over time, but also AI-related investment that could raise input costs now.&lt;/p&gt;
&lt;h2 id="why-this-matters-for-the-june-fomc"&gt;Why This Matters for the June FOMC
&lt;/h2&gt;&lt;p&gt;The June FOMC is scheduled for June 16-17, with the decision at 2:00 p.m. ET and press conference at 2:30 p.m. ET on June 17. The minutes are scheduled for July 8. The current target range is &lt;strong&gt;3.50-3.75%&lt;/strong&gt;.&lt;/p&gt;
&lt;p&gt;Recent data do not give the Fed an easy easing case:&lt;/p&gt;
&lt;table&gt;
 &lt;thead&gt;
 &lt;tr&gt;
 &lt;th&gt;Indicator&lt;/th&gt;
 &lt;th style="text-align: right"&gt;Latest reading&lt;/th&gt;
 &lt;th&gt;Policy read&lt;/th&gt;
 &lt;/tr&gt;
 &lt;/thead&gt;
 &lt;tbody&gt;
 &lt;tr&gt;
 &lt;td&gt;May CPI&lt;/td&gt;
 &lt;td style="text-align: right"&gt;+0.5% MoM, +4.2% YoY&lt;/td&gt;
 &lt;td&gt;Headline inflation uncomfortable&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;May core CPI&lt;/td&gt;
 &lt;td style="text-align: right"&gt;+0.2% MoM, +2.9% YoY&lt;/td&gt;
 &lt;td&gt;Better monthly core, still above target YoY&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;April PCE&lt;/td&gt;
 &lt;td style="text-align: right"&gt;+3.8% YoY&lt;/td&gt;
 &lt;td&gt;Fed&amp;rsquo;s preferred inflation gauge still high&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;April core PCE&lt;/td&gt;
 &lt;td style="text-align: right"&gt;+3.3% YoY&lt;/td&gt;
 &lt;td&gt;Core still too high&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;May payrolls&lt;/td&gt;
 &lt;td style="text-align: right"&gt;+172k&lt;/td&gt;
 &lt;td&gt;Labor market not weak&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;May unemployment&lt;/td&gt;
 &lt;td style="text-align: right"&gt;4.3%&lt;/td&gt;
 &lt;td&gt;Stable&lt;/td&gt;
 &lt;/tr&gt;
 &lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;The market already expects a hold. The real issue is what kind of hold.&lt;/p&gt;
&lt;h2 id="base-case"&gt;Base Case
&lt;/h2&gt;&lt;p&gt;My base case is a &lt;strong&gt;hawkish hold&lt;/strong&gt;.&lt;/p&gt;
&lt;table&gt;
 &lt;thead&gt;
 &lt;tr&gt;
 &lt;th&gt;Item&lt;/th&gt;
 &lt;th&gt;Base case&lt;/th&gt;
 &lt;/tr&gt;
 &lt;/thead&gt;
 &lt;tbody&gt;
 &lt;tr&gt;
 &lt;td&gt;Policy rate&lt;/td&gt;
 &lt;td&gt;Hold at 3.50-3.75%&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Statement&lt;/td&gt;
 &lt;td&gt;Data dependent, inflation still elevated&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Easing bias&lt;/td&gt;
 &lt;td&gt;Weakened or removed&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;2026 dot&lt;/td&gt;
 &lt;td&gt;Moves toward no cuts in 2026&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Longer-run dot&lt;/td&gt;
 &lt;td&gt;Holds at 3.1% or edges higher&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Press conference&lt;/td&gt;
 &lt;td&gt;AI productivity is possible, but not a current policy basis&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Minutes&lt;/td&gt;
 &lt;td&gt;More hawkish than the statement&lt;/td&gt;
 &lt;/tr&gt;
 &lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;The March SEP had a 2026 federal funds median of &lt;strong&gt;3.4%&lt;/strong&gt;, roughly implying one cut from the current midpoint. If the median shifts toward &lt;strong&gt;3.625%&lt;/strong&gt;, markets should read it as &amp;ldquo;no 2026 cuts.&amp;rdquo;&lt;/p&gt;
&lt;h2 id="scenario-map"&gt;Scenario Map
&lt;/h2&gt;&lt;table&gt;
 &lt;thead&gt;
 &lt;tr&gt;
 &lt;th&gt;Scenario&lt;/th&gt;
 &lt;th style="text-align: right"&gt;Probability&lt;/th&gt;
 &lt;th&gt;What happens&lt;/th&gt;
 &lt;th&gt;Market read&lt;/th&gt;
 &lt;/tr&gt;
 &lt;/thead&gt;
 &lt;tbody&gt;
 &lt;tr&gt;
 &lt;td&gt;Base: hawkish hold&lt;/td&gt;
 &lt;td style="text-align: right"&gt;65%&lt;/td&gt;
 &lt;td&gt;Hold, weaker easing bias, higher 2026 dots&lt;/td&gt;
 &lt;td&gt;Short rates supported, dollar firm, equity multiples capped&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Hawkish surprise&lt;/td&gt;
 &lt;td style="text-align: right"&gt;25%&lt;/td&gt;
 &lt;td&gt;Some hike dots, longer-run dot up, r-star language&lt;/td&gt;
 &lt;td&gt;Front-end yields rise, risk assets volatile&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Dovish surprise&lt;/td&gt;
 &lt;td style="text-align: right"&gt;10%&lt;/td&gt;
 &lt;td&gt;One-cut median stays, energy shock treated as temporary&lt;/td&gt;
 &lt;td&gt;Yields fall, dollar eases, risk relief&lt;/td&gt;
 &lt;/tr&gt;
 &lt;/tbody&gt;
&lt;/table&gt;
&lt;h2 id="what-to-watch"&gt;What to Watch
&lt;/h2&gt;&lt;p&gt;The key post-FOMC checklist:&lt;/p&gt;
&lt;table&gt;
 &lt;thead&gt;
 &lt;tr&gt;
 &lt;th&gt;Signal&lt;/th&gt;
 &lt;th&gt;Dovish&lt;/th&gt;
 &lt;th&gt;Hawkish&lt;/th&gt;
 &lt;/tr&gt;
 &lt;/thead&gt;
 &lt;tbody&gt;
 &lt;tr&gt;
 &lt;td&gt;2026 median dot&lt;/td&gt;
 &lt;td&gt;Stays near 3.4%&lt;/td&gt;
 &lt;td&gt;Moves toward 3.625%&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Longer-run dot&lt;/td&gt;
 &lt;td&gt;Stays 3.1%&lt;/td&gt;
 &lt;td&gt;Moves above 3.1%&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;AI language&lt;/td&gt;
 &lt;td&gt;Productivity and disinflation&lt;/td&gt;
 &lt;td&gt;CapEx, input costs, r-star&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Warsh tone&lt;/td&gt;
 &lt;td&gt;Patience toward headline inflation&lt;/td&gt;
 &lt;td&gt;Credibility and elevated inflation&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;July 8 minutes&lt;/td&gt;
 &lt;td&gt;Productivity gains emphasized&lt;/td&gt;
 &lt;td&gt;Input costs, asset valuations, policy firming&lt;/td&gt;
 &lt;/tr&gt;
 &lt;/tbody&gt;
&lt;/table&gt;
&lt;h2 id="market-implication"&gt;Market Implication
&lt;/h2&gt;&lt;p&gt;This is not a stock-picking note. At the macro level, the message is:&lt;/p&gt;
&lt;div class="highlight"&gt;&lt;pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;-webkit-text-size-adjust:none;"&gt;&lt;code class="language-text" data-lang="text"&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;AI can raise long-run earnings potential
&lt;/span&gt;&lt;/span&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;but if it also raises r-star and real rates
&lt;/span&gt;&lt;/span&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;the same future earnings deserve a lower multiple
&lt;/span&gt;&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;&lt;/div&gt;&lt;p&gt;That is the tension. AI can be good for growth and still bad for rate-cut hopes.&lt;/p&gt;
&lt;h2 id="final-view"&gt;Final View
&lt;/h2&gt;&lt;p&gt;AI may eventually become a 1996-style productivity disinflation force. But for this FOMC, the observable facts are closer to 1999: CapEx, power demand, input costs, financial conditions, and expectations are moving before economy-wide productivity disinflation is proven.&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;AI being good is not the same thing as rates going down.&lt;/strong&gt; Until productivity shows up in broad unit labor costs and core services inflation, the Fed is more likely to treat AI as a neutral-rate and investment-demand issue than as a reason to ease.&lt;/p&gt;
&lt;p&gt;Sources: &lt;a class="link" href="https://www.federalreserve.gov/newsevents/2026-june.htm" target="_blank" rel="noopener"
 &gt;Fed June calendar&lt;/a&gt;, &lt;a class="link" href="https://www.federalreserve.gov/newsevents/2026-july.htm" target="_blank" rel="noopener"
 &gt;Fed July calendar&lt;/a&gt;, &lt;a class="link" href="https://www.federalreserve.gov/monetarypolicy/fomcminutes20260429.htm" target="_blank" rel="noopener"
 &gt;April FOMC minutes&lt;/a&gt;, &lt;a class="link" href="https://www.federalreserve.gov/monetarypolicy/fomcprojtabl20260318.htm" target="_blank" rel="noopener"
 &gt;March SEP&lt;/a&gt;, &lt;a class="link" href="https://www.federalreserve.gov/newsevents/speech/cook20260224a.htm" target="_blank" rel="noopener"
 &gt;Governor Cook on AI&lt;/a&gt;, &lt;a class="link" href="https://www.bls.gov/news.release/cpi.nr0.htm" target="_blank" rel="noopener"
 &gt;BLS CPI&lt;/a&gt;, &lt;a class="link" href="https://www.bls.gov/news.release/empsit.nr0.htm" target="_blank" rel="noopener"
 &gt;BLS jobs&lt;/a&gt;, &lt;a class="link" href="https://www.bea.gov/news/2026/personal-income-and-outlays-april-2026" target="_blank" rel="noopener"
 &gt;BEA PCE&lt;/a&gt;, &lt;a class="link" href="https://academic.oup.com/qje/article/140/2/889/7990658" target="_blank" rel="noopener"
 &gt;QJE Generative AI at Work&lt;/a&gt;, &lt;a class="link" href="https://www.federalreserve.gov/econres/notes/feds-notes/monitoring-ai-adoption-in-the-u-s-economy-20260403.html" target="_blank" rel="noopener"
 &gt;Fed FEDS Notes&lt;/a&gt;, &lt;a class="link" href="https://www.kansascityfed.org/research/economic-bulletin/a-new-us-productivity-chapter-what-industry-data-say-about-ai/" target="_blank" rel="noopener"
 &gt;Kansas City Fed&lt;/a&gt;, &lt;a class="link" href="https://www.frbsf.org/research-and-insights/publications/economic-letter/2026/02/ai-moment-possibilities-productivity-policy/" target="_blank" rel="noopener"
 &gt;San Francisco Fed&lt;/a&gt;.&lt;/p&gt;</description></item><item><title>Is AI Productivity Real? From 15% Workplace Gains to Macro Diffusion</title><link>https://koreainvestinsights.com/post/ai-productivity-real-evidence-generative-ai-at-work-fed-2026-06-17/</link><pubDate>Wed, 17 Jun 2026 10:30:00 +0900</pubDate><guid>https://koreainvestinsights.com/post/ai-productivity-real-evidence-generative-ai-at-work-fed-2026-06-17/</guid><description>
 &lt;blockquote&gt;
 &lt;p&gt;Context&lt;br&gt;
This note follows our prior work on the AI supercycle, AI data-center capex, memory demand, and rate risk. The question here is more basic: &lt;strong&gt;is AI actually raising productivity, and is that effect already visible at the macro level?&lt;/strong&gt;&lt;/p&gt;

 &lt;/blockquote&gt;
&lt;h2 id="tldr"&gt;TL;DR
&lt;/h2&gt;&lt;ul&gt;
&lt;li&gt;The right answer is &lt;strong&gt;micro yes, macro not yet fully confirmed&lt;/strong&gt;.&lt;/li&gt;
&lt;li&gt;Brynjolfsson, Li, and Raymond&amp;rsquo;s &lt;em&gt;Generative AI at Work&lt;/em&gt; finds that AI assistance raised customer-support agents&amp;rsquo; productivity by about &lt;strong&gt;15%&lt;/strong&gt; on average, with the largest benefits for less experienced and lower-skilled workers.&lt;/li&gt;
&lt;li&gt;The Fed&amp;rsquo;s FEDS Notes shows that AI adoption is no longer trivial: about &lt;strong&gt;18%&lt;/strong&gt; of firms had adopted AI by year-end 2025, work-related GenAI usage among individuals was around &lt;strong&gt;41%&lt;/strong&gt;, and employment-weighted survey estimates suggest that &lt;strong&gt;78%&lt;/strong&gt; of workers were at AI-adopting firms and &lt;strong&gt;54%&lt;/strong&gt; at firms using LLMs.&lt;/li&gt;
&lt;li&gt;The Kansas City Fed finds that U.S. labor productivity has moved above its pre-pandemic trend since late 2022, but the pickup is not broad-based. AI adoption aligns with faster productivity growth across industries, but explains little of the aggregate shift so far.&lt;/li&gt;
&lt;li&gt;The San Francisco Fed&amp;rsquo;s policy lesson is that aggregate productivity, labor, and inflation data are not enough. Policymakers and investors need disaggregated evidence and business-level signals before the macro data fully reveal the transformation.&lt;/li&gt;
&lt;/ul&gt;
&lt;div class="thesis-callout"&gt;
 &lt;div class="thesis-callout__label"&gt;Core Takeaway&lt;/div&gt;
 &lt;div class="thesis-callout__body"&gt;
 AI productivity is not a yes-or-no debate. It is a diffusion timeline. The workplace evidence is real, adoption is rising quickly, but economy-wide productivity gains are still narrow and uneven.
 &lt;/div&gt;
&lt;/div&gt;
&lt;h2 id="the-four-sources-fit-together"&gt;The Four Sources Fit Together
&lt;/h2&gt;&lt;table&gt;
 &lt;thead&gt;
 &lt;tr&gt;
 &lt;th&gt;Source&lt;/th&gt;
 &lt;th&gt;Layer&lt;/th&gt;
 &lt;th&gt;Main question&lt;/th&gt;
 &lt;th&gt;Core answer&lt;/th&gt;
 &lt;/tr&gt;
 &lt;/thead&gt;
 &lt;tbody&gt;
 &lt;tr&gt;
 &lt;td&gt;&lt;em&gt;Generative AI at Work&lt;/em&gt;&lt;/td&gt;
 &lt;td&gt;Workplace task&lt;/td&gt;
 &lt;td&gt;Does AI raise productivity on the job?&lt;/td&gt;
 &lt;td&gt;Yes, in specific workflows&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Fed FEDS Notes&lt;/td&gt;
 &lt;td&gt;Adoption&lt;/td&gt;
 &lt;td&gt;How widely is AI used?&lt;/td&gt;
 &lt;td&gt;Meaningfully, but unevenly&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Kansas City Fed&lt;/td&gt;
 &lt;td&gt;Industry productivity&lt;/td&gt;
 &lt;td&gt;Is the U.S. productivity pickup AI-driven?&lt;/td&gt;
 &lt;td&gt;Related, but not fully explained by AI&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;San Francisco Fed&lt;/td&gt;
 &lt;td&gt;Policy&lt;/td&gt;
 &lt;td&gt;What should policymakers watch?&lt;/td&gt;
 &lt;td&gt;Micro and sector data, not only aggregates&lt;/td&gt;
 &lt;/tr&gt;
 &lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;The combined message is straightforward:&lt;/p&gt;
&lt;div class="highlight"&gt;&lt;pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;-webkit-text-size-adjust:none;"&gt;&lt;code class="language-text" data-lang="text"&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;AI raises productivity in real workplace settings
&lt;/span&gt;&lt;/span&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;→ adoption is spreading
&lt;/span&gt;&lt;/span&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;→ macro productivity gains are still narrow
&lt;/span&gt;&lt;/span&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;→ policy and investment should track diffusion quality
&lt;/span&gt;&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;&lt;/div&gt;&lt;h2 id="workplace-evidence-what-ai-actually-changed"&gt;Workplace Evidence: What AI Actually Changed
&lt;/h2&gt;&lt;p&gt;The QJE version of &lt;em&gt;Generative AI at Work&lt;/em&gt; studies 5,172 customer-support agents at a Fortune 500 firm using a GPT-3-based conversational assistant. The tool offered real-time response suggestions that agents could accept, edit, or ignore.&lt;/p&gt;
&lt;p&gt;The headline result is a roughly &lt;strong&gt;15%&lt;/strong&gt; increase in issues resolved per hour. But the deeper result is distributional. Lower-skilled and newer workers benefited most, while experienced high performers saw smaller speed gains and some quality trade-offs.&lt;/p&gt;
&lt;p&gt;This matters because AI did not merely make the best workers even better. It compressed skill gaps.&lt;/p&gt;
&lt;table&gt;
 &lt;thead&gt;
 &lt;tr&gt;
 &lt;th&gt;Finding&lt;/th&gt;
 &lt;th&gt;Economic meaning&lt;/th&gt;
 &lt;/tr&gt;
 &lt;/thead&gt;
 &lt;tbody&gt;
 &lt;tr&gt;
 &lt;td&gt;Larger gains for less experienced workers&lt;/td&gt;
 &lt;td&gt;Faster onboarding and lower training cost&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Text style shifted toward high-skill agents&lt;/td&gt;
 &lt;td&gt;Tacit knowledge became more transferable&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Customer sentiment improved&lt;/td&gt;
 &lt;td&gt;Quality and work experience also changed&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Stronger effects for moderately rare issues&lt;/td&gt;
 &lt;td&gt;AI helps when the firm has patterns but humans lack repeated exposure&lt;/td&gt;
 &lt;/tr&gt;
 &lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;This is not full automation. It is workflow augmentation. The first-order effect is not &amp;ldquo;remove the human&amp;rdquo;; it is &amp;ldquo;raise the average worker&amp;rsquo;s output and quality.&amp;rdquo;&lt;/p&gt;
&lt;h2 id="adoption-ai-is-already-inside-the-workplace"&gt;Adoption: AI Is Already Inside the Workplace
&lt;/h2&gt;&lt;p&gt;The Fed&amp;rsquo;s FEDS Notes reconciles three different survey lenses. The numbers vary because they use different units, but together they show that AI has moved beyond experimentation.&lt;/p&gt;
&lt;table&gt;
 &lt;thead&gt;
 &lt;tr&gt;
 &lt;th style="text-align: right"&gt;Estimate&lt;/th&gt;
 &lt;th&gt;Unit&lt;/th&gt;
 &lt;th&gt;Interpretation&lt;/th&gt;
 &lt;/tr&gt;
 &lt;/thead&gt;
 &lt;tbody&gt;
 &lt;tr&gt;
 &lt;td style="text-align: right"&gt;18%&lt;/td&gt;
 &lt;td&gt;Firm count&lt;/td&gt;
 &lt;td&gt;Share of firms reporting AI adoption at year-end 2025&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td style="text-align: right"&gt;41%&lt;/td&gt;
 &lt;td&gt;Individuals&lt;/td&gt;
 &lt;td&gt;Workers reporting work-related GenAI usage&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td style="text-align: right"&gt;78%&lt;/td&gt;
 &lt;td&gt;Employment-weighted firms&lt;/td&gt;
 &lt;td&gt;Workers employed at AI-adopting firms&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td style="text-align: right"&gt;54%&lt;/td&gt;
 &lt;td&gt;Employment-weighted LLM firms&lt;/td&gt;
 &lt;td&gt;Workers employed at firms using LLMs&lt;/td&gt;
 &lt;/tr&gt;
 &lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;The high employment-weighted numbers matter. Large firms employ many workers, so AI adoption by large organizations can reach a large share of the workforce even if the firm-count adoption rate looks modest.&lt;/p&gt;
&lt;p&gt;The industry pattern also matters. AI adoption is strongest in finance and professional services, the high-value cognitive work sectors where research, analysis, compliance, coding, legal drafting, accounting, and client communication are core workflows.&lt;/p&gt;
&lt;h2 id="macro-productivity-promising-but-not-broad-based-yet"&gt;Macro Productivity: Promising but Not Broad-Based Yet
&lt;/h2&gt;&lt;p&gt;The Kansas City Fed provides the necessary caution. U.S. labor productivity has strengthened since late 2022 and moved above its pre-pandemic trend. That period overlaps with the commercial emergence of widely used generative AI tools.&lt;/p&gt;
&lt;p&gt;But the pickup is not broad-based. A relatively small set of industries accounts for much of the improvement. Higher AI adoption is associated with faster productivity growth across industries, but it explains little of the aggregate productivity shift so far.&lt;/p&gt;
&lt;p&gt;That is not a bearish conclusion. It is an early-diffusion conclusion.&lt;/p&gt;
&lt;div class="highlight"&gt;&lt;pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;-webkit-text-size-adjust:none;"&gt;&lt;code class="language-text" data-lang="text"&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;Task-level gains
&lt;/span&gt;&lt;/span&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;→ firm workflow redesign
&lt;/span&gt;&lt;/span&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;→ industry-level adoption
&lt;/span&gt;&lt;/span&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;→ macro productivity data
&lt;/span&gt;&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;&lt;/div&gt;&lt;p&gt;The economy is still moving through the middle of that chain.&lt;/p&gt;
&lt;h2 id="policy-lesson-do-not-wait-for-aggregates"&gt;Policy Lesson: Do Not Wait for Aggregates
&lt;/h2&gt;&lt;p&gt;The San Francisco Fed frames the AI moment through the historical lens of the 1990s productivity acceleration. The lesson is not that AI will repeat the 1990s exactly. The lesson is that aggregate data often lag transformative business change.&lt;/p&gt;
&lt;p&gt;For monetary policy, AI is complicated because it can create short-term demand pressure and long-term supply benefits at the same time.&lt;/p&gt;
&lt;table&gt;
 &lt;thead&gt;
 &lt;tr&gt;
 &lt;th&gt;Channel&lt;/th&gt;
 &lt;th&gt;Short-term effect&lt;/th&gt;
 &lt;th&gt;Long-term possibility&lt;/th&gt;
 &lt;/tr&gt;
 &lt;/thead&gt;
 &lt;tbody&gt;
 &lt;tr&gt;
 &lt;td&gt;Data-center capex&lt;/td&gt;
 &lt;td&gt;Raises demand for power, equipment, land, financing&lt;/td&gt;
 &lt;td&gt;Builds productive capacity&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;AI tools&lt;/td&gt;
 &lt;td&gt;Implementation costs and disruption&lt;/td&gt;
 &lt;td&gt;Lower unit labor cost&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Labor market&lt;/td&gt;
 &lt;td&gt;Task displacement and job redesign&lt;/td&gt;
 &lt;td&gt;Higher output per worker&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Inflation&lt;/td&gt;
 &lt;td&gt;Infrastructure bottlenecks&lt;/td&gt;
 &lt;td&gt;Disinflation from productivity&lt;/td&gt;
 &lt;/tr&gt;
 &lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;This is why the Fed cannot simply wait for aggregate productivity to settle the debate. It needs firm-level and industry-level signals.&lt;/p&gt;
&lt;h2 id="investment-read-through"&gt;Investment Read-Through
&lt;/h2&gt;&lt;p&gt;For investors, the key implication is that AI capex can still be justified, but not by model benchmarks alone. It needs workflow productivity.&lt;/p&gt;
&lt;p&gt;Positive evidence:&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;real task-level productivity gains&lt;/li&gt;
&lt;li&gt;rising work-related adoption&lt;/li&gt;
&lt;li&gt;early penetration into high-value cognitive sectors&lt;/li&gt;
&lt;li&gt;signs of tacit knowledge transfer and faster onboarding&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;Remaining risks:&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;usage does not equal value creation&lt;/li&gt;
&lt;li&gt;workflow redesign takes time&lt;/li&gt;
&lt;li&gt;aggregate productivity gains remain narrow&lt;/li&gt;
&lt;li&gt;capex may outrun proven ROI&lt;/li&gt;
&lt;li&gt;quality, compliance, and liability issues can slow adoption&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;For Korea, this supports the long-term logic behind AI infrastructure: HBM, server DRAM, eSSD, networking, power equipment, data-center construction, AI cloud, and enterprise workflow software. But it also raises the bar. The market should reward companies that turn AI into repeatable productivity workflows, not companies that merely attach AI labels to old businesses.&lt;/p&gt;
&lt;h2 id="final-view"&gt;Final View
&lt;/h2&gt;&lt;p&gt;AI productivity is real at the task level. Adoption is high enough to matter. But the macro productivity revolution is not yet fully proven.&lt;/p&gt;
&lt;p&gt;That is exactly why the next phase of the AI cycle is about diffusion, not demo performance. The question is no longer whether AI can answer questions. The question is whether it changes workflows deeply enough to show up in margins, output per worker, and eventually the national productivity data.&lt;/p&gt;
&lt;h2 id="sources"&gt;Sources
&lt;/h2&gt;&lt;ul&gt;
&lt;li&gt;&lt;a class="link" href="https://academic.oup.com/qje/article/140/2/889/7990658" target="_blank" rel="noopener"
 &gt;Brynjolfsson, Li, Raymond, &lt;em&gt;Generative AI at Work&lt;/em&gt;, QJE&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a class="link" href="https://www.frbsf.org/research-and-insights/publications/economic-letter/2026/02/ai-moment-possibilities-productivity-policy/" target="_blank" rel="noopener"
 &gt;San Francisco Fed, &lt;em&gt;The AI Moment? Possibilities, Productivity, and Policy&lt;/em&gt;&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a class="link" href="https://www.kansascityfed.org/research/economic-bulletin/a-new-us-productivity-chapter-what-industry-data-say-about-ai/" target="_blank" rel="noopener"
 &gt;Kansas City Fed, &lt;em&gt;A New U.S. Productivity Chapter? What Industry Data Say About AI&lt;/em&gt;&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a class="link" href="https://www.federalreserve.gov/econres/notes/feds-notes/monitoring-ai-adoption-in-the-u-s-economy-20260403.html" target="_blank" rel="noopener"
 &gt;Federal Reserve FEDS Notes, &lt;em&gt;Monitoring AI Adoption in the U.S. Economy&lt;/em&gt;&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;</description></item><item><title>AI Supercycle Midgame: Rate Risk Is Rising, But the Red Flags Are Not Yet Observed</title><link>https://koreainvestinsights.com/post/ai-supercycle-midgame-rate-risk-yellow-not-red-2026-06-06/</link><pubDate>Sat, 06 Jun 2026 19:30:00 +0900</pubDate><guid>https://koreainvestinsights.com/post/ai-supercycle-midgame-rate-risk-yellow-not-red-2026-06-06/</guid><description>
 &lt;blockquote&gt;
 &lt;p&gt;Context: This note follows &lt;a class="link" href="https://koreainvestinsights.com/post/ai-datacenter-capex-5p3t-korea-power-substrate-storage-bottleneck-2026-06-05/" &gt;the $5.3T AI data-center CapEx map&lt;/a&gt;, &lt;a class="link" href="https://koreainvestinsights.com/post/us-jobs-rate-hike-shock-kospi-macro-gate-2026-06-06/" &gt;the U.S. jobs shock / KOSPI 8,000 gate&lt;/a&gt;, &lt;a class="link" href="https://koreainvestinsights.com/post/jensen-huang-hbm4-three-vendors-qualified-market-missed-2026-06-05/" &gt;Jensen Huang&amp;rsquo;s HBM4 three-vendor qualification remark&lt;/a&gt;, and &lt;a class="link" href="https://koreainvestinsights.com/post/sam-hama-parity-follow-up-ai-chip-memory-pe-map-2026-06-05/" &gt;the Samsung-Hynix-Micron parity follow-up&lt;/a&gt;.&lt;/p&gt;

 &lt;/blockquote&gt;
&lt;h2 id="tldr"&gt;TL;DR
&lt;/h2&gt;&lt;ul&gt;
&lt;li&gt;The bearish counterargument is valid. But the observed fact is not an AI-demand collapse. It is that higher rates and capital costs are starting to cap AI infrastructure investment expectations and AI-stock multiples.&lt;/li&gt;
&lt;li&gt;The current regime is &lt;strong&gt;yellow, not red&lt;/strong&gt;. Long-rate pressure, more leveraged data-center financing structures and hyperscaler FCF pressure are visible.&lt;/li&gt;
&lt;li&gt;The red flags are not yet clearly visible: hyperscaler AI capex cuts, HBM order cancellations, DRAM/NAND contract-price rollovers, rising data-center vacancy, and lease-backed financing failures.&lt;/li&gt;
&lt;li&gt;Low forward P/E ratios in memory stocks are helpful, but not sufficient. In memory, low P/E can also mean the market is discounting peak-cycle earnings.&lt;/li&gt;
&lt;/ul&gt;
&lt;div class="thesis-callout"&gt;
 &lt;div class="thesis-callout__label"&gt;Core Point&lt;/div&gt;
 &lt;div class="thesis-callout__body"&gt;
 The AI supercycle has not broken. What we can observe is a yellow light: rates, capital costs and data-center financing are starting to cap the upside for AI infrastructure multiples.
 &lt;/div&gt;
&lt;/div&gt;
&lt;h2 id="1-the-question"&gt;1. The Question
&lt;/h2&gt;&lt;p&gt;This is not a simple &amp;ldquo;AI is real&amp;rdquo; versus &amp;ldquo;AI is a bubble&amp;rdquo; debate. AI demand is showing up in earnings. Micron, Samsung Electronics and SK Hynix have all pointed to AI-related memory demand, HBM, server DRAM and enterprise SSD strength. (&lt;a class="link" href="https://investors.micron.com/static-files/9c0becf5-df56-4eec-bd67-453dda68b273" title="Micron FY2026 Q2 presentation"
 target="_blank" rel="noopener"
 &gt;Micron&lt;/a&gt;, &lt;a class="link" href="https://news.samsung.com/be/samsung-electronics-announces-first-quarter-2026-results" title="Samsung Electronics Announces First Quarter 2026 Results"
 target="_blank" rel="noopener"
 &gt;Samsung&lt;/a&gt;, &lt;a class="link" href="https://www.reuters.com/world/asia-pacific/nvidia-supplier-sk-hynix-q1-profit-rises-406-meets-forecasts-2026-04-22/" title="Reuters - SK Hynix Q1 2026 profit and AI chip demand"
 target="_blank" rel="noopener"
 &gt;Reuters/SK Hynix&lt;/a&gt;)&lt;/p&gt;
&lt;p&gt;The real question is different:&lt;/p&gt;

 &lt;blockquote&gt;
 &lt;p&gt;Even if AI is real, can higher capital costs force ROI discipline and pressure equity multiples, data-center projects and eventually chip-order expectations?&lt;/p&gt;

 &lt;/blockquote&gt;
&lt;p&gt;Yes, that path is real. But it is not yet a confirmed red-light scenario.&lt;/p&gt;
&lt;h2 id="2-what-is-visible"&gt;2. What Is Visible
&lt;/h2&gt;&lt;table&gt;
 &lt;thead&gt;
 &lt;tr&gt;
 &lt;th&gt;Signal&lt;/th&gt;
 &lt;th&gt;Status&lt;/th&gt;
 &lt;th&gt;Interpretation&lt;/th&gt;
 &lt;/tr&gt;
 &lt;/thead&gt;
 &lt;tbody&gt;
 &lt;tr&gt;
 &lt;td&gt;U.S. 10-year near 4.5% and 30-year near 5%&lt;/td&gt;
 &lt;td&gt;Visible&lt;/td&gt;
 &lt;td&gt;Higher discount rates pressure AI and growth multiples. (&lt;a class="link" href="https://fred.stlouisfed.org/series/DGS10" title="FRED - 10-Year Treasury Constant Maturity Rate"
 target="_blank" rel="noopener"
 &gt;FRED&lt;/a&gt;)&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Data-center financing complexity&lt;/td&gt;
 &lt;td&gt;Visible&lt;/td&gt;
 &lt;td&gt;Debt, JV, lease-backed financing, private credit and CMBS structures are increasingly important. (&lt;a class="link" href="https://www.cbre.com/insights/books/north-america-data-center-trends-h2-2025" title="CBRE - North America Data Center Trends H2 2025"
 target="_blank" rel="noopener"
 &gt;CBRE&lt;/a&gt;, &lt;a class="link" href="https://www.jll.com/en-us/insights/market-dynamics/north-america-data-centers" title="JLL - North America Data Center Report Year-end 2025"
 target="_blank" rel="noopener"
 &gt;JLL&lt;/a&gt;)&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Hyperscaler FCF pressure&lt;/td&gt;
 &lt;td&gt;Visible&lt;/td&gt;
 &lt;td&gt;AI capex is large enough to affect free cash flow and balance-sheet choices.&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Data-center vacancy / demand collapse&lt;/td&gt;
 &lt;td&gt;Not visible&lt;/td&gt;
 &lt;td&gt;North American vacancy and pre-commitment data still point to supply tightness. (&lt;a class="link" href="https://www.jll.com/en-us/insights/market-dynamics/north-america-data-centers" title="JLL - North America Data Center Report Year-end 2025"
 target="_blank" rel="noopener"
 &gt;JLL&lt;/a&gt;)&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;HBM / DRAM order stress&lt;/td&gt;
 &lt;td&gt;Not visible&lt;/td&gt;
 &lt;td&gt;The current evidence still points to tight AI-memory supply.&lt;/td&gt;
 &lt;/tr&gt;
 &lt;/tbody&gt;
&lt;/table&gt;
&lt;h2 id="3-memory-pe-is-not-enough"&gt;3. Memory P/E Is Not Enough
&lt;/h2&gt;&lt;p&gt;Samsung Electronics, SK Hynix and Micron trade at low forward P/E levels relative to many AI-chip names. That is attractive, especially if HBM and server-memory earnings persist.&lt;/p&gt;
&lt;p&gt;But memory is cyclical. A low forward P/E can mean two things:&lt;/p&gt;
&lt;table&gt;
 &lt;thead&gt;
 &lt;tr&gt;
 &lt;th&gt;Interpretation&lt;/th&gt;
 &lt;th&gt;Meaning&lt;/th&gt;
 &lt;/tr&gt;
 &lt;/thead&gt;
 &lt;tbody&gt;
 &lt;tr&gt;
 &lt;td&gt;Structural upside is underpriced&lt;/td&gt;
 &lt;td&gt;Bullish if HBM / DDR5 / eSSD demand stays tight&lt;/td&gt;
 &lt;/tr&gt;
 &lt;tr&gt;
 &lt;td&gt;Current EPS is peak-cycle EPS&lt;/td&gt;
 &lt;td&gt;Risky if prices roll over or customers over-ordered&lt;/td&gt;
 &lt;/tr&gt;
 &lt;/tbody&gt;
&lt;/table&gt;
&lt;p&gt;The practical test is EPS durability: HBM allocation, contract pricing, DRAM/NAND contract prices, customer inventory, supply capex and 2027-2028 oversupply risk.&lt;/p&gt;
&lt;h2 id="4-the-rate-transmission-channel"&gt;4. The Rate Transmission Channel
&lt;/h2&gt;&lt;div class="highlight"&gt;&lt;pre tabindex="0" style="color:#f8f8f2;background-color:#272822;-moz-tab-size:4;-o-tab-size:4;tab-size:4;-webkit-text-size-adjust:none;"&gt;&lt;code class="language-text" data-lang="text"&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;Higher long rates
&lt;/span&gt;&lt;/span&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;→ lower AI / growth multiples
&lt;/span&gt;&lt;/span&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;→ higher data-center financing costs
&lt;/span&gt;&lt;/span&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;→ tougher project IRR and ROI discipline
&lt;/span&gt;&lt;/span&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;→ slower marginal AI-campus / neocloud buildout
&lt;/span&gt;&lt;/span&gt;&lt;span style="display:flex;"&gt;&lt;span&gt;→ lower growth expectations for GPU / HBM / SSD orders
&lt;/span&gt;&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;&lt;/div&gt;&lt;p&gt;The first two steps are already visible. The last two are not yet confirmed.&lt;/p&gt;
&lt;h2 id="5-korea-translation"&gt;5. Korea Translation
&lt;/h2&gt;&lt;p&gt;For Korean equities, the framework is:&lt;/p&gt;
&lt;ul&gt;
&lt;li&gt;Samsung Electronics and SK Hynix remain the cleanest large-cap AI-memory exposure, but the key is EPS durability, not just low P/E.&lt;/li&gt;
&lt;li&gt;Samsung Electro-Mechanics, Hanmi Semiconductor, Gigavis and other AI-infrastructure bottleneck names must prove orders, margins and customer diversification.&lt;/li&gt;
&lt;li&gt;Power equipment and data-center infrastructure are AI beneficiaries, but they are also closer to the rate-sensitive financing layer.&lt;/li&gt;
&lt;/ul&gt;
&lt;h2 id="final-view"&gt;Final View
&lt;/h2&gt;&lt;p&gt;The AI semiconductor rally is still earnings-backed. But the durability of those earnings is increasingly tied to rates and data-center financing conditions.&lt;/p&gt;
&lt;p&gt;This is not the end of the AI bull thesis. It is the moment when the rate bear thesis becomes a monitorable risk.&lt;/p&gt;
&lt;h2 id="fact--inference--blocked"&gt;Fact / Inference / Blocked
&lt;/h2&gt;&lt;ul&gt;
&lt;li&gt;[Fact] Long rates, data-center financing structures and AI-memory earnings data support the yellow-light framing. (&lt;a class="link" href="https://fred.stlouisfed.org/series/DGS10" title="FRED - 10-Year Treasury Constant Maturity Rate"
 target="_blank" rel="noopener"
 &gt;FRED&lt;/a&gt;, &lt;a class="link" href="https://www.cbre.com/insights/books/north-america-data-center-trends-h2-2025" title="CBRE - North America Data Center Trends H2 2025"
 target="_blank" rel="noopener"
 &gt;CBRE&lt;/a&gt;, &lt;a class="link" href="https://www.jll.com/en-us/insights/market-dynamics/north-america-data-centers" title="JLL - North America Data Center Report Year-end 2025"
 target="_blank" rel="noopener"
 &gt;JLL&lt;/a&gt;, &lt;a class="link" href="https://investors.micron.com/static-files/9c0becf5-df56-4eec-bd67-453dda68b273" title="Micron FY2026 Q2 presentation"
 target="_blank" rel="noopener"
 &gt;Micron&lt;/a&gt;)&lt;/li&gt;
&lt;li&gt;[Inference] Rate risk will likely show up first in data-center developers, colocation, neocloud, GPU cloud and leveraged AI infrastructure, before core memory suppliers.&lt;/li&gt;
&lt;li&gt;[Blocked] Customer-level HBM4 allocation, contract pricing, AI data-center project IRR and lease-backed financing spreads are not fully public.&lt;/li&gt;
&lt;/ul&gt;
&lt;p&gt;&lt;em&gt;Disclaimer: Research and information only. Not investment advice.&lt;/em&gt;&lt;/p&gt;</description></item></channel></rss>