● Short Storm

Michael Burry, Not Retiring but ‘Changing His Position’. The True Scale of Palantir Put Options, the Intention Behind the Early 13F, and the Core of the 6-Year GPU Depreciation Debate and Investment Strategy.

This article contains four topics.

1) What Michael Burry’s retirement declaration actually means and his strategy shift going forward.

2) Clarifying the misunderstanding between “notional vs. actual expenditure” regarding Palantir (PLTR) put options.

3) Reinterpreting the GPU 6-year depreciation debate from the perspectives of accounting and data center operations.

4) AI computing cost per unit, demand scenarios, and an investment strategy checklist.

We have connected the global economy, interest rates, inflation, the stock market, and investment strategies in a news format for you.

[Breaking] Burry “Liquidates Fund”. Signal of Transition to a Family Office.

Burry announced the liquidation of his fund and the return of capital, effectively ceasing the management of external funds.

This signals that he intends to operate his positions at his own pace, free from regulatory obligations and quarterly reporting pressures.

Many interpret this as a prelude to maintaining contrarian positions for the long term, similar to his “Big Short” days.

The key point is not ‘retirement’ but a ‘change of playground’.

Palantir Put Options, Setting the Record Straight: 9.12M vs. 912M.

Burry revealed that he bought 50,000 put option contracts on Palantir at approximately $1.84 per contract.

Since each option contract typically represents 100 shares, the actual cash expenditure is about $9.12 million.

The misunderstood “hundreds of millions of dollars” figure represents only the notional value (number of shares × share price) of the underlying asset, not the premium paid.

The point is that it is a directional bet with leverage, and the cash deployed is relatively small compared to holding the underlying asset.

Based on the strike price and expiration (Burry mentioned a 2027 maturity with a $50 put), this represents a long-term hedge or aggressive short position focused on the downside scenario.

The profit/loss will be largely influenced by the price path until expiration, volatility, time value (theta), and implied volatility (vega).

The Unusual Early 13F Filing: Why Was It Disclosed ‘Early’.

Most institutions delay exposing their positions until just before the deadline.

Burry disclosed his positions ahead of the deadline.

There are two possible interpretations.

1) Increased freedom in communication due to the cessation of external fund management.

2) A strategic disclosure aimed at controlling the narrative in the market battle.

In other words, it can be seen as the beginning of a psychological warfare preparing for a long-term short or structural bearish bet.

The 6-Year GPU Depreciation Debate: The Intersection of Accounting and Facility Operations.

The controversy centers on the gap between the “technological lifecycle of AI semiconductors (2–3 years?)” and the “economic useful life of data center assets (5–6 years?)”.

Burry pointed out that extending the depreciable life can superficially inflate short-term earnings.

The counterarguments are as follows.

– A data center includes more than just GPUs.

– Systems that include racks, networks, storage, cooling, and power infrastructure typically have a lifecycle of 5–7 years.

– Even if older GPUs are supplanted in the first tier for training, they can be redeployed for inference, graphics/media, and enterprise workloads.

– Cloud providers monetize older instances as products over an extended period.

Therefore, the accounting depreciation period of 5–6 years can be justified by actual operational practices.

However, some cautions remain.

– If the performance per watt gap between new and old GPUs widens significantly, margins may be diluted under the same pricing.

– If the redeployment of surplus GPUs fails, their residual value declines, and longer depreciation periods may pressure ROIC.

The conclusion is that it is not merely the “number of years” but the “reassignment rate, utilization rate, and pricing policy” that determine the quality of earnings.

How Many GPU Training Sessions Are ‘Optimal’: The Triangle of Demand, Unit Price, and Revenue Model.

The key question is “How many GPUs need to be deployed for the model/service to turn profitable?”

The evaluation framework is briefly outlined as follows.

– On the price aspect.

Combine the GPU hourly cost (on-demand/reserved) × total training hours with data and engineering expenses.

The model size (parameters), token count (training/fine-tuning), and distributed efficiency determine the total cost.

– On the demand aspect.

Estimate the revenue potential based on post-launch DAU/MAU, API call volumes, token usage, and average revenue per user (ARPU).

– On the revenue model.

SaaS subscriptions, API billing, on-premise licensing, and enhanced advertising/search revenue must have a clear cash flow structure.

The term ‘optimal’ means that the scale results in a positive ROI.

Thus, even though initial large-scale training is inevitable, cost recovery must be achievable through unit cost reductions and high reuse rates during the inference phase.

Key financial metrics include CAC versus LTV, margin per token, and the pace of decline in the inference cost (prompt+generated token cost).

Connecting Macroeconomic Variables: The Impact on the Global Economy, Interest Rates, Inflation, and the Stock Market.

If interest rates rise, the present value of assets with long durations is discounted, and the capital cost for data center CAPEX increases.

If inflation becomes entrenched, costs for power, servers, and labor rise, increasing the fixed cost burden of AI infrastructure.

A slowdown in global economic growth enhances the cyclical and discretionary nature of IT spending, affecting order visibility.

Conversely, a decline in interest rates and a disinflationary phase can drive multiple expansion and CAPEX reactivation, restoring risk asset appetite in the stock market.

In short, even if the AI trend remains robust, capital cost structures determine margins.

What Burry’s Bet Indicates: An Investment Strategy Checklist.

– Quality of Earnings.

Check for disclosures on depreciation period, residual value, inventory/refurbishment, and utilization rates.

– Cash Flow.

Examine the gap between operating profits, inflated by non-cash depreciation, and free cash flow (FCF).

– Customer Mix.

Determine whether the focus is on high-end training or diversified large-scale inference/enterprise workloads.

– Supply Chain Constraints.

Track potential bottlenecks such as GPUs, network switches, CoWoS, HBM memory, and power/cooling infrastructure.

– Valuation/Risk Management.

Utilize a mix of long-dated put/call spreads and volatility trading strategies to hedge in overvalued ranges.

– Event Risks.

Monitor policy regulations, export controls, data sovereignty issues, and potential privacy/copyright lawsuits.

News in a Nutshell.

Burry has freed himself from the shackles of external funds and entered a long-term mode.

The Palantir put options represent a leveraged strategy that takes a significant downside bet with a limited cash outlay, rather than an enormous notional exposure.

The 6-year GPU depreciation, despite accusations of being an accounting gimmick, is a defensible choice depending on data center operations and revenue models; ultimately, cash flow and pricing policy are key.

Key Points Other YouTube/News Outlets May Have Missed.

– It’s not about the ‘number of years’ but rather how effectively GPUs are redeployed and their utilization rates that determine the quality of earnings.

– The turning point for AI revenue models is the pace at which inference costs decline and repeated enterprise/API billing.

Earnings from inference, rather than training, ultimately determine total returns.

– The early 13F filing is a psychological maneuver aimed at dominating the narrative.

Burry is not just setting a price but also controlling the discourse.

– Accounting quality checks should extend beyond individual companies and include cross-checks across the value chain.

Monitoring GPU supply, module assembly, power infrastructure, and cloud productization minimizes distortions.

Timeline Guidance and Points to Watch.

– Quarterly Earnings Season.

Monitor changes in AI CAPEX guidance, depreciation policies, FCF, and instance mix.

– Product Roadmap.

Check whether the release intervals and performance-per-watt improvements of next-generation GPUs/accelerators are reflected in financial assumptions.

– Policies/Regulations.

Export controls, power infrastructure approvals, and data regulation issues can increase value chain volatility.

– Market Perspective.

Interest rate trends, inflation outlook, and global economic indicators have a direct impact on market multiples.

Portfolio Action Guidelines.

– Core/Exploration Separation.

Maintain a separation between core holdings with verified cash flows and exploratory positions in high-growth AI.

– Hedge Layering.

Manage volatility risk in overvalued phases by utilizing staggered expiration put/call spreads.

– Chain Diversification.

Diversify across each segment: GPUs, memory, networking, power, cooling, software, and services.

– Metrics Follow-Up.

Consistently track utilization rates, inventory days, lease/subscription retention, and margin per token.

Data at a Glance: Glossary.

– Notional Value.

The total scale of the underlying asset that the option can affect.

– Premium.

The actual cost that the option buyer has paid.

– Useful Life.

The period for accounting depreciation, which may or may not coincide with the asset’s actual economic use.

– Inference vs. Training.

Training is capital intensive, while inference is driven by economies of scale and cost reductions.

Conclusion.

Burry is not ‘retiring’ but is changing the rules to prepare for a long-term battle.

When stripped of the illusion of inflated option numbers, his bet is a leveraged strategy that takes significant downside risk with limited cash.

The 6-year GPU depreciation is a defensible choice depending on operational capabilities, and ultimately, cash flow and pricing policy will decide the outcome.

What is needed now are not just compelling stories but key metrics.

FCF, utilization rates, inference costs, customer mix, and interest rate trends will provide the answers.

< Summary >

Burry has ditched external funds and opted for a long-term, family office-style approach.

The Palantir put options represent a leveraged strategy with a $9.12 million cash outlay, betting big on the notional downside.

The 6-year GPU depreciation, despite being criticized as an accounting gimmick, relies on actual redeployment, utilization, and pricing policies.

Investment strategies should be realigned around the quality of earnings, FCF, CAPEX recovery, and interest rate trends.

[Related Articles…]

The Truth About the NVIDIA Cycle and Data Center CAPEX

Global Interest Rate Outlook 2025 and Its Impact on the Stock Market

*Source: [ 내일은 투자왕 – 김단테 ]

– 마이클 버리 은퇴 선언? 아니, 그는 ‘더 큰 것’을 준비 중!

---

● Shutdown Shock, Debt Flood, Liquidity Squeeze, AI Hype

After Shutdown: ‘Supreme Court Ruling → Government Bond Issuance → Liquidity Drive’ and Political Volatility Driven until the 2026 Midterm Elections, the True Nature of the AI Bubble Theory

In this article, it is presented in a news format that the cause of the short-term slump is not the ‘AI bubble’ but rather ‘political uncertainty.’ It specifically links the large-scale refund possibility following the Supreme Court ruling to the government bond issuance and the subsequent re-liquidity pathway. It outlines where the liquidity-driven market may falter if inflation reignites at 5–6%, and provides the most realistic trigger and checklist. It separates the fundamentals of the AI trend from the mechanism of M7 volatility, and even touches on the often-overlooked ‘plumbing’ of liquidity supply.

Today’s News Summary

• The primary cause of the market plunge is not the collapse of the AI bubble but the expanded ‘policy risk premium’ due to the fallout from the shutdown and the uncertainty following the Supreme Court ruling.
• The possibility of an invalidation of Trump’s tariffs is emerging, giving rise to a scenario where large-scale refunds and government bond issuance for this purpose may change the liquidity pathway.
• The shutdown has created a data blackout in official employment and inflation figures, leading to increased estimation errors in the Fed’s rate policy and concentrated risk aversion in both stock and crypto markets.
• Even though the motive for the ‘liquidity drive’ ahead of the 2026 midterm elections is clear, a re-ignition of inflation into the 5–6% range could prematurely end the drive.
• While the AI-related M7 stands as a representative beneficiary of the liquidity market, it also exhibits a dual character as the first target for profit-taking during corrections.

Political Uncertainty vs. AI Bubble Theory: What is Shaking the Stock Market?

• Key Point: The short-term correction is not due to the collapse of the AI bubble, but is driven by risk aversion amid policy volatility.
• The ‘Excuse Effect’ of the bubble theory: In a downturn, the narrative of the ‘bubble’ serves merely as a rhetoric to justify selling, while the actual trigger is a political event.
• Structure of Volatility: Even as the liquidity-driven market persists, policy headlines unsettle the premium, creating a rollercoaster effect.
• Investor Behavior: The top-rising large tech stocks (M7) become the “first target for profit-taking,” and the narrative of the downturn is wrapped in the guise of an AI bubble.

Supreme Court Ruling → Refund Shock → Government Bond Issuance → The Actual Path of Liquidity Re-supply

• The Framework of Assumptions: If Trump’s tariffs are invalidated under the logic of the abuse of emergency powers such as IEEPA, a large-scale refund issue for corporations could occur.
• Fiscal Response: To secure the funds for refunds, the Treasury will inevitably resort to additional government bond issuance.
• The Importance of ‘Plumbing’ Details:
• If the issuance is centered on short-term bonds, funds from money market funds move from RRP to government bonds, resulting in either neutral or slightly positive impacts on bank reserves’ liquidity.
• If there is a larger proportion of medium-to-long-term bonds, it would require banks and institutions to absorb the duration risk, causing a temporary tightening of financial conditions as reserves move out.
• While the refund increases private cash flow and may boost inflows into risk assets like stocks and crypto, the combination and timing of the issuance can create an asymmetry of ‘temporary tightening → gradual easing.’
• Policy Chain:
• As the termination of QT draws near (assuming an end), the sensitivity of reserves is mitigated, and there is room to deploy liquidity bridges using SRF or temporary operations if needed.
• While the motive behind the liquidity drive is to foster a risk-on environment until the 2026 midterm elections, only inflation serves as the brake.

Aftermath of the Shutdown: Data Blackout and the Fed’s Policy Uncertainty

• Data Blackout: With delays in official employment and inflation indicators, the Fed must rely on estimates, and the market becomes overly sensitive to alternative private indicators.
• Weakening Policy Signals: Uncertainty in rate decisions increases, the dot plots and forward guidance lose weight, and volatility intensifies.
• Market Behavior: In this phase of ‘information asymmetry,’ cash holding increases and hedging demands in the stock and crypto markets often result in overreactions.

Trigger for Premature End if Inflation Reignites at 5–6%

• Risk Triggers:
• Should the tariff structure change or if item-specific tariffs (Plan B) are introduced, it could deteriorate importers’ margins → lead to price pass-through → and cause secondary ripple effects across services.
• If oil prices remain above $90, coupled with re-pricing of freight and insurance, and persistent wage increases, core inflation could rise.
• Slow-moving items such as housing costs, insurance, and medical expenses exhibit downward rigidity.
• Trigger Levels:
• If headline inflation re-accelerates into the 5–6% range, or if the 5-year/10-year Breakeven Inflation Rates (BEI) structurally settle in the high 3% range, the liquidity drive will halt.
• In this scenario, expectations for rate cuts will retreat, and maintaining liquidity neutrality takes precedence over restarting QE.

Base, Upside, and Downside Scenarios until the 2026 Midterms

• Base (Highest Probability):
• The Supreme Court ruling and shutdown fallout are resolved within the first half of 2025, with a shift toward a higher proportion of short-term bonds in the issuance mix.
• Gradual termination of QT, stable reserves, and a recovery of the liquidity-driven market.
• The stock market style remains centered on big tech, but condensed into those that visibly benefit from AI.
• Upside:
• A refund and issuance combination that is market-friendly, stable oil prices, slowing wage growth, and anchored inflation expectations.
• An accelerated timeline for rate cuts and a resumption of re-rating multiples.
• Downside:
• Negative shocks from Plan B tariffs on specific items, a rebound in oil prices, and sticky service inflation.
• A reinvigoration of inflation into the 5–6% range that prematurely ends the liquidity drive.
• Expansion of duration losses and a re-adjustment of stock market valuations.

AI Trend: Not a Bubble but an ‘Early S-Curve’ with Downward Cost Trends

• From CapEx to OpEx: The period 2024–2026 marks the phase when hyperscalers hit the peak in accelerator and network investments, followed by a transition where enterprises move to managed AI services.
• Clues for Profitability:
• If data center power and cooling constraints become bottlenecks, unit token costs will decline through model simplification and the adoption of low-precision computations (DSP/token optimization).
• As AI functionalities become embedded in search, commerce, and SaaS, ARPU gradually improves, and a hybrid of advertising and subscription revenues increases.
• M7 Volatility:
• It is an excess beneficiary in a liquidity-driven market.
• During corrections, it becomes the ‘first winner to be cashed out’ due to its liquidity, intensifying its volatility.
• Selection Criteria:
• Filter by an increasing AI revenue ratio, transition from GPU to ASIC roadmaps, confident upside management of power agreements, growth in RPO (unbilled orders), and FCF coverage.

Investment Checklist: 7 Things to Check Right Now

• Check the short-term/long-term mix and the refinancing calendar in the Treasury’s refund and issuance notices.
• Monitor both the RRP balance and the flow of reserves to capture the transition from ‘temporary tightening → gradual easing.’
• Keep an eye on oil prices, diesel spreads, global freight indices, and the detailed CPI items for health insurance.
• Use BEI (5Y/10Y) and wage indicators (such as the Atlanta Wage Tracker) to verify the anchoring of inflation expectations.
• Review the Fed’s minutes for hints on SRF usage, QT pace, and the tone regarding appropriate reserve levels.
• Compare the hyperscalers’ CapEx guidance with data center power procurement plans.
• Check the AI revenue contributions in the M7 earnings announcements and evidence of unit cost declines.

Key Points Often Overlooked by Other Media (4 Core Aspects)

• Although refunds add to ‘private cash flow,’ the duration mix of the government bond issuance needed to facilitate them can, in turn, tighten short-term financial conditions.
• The end of QT does not automatically trigger an ‘automatic QE.’
• A liquidity bridge can be established using SRF or short-term operations, enabling a recovery in risk appetite without resorting to large-scale QE.
• The Fed Chairmanship transition issue until May 2026 may affect the ‘policy tone,’ but if inflation reaches 5–6%, the liquidity drive will halt regardless of the personnel changes.
• The true essence of the AI cycle is the downward trend in cost curves.
• For enterprises to see explosive usage, the cost per token must drop.
• Without this indicator, debates around a ‘bubble/bright future’ remain at a standstill.

Positioning Ideas (Scenario-Based)

• Base: Hold core large tech and AI performance visibility stocks, while keeping 10–20% in cash and maintaining funds in short-term bonds/BIL for volatility periods.
• Risk Hedge: Underweight oil-sensitive sectors, incorporate inflation swaps/curve flatteners, and embrace defensive quality factors.
• Upside Call: If the combined refund and issuance mix is favorable, beta expands; leverage on AI infrastructure chains like semiconductor power management, thermal management, and photomodule industries.
• Downside Call: In the event of inflation re-acceleration, pivot towards essential consumer goods, healthcare, and high cash flow quality stocks with lower earnings sensitivity.

< Summary >

• The essence of the short-term correction is not the collapse of the AI bubble but political uncertainty.
• The impact of the post-Supreme Court refund and government bond issuance pathway on liquidity is determined by the ‘issuance duration mix.’
• The data blackout from the shutdown has widened the Fed’s rate path uncertainty and intensified market volatility.
• If inflation re-accelerates into the 5–6% range, the liquidity-driven market may end prematurely.
• Although the motive behind the liquidity drive holds strong until the 2026 midterms, the anchoring of inflation expectations is the key determinant.
• The AI trend is essentially driven by downward cost curves and profit visibility, while M7 exhibits the duality of being both a beneficiary and the first target for profit-taking.

[Related Articles…]

• The Real Path of Liquidity until 2026 through Government Bond Issuance and RRP
• The Tug-of-War between the AI Cycle and Inflation: How to Read the Volatility of M7

*Source: [ 경제 읽어주는 남자(김광석TV) ]

– 셧다운 종료 이후, 대법원 판결과 26년 중간선거까지 이어질 트럼프의 ‘유동성 드라이브’ 시나리오 | 클로즈업 – 칠판강의 3편

---

● Pension-Strangled KOSPI, TAA Cap Sparks Dollar Surge

The Real Reasons the National Pension Fund Can No Longer Buy More KOSPI and the 2026 Outlook: TAA Limit, Exchange Rate, and AI Cycle Transition

This article covers the direct impact of the National Pension Fund’s TAA limit and changes in strategic allocation on KOSPI supply and demand, the chain reaction of exchange rates and policy cards, global economic scenarios for 2025–2026, and even the “next players” in AI investment trends.It explains both the 5,000 KOSPI scenario and the 3,000 risk while organizing a practical data checklist and positioning guidance for current evaluation.Notably, it separately unravels the hidden connection between “pension funds – exchange rates – foreign investor flows” that most news outlets overlook.

Today’s News Summary: Supply and Demand Terrain and the Reason for KOSPI’s Hesitation

• After a sharp rise in September–October, the KOSPI lost momentum in November and fluctuated up and down.
• Foreign investors maintained a net selling trend on a monthly basis, while individual investors bought heavily during steep declines before switching to profit-taking and wait-and-see stances recently.
• Among institutions, securities (investment firms) defended the index, but pension funds appeared to engage in net selling and rotation trades.
• The interpretation that “the market outlook isn’t bad, but the domestic stock limit is essentially reached, making further net buying difficult” for the National Pension Fund is gaining momentum.
• From a policy perspective, discussions include dividend tax reform with separation taxation, additional measures to boost the KOSPI, and the potential use of the TAA limit.

The Reason the National Pension Fund Can No Longer Buy More KOSPI: Strategic Allocation and the TAA Limit

• Strategic allocation adjustment. The 2025 strategy set a direction to slightly lower the domestic equity proportion and increase the overseas equity proportion.
• Tactical Asset Allocation (TAA) limit. While the basic limit (e.g., 14.9%) allows a ±3% range (up to approximately 17.9%), it is widely speculated that the surge in September–October had pushed it close to the upper limit.
• As a result, instead of additional net buying, the focus shifted to rotation trades, and even if there is available capital, regulations render it difficult to acquire more.
• However, there is an observation that if policies expand TAA flexibility or readjust the strategic ratio, there could be an additional buying capacity of around 30 trillion won.
• Gains and losses from an investor standpoint. Although there is a short-term upswing supply benefit, if prices reverse from their highs, the pension fund’s risk budget and volatility limits could trigger a ‘pro-cyclical’ sell-off, posing a medium-term risk.

Policy Card and Market Impact: Which Will Be Used in Reality First?

• Priority card: Implementing TAA and improving dividend separation taxation. This combination simultaneously enhances the expected returns and after-tax returns on domestic stocks.
• Secondary card: Micro-level institutional enhancements such as inducing corporate share buybacks/dividends, improving listing regulations, and liquidity support.
• Controversial card: Strengthening taxation on overseas stocks may only gain political traction during stressful periods such as an exchange rate around 1,500 won, but its side effects are significant, making its realism low.
• Exchange rate trajectory. The pension fund’s increase in overseas allocation structurally raises dollar demand, acting as a factor for exchange rate appreciation. As the exchange rate rises, foreigners may face greater currency hedging costs on won-denominated assets, leading to a dilemma where stock market buying momentum is weakened again.

Macro Scenarios for 2025–2026: Risks of Interest Rates, Exchange Rates, Inflation, and Recession

• Interest rates. If policy rate cuts are delayed, valuation re-rating in the stock market will be limited. Persistently high long-term rates increase volatility in sectors sensitive to valuation.
• Inflation. Unless entrenched inflation in service prices and housing costs eases, central banks will slow down easing, which will favor value stocks relatively more than growth stocks suffering from discounting.
• Exchange rate. A complex interplay is at work between the pension fund’s increased overseas allocation, current account and export cycles, and the direction of U.S. interest rates. In a phase of won depreciation, export stocks tend to be stronger while domestic and financial stocks lag.
• Recession. If signals of a slowdown in U.S. consumer spending (declining card spending, rising delinquency rates) accumulate, EPS upward revisions could falter, and volatility in risk assets may increase. Preparation is needed for all three tracks: soft landing, shallow recession, or hard landing.

AI Trend 2025–2026: From Infrastructure ‘Expansion’ to ‘Efficiency’ and ‘Edge’

• Changes in the CAPEX mix. In 2023–2024, GPU and data center expansions dominated, and in 2025–2026, investment focus will shift towards improvements in power, cooling, networking, and storage efficiency.
• Act II of the memory supercycle. The transition to HBM3E and HBM4, expansion of TSV/COWOS supply chains, and the shift to DDR5/LPDDR5X will lead to structural improvement in ASP and mix. In the Korean stock market, memory, advanced packaging, substrates, and materials form the core.
• Networking and interconnect. Technologies such as 800G/1.6T Ethernet, optical modules, and switch ASICs will be key to alleviating bottlenecks. Power semiconductors, VRM, and immersion/direct liquid cooling will also be beneficiary sectors as part of the ‘power-cooling’ axis.
• Edge AI and on-device applications. The demand for inference expands with AI PCs, smartphones with NPUs, and AI for vehicles. Analog/mixed-signal, PMIC, sensors, and high-performance PCBs will see demand.
• Software monetization. Instead of a race for models, subscription-based revenue models for “agents + business automation” are becoming tangible. There is also potential for revaluation of software/platforms relative to infrastructure due to lower valuation burdens.

Korean Investment Strategy: Base, Risk, and Hedge Playing

• Base case (soft landing). Increase the proportion of memory, advanced packaging, server components, power/cooling, and networking equipment. Concurrently invest in value stocks with strong dividends and cash flows. Assume the exchange rate to remain in the high 1,300s to 1,400 won range.
• Risk case (sharp decline in consumer spending/recession). If high-valuation growth stocks lose momentum, increase the share of defensive stocks, dividend stocks, and cash. In the event of a rising exchange rate, favor export stocks, while decreasing domestic and financial stock proportions.
• Hedge. Manage the ratio of dollar/won hedges, keep gold and cash as standby assets, and consider defensive instruments such as inverse funds or puts during periods of sharply rising volatility. Avoid excessive leverage.
• Checklist.1) Post-Black Friday consumer data (YoY card spending).2) U.S. EPS revision trends and semiconductor industry indicators (inventory/ASP/operating rates).3) Mortgage and credit card delinquency rates and unemployment figures.4) Whether the won/dollar exchange rate approaches 1,500 won and signals from authorities.5) Official updates on the National Pension Fund’s TAA/strategic allocation.

The Most Important Aspect That Other YouTube/News Outlets Overlook

• FX – Pension Fund – Foreign Investor Flow Asymmetry. The pension fund’s increase in overseas allocation triggers won depreciation (exchange rate rise), which in turn increases foreign investors’ currency hedging costs on won assets, reducing the momentum of net stock buying. Supporting domestic stocks and stabilizing the exchange rate may not move in the same direction.
• Risk budget and pro-cyclicality. Although raising the stock allocation in a pension fund helps supply in a rising market, in times of surging volatility, mechanical reduction due to risk limits can amplify a market downturn. “Increases do not always mean benefits.”
• The critical point of liquidity elasticity. Once critical thresholds such as an exchange rate level around 1,500 won, deepening inversion of short- and long-term interest rates, and increased credit card/mortgage delinquencies are crossed, policy and supply reactions can be non-linear. A strategy that tracks “critical point triggers” rather than index logic is necessary.
• Inducing dividends/share buybacks does not control the exchange rate. Tax and governance reforms are effective for medium- to long-term valuation improvement, but they are insufficient to create short-term reversals in exchange rates and capital flows. The FX cycle must be managed separately.

Data Check Points: The Next Month Is Crucial

• Post-Black Friday/Year-end consumption statistics. Underperformance relative to expectations could pressure growth and high-valuation stocks, while overperformance might extend a rally.
• Semiconductor leading indicators. HBM orders, packaging capacity, memory price indicators, and guidance from suppliers.
• Inflation and interest rates. Directions for service prices, wages, and long-term rates. Reconstructing expectations of rate cuts is key.
• Exchange rates. Execution/mentions of the pension fund’s overseas allocation and won levels. A surge in the exchange rate could trigger a change in the relative strength between export and domestic stocks.
• Domestic policy schedule. Announcements on dividend separation taxation rates, updates on TAA/strategic allocation, and short-term liquidity support measures.

KOSPI 5,000 vs. 3,000 Scenario: A Simple Roadmap

• 5,000 Track. Expansion of TAA flexibility + improvement of dividend taxation + upward revision of semiconductor performance + exchange rate stabilization (won appreciation) + re-emergence of expectations for rate cuts.
• 3,000 Track. Sharp decline in U.S. consumer spending + downward EPS revisions + re-crossing the 1,500 won exchange rate + fixed high interest rates + reduction in pension fund risk budgets leading to adverse supply and demand.
• Conclusion. Rather than “predicting” the direction, tracking triggers and adjusting position sizes will increase the chances of success.

Practical Action Checklist

• Technical. Whether the index reclaims its high (e.g., around 4,200), expansion in trading volume, and confirmation of simultaneous net buying by foreign investors in both futures and cash markets.
• Fundamental. Visibility in semiconductor ASPs/inventory, data center CAPEX, AI PC shipments, and power/cooling CAPEX.
• Supply/Policy. Latest news on TAA/strategic allocation, dividend/share buyback policies, and fund inflows at year-end and beginning.
• Risk Management. After accumulating two to three failures to break the high, consider profit-taking or position reduction, monitor shifts in export/domestic ratios if the exchange rate surges, and activate hedges when volatility expands.

< Summary >

• The National Pension Fund finds it difficult to make additional net buys not due to a pessimistic market outlook, but because of the TAA/strategic allocation limits.
• If policies widen TAA flexibility, there will be short-term supply benefits, but volatility surges could increase ‘pro-cyclical’ risks.
• The pension fund’s increased overseas allocation triggers exchange rate rises, which can weaken the momentum of net buying by foreign investors.
• AI investments for 2025–2026 will shift focus to “power, cooling, networking, and memory,” with edge AI and software monetization coming to the forefront.
• The KOSPI 5,000 vs. 3,000 scenario depends on triggers such as interest rates, exchange rates, consumer behavior, and semiconductor performance. Adjust positions based on data.

[Related Articles…]

• The Decisive Factor of Exchange Rates on the Stock Market
• The Next Player in the AI Infrastructure Investment Cycle

*Source: [ Jun’s economy lab ]

– 국민연금이 코스피를 더 사지 못하는 이유

---