2026 03 03T08 18 53 357Z Gmi March 2026 Monthly Red

Author: Raoul Pal (Global Macro Investor) Date: 2026-03-03 Type: r2 Evidence: 39 Themes: 17

copper-specialty-commodities-bottleneck

🟢 [E4016] Copper spec positioning is telling you ISM is heading to 60 this year. South Korean exports are absolutely booming. Japan Machine Tool Orders continue to firm. These are leading indicators of the second-order CapEx and infrastructure buildout phase.
supporting · 2026-03-03

us-dollar-fx-structural-bear

🟢 [E4008] Pal expects another leg lower in dollar with DXY targeting low 90s based on past Trump administration impact. Trump and Bessent want both lower rates and weaker dollar — expecting re-run of Greenspan years. Truflation shows CPI heading below 2%. Combination of lower rates and lower dollar will push up financial conditions.
supporting · 2026-03-03

regional-opportunistic-trades

🟢 [E4017] South Korea exports are 'absolutely booming' and Japan Machine Tool Orders continue to firm — both confirming global manufacturing cycle acceleration. Global Semiconductor Sales have exploded to fastest YoY growth rate since 2009. Empire Manufacturing ripped to highest since February 2023, with CapEx intentions central to 2026 thesis.
supporting · 2026-03-03

inflationary-bust-commodity-barbell

🔴 [E4052] Pal frames current regime as deflationary for services via AI. Real-time inflation measures collapsing with Truflation understating CPI but showing clear path below 2%. The Everything Code predicts debasement (~8% annually) but productivity gains from AI could eventually match this, transforming the economic model entirely.
challenging · 2026-03-03

equity-market-correction-positioning

🟡 [E4012] NDX still looks constructive — above 200-day MA with bullish ascending triangle consolidating since October. Decisive break below 200-day MA not base case but would trigger deeper pullback like 2024. The 1999-2000 analogue suggests 3 months of pullback but policy backdrop is opposite (easing vs tightening). Better analogues are 2021 or early 1990s.
contested · 2026-03-03

crypto-ai-onchain-economy

🟢 [E3997] AI agents are arriving as economic actors — Gartner projects non-human identities will outnumber human employees 80:1 in enterprise. McKinsey estimates 60% of crypto trades will be agent-managed by end of 2026. MCP (Model Context Protocol) is the open standard for AI agents, adopted by OpenAI, Google, Microsoft. Google announced WebMCP browser technology February 10, 2026.
supporting · 2026-03-03
🟢 [E4033] Tokenization is the operating system of the machine economy — dollars become stablecoins, bonds become tokenized RWAs, identity becomes credential tokens. Every major crypto narrative (stablecoins, DeFi, RWA, gaming, storage, identity) resolves into a single story: coordination infrastructure for an economy where majority of participants are not human.
supporting · 2026-03-03
🟢 [E3998] Stablecoins settled $35T in 2025; current run rate for 2026 is $110T. Stripe's annual letter declared agents will handle most internet transactions. Blockchains will need 1M+ TPS (current ~1,500). Stripe built Tempo payments blockchain for agent settlement with Visa, Nubank, Shopify, Klarna testing. Every TAM calculation in crypto is wrong by orders of magnitude because it assumes human end-users.
supporting · 2026-03-03

ai-pricing-sovereignty-local-models

🟢 [E4015] Anthropic is the fastest-growing revenue business of all time, demonstrating Reed's Law in action. Even on a log scale, the revenue growth is exponential. This translates into accelerating AI capability — the AI revolution is proceeding faster than anyone anticipated. Patrick Collison (Stripe CEO) suggests Q1 2026 will be looked back on as 'the first quarter of the singularity.'
supporting · 2026-03-03

ai-disruption-knowledge-economy

🟢 [E4043] Citrini Research speculative scenario about 10.2% unemployment and 38% S&P crash by June 2028 caused DoorDash and Uber to plunge. White House commented calling it science fiction. Amazon leaked documents show company believes it can avoid ~600,000 hires by 2030 thanks to robots.
supporting · 2026-03-03
🟢 [E4006] AI capability exploding: METR Time Horizons chart shows frontier models can complete 15+ hour knowledge work tasks (up from 10 minutes in 2023, 1 hour in early 2025). Latest models diverging from logarithmic trend — going vertical. SWE-bench shows Claude Opus 4.5 solving 72% of real-world software engineering tasks (up from 45% in late 2024).
supporting · 2026-03-03
💬 [E4044] Mattin argues markets cannot process the Economic Singularity — they lack tools, metrics, or frameworks to price civilizational transformation. Two stories colliding: Story One ('CapEx in search of return') is legible to markets; Story Two (civilizational transformation) is not. Market incoherence stems from these stories colliding.
commentary · 2026-03-03
🟢 [E4005] $830B software stocks selloff driven by Anthropic's Claude integrations targeting legal, sales, marketing, and data analysis. IBM had worst day in 25 years after Anthropic revealed Claude Code can modernize COBOL. Block cutting 40% of staff (4,000+ people) with Dorsey citing 'intelligence tools have changed what it means to build and run a company.' Stock soared 24%.
supporting · 2026-03-03

tesla-robotics-autonomy

🟢 [E4009] Trade recommendation: BUY TSLA within Long-Term Portfolio. Tesla is AI and robotics platform wrapped in automotive business. Daily 13 buy signal triggered, weekly 9 buy setup on 200-day MA. As ISM turns up, vehicle demand and sales growth reaccelerates. Base case TSLA trades toward +1 to +2 standard deviation range, suggesting $1,500 to $2,500.
supporting · 2026-03-03

crypto-industry-coinbase-equity

🟢 [E4011] Trade recommendation: BUY HOOD within Core Portfolio. HOOD completed 5-wave lower but no 13-count yet. First weekly 9 buy setup since November 2023. Testing -2 standard deviation band of long-term regression channel. Fifth 40-50% corrective move this cycle — deeply oversold for 70-vol equity equivalent to 10% SPX pullback.
supporting · 2026-03-03
🟢 [E4010] Coinbase completed 5-wave move lower into DeMark wave 5 projection at $146 with daily 13 buy counts cluster. Weekly 9 buy setup in play with price reversing below TDST at $160. Added to long-term portfolio in February Flash Update. Circle (CRCL) up 48% since added last month — dominant stablecoin player.
supporting · 2026-03-03

global-liquidity-cycle-macro-regime

🟢 [E3992] Financial Conditions lead liquidity by 3 months and business cycle by 9 months. Pal expects further dollar weakness (DXY to low 90s) and rates falling to 2.5% in the 5-year sector, which will push up financial conditions and extend the business cycle into H2 2027 for liquidity and H1 2027 for the ISM.
supporting · 2026-03-03
🟢 [E3990] Pal argues liquidity is the single most dominant macro factor, explaining 96% of tech stock prices and 91% of BTC prices. 2026 is the critical year for debt rollover with nearly $10T to refinance, requiring accelerated money printing. The average weighted debt duration is 5.4 years, meaning 2021 debt must be rolled now. GMI Global Total Liquidity Index shows acceleration in both level and rate of change.
supporting · 2026-03-03
🟢 [E3991] US Total Liquidity and Global Liquidity diverged in July 2025 when TGA was rebuilt, causing 'alligator jaws' between BTC/NDX and liquidity. The measures realigned on December 3rd after the government shutdown ended. Pal expects the 3-month liquidity lead to reassert by March 3rd, closing the gap between crypto prices and liquidity.
supporting · 2026-03-03

crypto-regulatory-stablecoin-catalyst

🟢 [E3999] The Clarity Act is expected to be agreed in the next 2-3 weeks as the Administration pushes hard. It opens the door for everyone to build on blockchains and invest in them. Stablecoin issuance growing at 60% per annum and expected to accelerate sharply. Coinbase, Bitwise report enormous interest from Wall St, pension funds, investment advisors, and sovereign wealth.
supporting · 2026-03-03

solana-sui-layer1-ecosystem

🟢 [E4002] Sui's Mysticeti v2 consensus hits 67ms median latency with sub-second full finality (vs Solana's 12.8s actual finality). Transaction costs stable at $0.01-0.02 despite 12x throughput increase. Single PTB can execute up to 1,024 operations across multiple protocols in one atomic unit. DeepBook, Walrus, Seal, Nautilus provide pre-built coordination stack.
supporting · 2026-03-03
🟢 [E4040] Sui today occupies similar position to Solana in April 2021: comparable FDV (~$9B vs $8-12B), comparable intelligence density, comparable early-stage infrastructure. Solana subsequently went $30 to $260. Three triggers to watch: stablecoin depth crossing 2.0x TVL, protocol count exceeding 300, broader wallet adoption.
supporting · 2026-03-03
🟢 [E4038] Only metric that predicts forward L1 returns is capital stickiness (velocity). High velocity = low coherence. Chains where capital churned fastest (Solana 208x, SUI 91x) delivered worst returns from Jan 2025 peak. Fee yield and TVL/FDV ratios had zero predictive power. Sui at 82x velocity sits between ETH (11x) and Solana (162x).
supporting · 2026-03-03
🟢 [E4000] Sui scores highest on Universal Code framework across all dimensions: 389 developers per $1B FDV (vs Solana 322), $0.0003 transaction costs, 77% staked with 24-hour exit (vs Solana 67%), 14.3 protocols per 100 developers (vs Solana 4.6). TVL per active address $2,120 vs Solana $1,648. Sui dominance is about to rise.
supporting · 2026-03-03
🟢 [E4001] Through an 83% token price drawdown (Dec 2024 to Feb 2026), Sui's per-user economic density held constant at ~$2,120 while stablecoins grew 50%, protocols grew 344%, developers grew 219%, categories grew 29%. This demonstrates 'agency' — the system continues complexifying when external energy is withdrawn. Sui is an 'organism' not a 'project'.
supporting · 2026-03-03
🟢 [E4039] Sui is the natural settlement layer for AI agent economy due to sub-second finality, stable predictable fees, native composability, and horizontal scalability. Mysten Labs collaborated with Google on Agentic Payments Standard. Beep launched as first agentic finance protocol with a402 protocol for zero-fee USDC micropayments between agents.
supporting · 2026-03-03

bitcoin-cycle-bear-phase

🔴 [E3994] Weekly DeMark counts on BTC, ETH, and SOL are on 6 of 9 lower, with SUI on week 12 of 13 lower. Daily 13 buy counts are present. Pal expects potential final flush before V-shaped recovery. Setting stink bids: ETH at $1,650, SOL at $65, SUI at $0.75. The $1,600 level is THE key support for ETH.
challenging · 2026-03-03
🔴 [E3993] Pal explicitly rejects the bear market narrative for crypto. Bitcoin's 50% drawdown is normal volatility for a 70% vol asset class — equivalent to Amazon's 25% correction. When vol-adjusted, BTC charts look identical to Amazon. The RSI sentiment is as depressed as the 2018 cycle low, approaching -2 standard deviations on the log regression channel comparable to 2015 lows. Fair value by end-2026 sits near $300,000.
challenging · 2026-03-03

portfolio-construction-income-allocation

🟢 [E4013] Core Portfolio positions: BTC (entry Jul 2022), ETH, SOL, Coinbase, Tesla, Google, Microsoft, Apple, NDX, TSMC, SUI, RKLB (+270.3%), TAN, DEEP, SCF, HOOD. Long-term portfolio adds: SUI at $0.97, Coinbase at $160.10, Circle at $56.33 (+48.1%), TSLA. Tactical stink bids placed: ETH $1,650, SOL $65, SUI $0.75.
supporting · 2026-03-03

macro-cycle-frameworks

🟢 [E3996] ISM Production subcomponent moved to 55.9, confirming inventory normalization. Average Weekly Hours broke above 0.3% ceiling that capped the past four years — hours lead jobs in the Employment Dominoes sequence. Global Semiconductor Sales exploded YoY to fastest growth rate since 2009.
supporting · 2026-03-03
🟢 [E4025] The traditional manufacturing cycle has fundamentally changed — AI has flipped causality. Compute is now the binding constraint; growth starts with chips and propagates outward. 2025 was about semiconductors, 2026 becomes second-order CapEx and real-world infrastructure buildout, which DM PMIs will respond to.
supporting · 2026-03-03
🟢 [E3995] ISM Manufacturing jumped to 52.6 in January 2026, with seven industries expanding MoM — the biggest breadth thrust since COVID 2020 reopening. Empire Manufacturing hit highest since February 2023. Chicago PMI surged to 57.7. Pal expects ISM heading to 60 in 2026. Regional Fed surveys aggregate suggest modest rise in February.
supporting · 2026-03-03
🔴 [E4054] Pal explicitly responds to 'rage quitters' who claim crypto is in a bear market. He argues critics are creating 'memetic narratives' explaining why things have happened rather than why things are happening. The supposed bear market is just normal volatility for a 70% vol asset — equivalent to Amazon's 25% correction.
challenging · 2026-03-03
🔴 [E4014] Pal directly addresses criticism that 'Everything Code is wrong' and 'GMI is wrong.' He dismisses critics who rely solely on technical analysis (Ben Cowen, Willy Woo) as using only charts with no macro framework. Claims 21 years of track record with unparalleled performance including fuck-ups. Framework has never failed and won't now.
challenging · 2026-03-03

ai-capex-infrastructure-bottleneck

🟢 [E4004] CapEx is backed by orders — hyperscalers surpassed $1 trillion in backlog orders, growing faster than CapEx in nominal terms. This is demand-led investment, not spending into a void. Energy is the binding constraint, not over-investment. IEA projects global data center electricity consumption to double by 2030.
supporting · 2026-03-03
🟢 [E4030] Energy is to the AI age what land was to agricultural age or capital to industrial age — the scarce resource around which everything orbits. The core metric is intelligence output per unit energy. Energy infrastructure is the foundational play: solar, fuel cells, cooling systems, grid build-out, and reshoring of critical supply chains.
supporting · 2026-03-03
🔴 [E4007] Pal/Mattin explicitly reject the bubble narrative. CapEx is backed by $1T+ backlog orders growing faster than spending. This is demand-led investment. The comparison to dotcom is flawed because in 1999-2000 the Fed hiked 175bps; today policy is easing not tightening. Better analogues are 2021 or early 1990s — consolidation within broader uptrend.
challenging · 2026-03-03
🟢 [E4029] Energy is the binding constraint for AI, not over-investment. Elon Musk has warned AI buildout will run out of electricity. Microsoft has $80B in Azure orders unfulfillable due to power constraints. Meta building Prometheus, first gigawatt-scale data center with Manhattan-sized land footprint. US/Europe grid interconnection queues are years long.
supporting · 2026-03-03
🟢 [E4003] Hyperscalers deploying ~$600B in planned CapEx in 2026, up dramatically from 2025. Amazon guiding $200B likely posting negative FCF. Hyperscaler spending exceeds 1.3% of US GDP — higher than Manhattan Project (0.4%) or Apollo (0.7%). Microsoft has $80B in Azure orders it cannot fulfill due to power constraints.
supporting · 2026-03-03