Global Economic Outlook 2025-2027 — AI Impact + Dollar Hegemony + China + Expert Comparison + Crisis Survival

Global growth slowing 2025-2027. US +2.0%, China +4.5%, Eurozone +1.2%, India +6.3%.

FRED actual: Q1 -0.6% → Q2 +3.8% → Q3 +4.4% → Q4 +0.5% → 2026Q1 +2.1% → Q2 +1.5%. Slowing but NOT recession.
Source: FRED A191RL1Q225SBEA

Key Data: CPI +3.0%, Core +3.2%. Down from 9.1% peak but above 2% target. "Sticky inflation."
Source: FRED CPIAUCSL

Key Data: Fed cut from 4.33% (Aug 2025) to 3.63% (Jul 2026) — 70bp. Paused since May.
Source: FRED FEDFUNDS

FRED actual: Rose to 4.5% (Nov 2025, Sahm trigger), fell to 4.2% (Jun 2026). Sahm Rule (100% accurate 1949-2023) "failed" first time. LFPR fell to 61.5%, true weakness understated.
Source: FRED UNRATE, CIVPART

Key Data: S&P 500 at 7,710. Forward PE ~22x. AI-driven. 15-20% correction risk.
Source: FRED SP500

Key Data: 10Y 4.63%, 2Y 4.20%. NORMALIZED (+44bp). Strongest signal recession risk declining.
Source: FRED DGS10, DGS2

Key Data: HY spread 2.75% (Q2 2026), near historical low. Credit markets completely normal.
Source: FRED BAMLH0A0HYM2

Key Data: Office vacancy 19.3%. CRE default ~$28B H1 2025. #1 structural risk. $1.5T CRE loans mature 2025-2028.
Source: Moody's, Cushman & Wakefield

Global debt/GDP: 321% peak (2023) → 298% (2024) → projected 320-335% (2026). Advanced economies 345-355%.

FRED actual: Michigan Consumer Sentiment crashed to 49.5 (Jun 2026). 1970s-level low. But consumption continues — "confidence-behavior divergence." Savings rate 2.7% historic low.
Source: FRED UMCSENT, PSAVERT

2023 +0.2% → 2024 +2.7% → 2025-2027 projected 1.5-2.2%. Trade fragmentation + geopolitical risks + tariff barriers.

1-10 scale: Consumer Confidence (9/10), CRE (8/10), AI Displacement (7/10), Income Inequality (8/10).
Source: FRED + IMF + BIS

2025Q4 Sahm triggers → 2026Q1-Q2 soft landing → 2027 risks: CRE maturity wall + AI structural unemployment. Soft landing likely but slower growth.

Baseline (65%): Soft landing. GDP 1.5-2.0%.
Pessimistic (25%): Mild recession. GDP -0.5 to -1.0%.
Severe (10%): Systemic shock.

FRED actual: R&D $916.3B/quarter. Computer new orders +37% YoY. Unprecedented. Engine AND biggest risk.
Source: FRED Y694RX1Q020SBEA, A31SNO

FRED CEU3200000001: Info sector: 4,792K → 4,740K (-52K) → 4,780K. Weekly pay: $37.02 → $37.64 (+1.7%). "Jobs down + wages up" = AI displacement signature.
Source: FRED CEU3200000001

FRED PRS85006173: 97.9% (Q1 2020) → 93.5% (Q1 2026) — 4.4pp collapse. Most underappreciated structural problem. → Savings rate 2.7% historic low.
Source: FRED PRS85006173, PSAVERT

2023 ChatGPT → 2024 NVIDIA → 2025 first layoffs (-12K) → 2026 labor share 93.5%, savings 2.7% → 2027 AI ROI questioning → 2028 unemployment could break 5%.

Mag7 = 32% of S&P 500 (Nasdaq 2000 peak: 33%). Nasdaq PE ~28x (2000 peak: 60x). R&D at record high. Three differences from 2000: ①AI real revenue ②PE 28x vs 60x ③But concentration similar.

AI Job Displacement (8/10): -12K. Market Concentration (7/10): Mag7=32%. AI ROI (6/10): $916B/qtr. Income Inequality (8/10): labor share 93.5%. Consumption (5/10): savings 2.7%.

FRED real-time: Initial claims 199K (historic low) / Avg hours 33.7h (stable) / Recession prob 0.60% (trending up) / Real income $18,056B (stagnant) / Financial stress -0.53 (very loose) / Building permits 1,374K (declining). "Split" indicators — layoff signals good, income signals bad.

Core contradiction: Real consumption $16,885B (growing) vs Savings rate 2.7% (historic low) vs Real income $18,056B (stagnant). Consumption from "eating savings" — unsustainable. Below 2% → must stall.
Trigger chain: Savings exhaustion → consumption cliff → retail decline → profit fall → layoffs → Sahm re-trigger → recession (3-6 months).

2026Q3: Income declining. Permits falling.
2026Q4: Savings exhausted → consumption pivot.
2027Q1: Sahm Rule re-triggers.
2027Q2-Q3: Full recession. GDP -1.0~-1.5%. CRE defaults. Fed cuts to <2.0%.
Conclusion: Timeline moved to 2027 Q1-Q2.

Initial claims lead unemployment by 3-6 months. Current 199K historic low → mass layoffs haven't started. But historically: once claims rise, 200K→250K+ within 3 months.
"Surface better than reality": unemployment 4.2% but participation 61.5%; GDP+1.5% but income stagnant; consumption up but savings 2.7% unsustainable; S&P 7,710 but Mag7=32%.
Historical parallel: Q3 2007 Fed said "solid" — 6 months later, GFC.

Key fact: Sahm Rule was 100% accurate (1949-2023). Nov 2025 triggered at 4.5% — recession didn't happen. Unemployment fell to 4.3% within 1 month. First time ever.

Reasons: ①Post-election gov spending surge ②AI creating jobs ③LFPR drop masking true unemployment.

What this means: No indicator 100% reliable during structural shifts. AI + gov spending changing traditional cycle patterns. If Sahm triggers again, will it "fail" again? Prediction framework needs "structural change" variable.

Impact on my prediction: 2027 Q1-Q2 crisis needs wider confidence intervals. If AI continues creating jobs, recession could delay to 2028.

Actual data: Q2 2026 GDP +1.5%, within baseline.

① Recession (25%): AI bubble + CRE → GDP -1.0%
② Baseline (50%): Soft landing → GDP 1.5-2.0%
③ Bull (25%): AI productivity boom → GDP 3.0%+

Key variables: AI productivity in GDP? Savings at 2.7%? Fed room (3.63%→2.0%)? CRE $1.5T maturity? AI spread beyond tech?

My view: Baseline 50% most reasonable, but AI creates "fat tails." Tail risks larger than historical.

Falsification conditions:
✅ 3+ conditions met → prediction wrong: ①Claims<180K ②Real income +6mo growth ③Savings>4% ④Labor share>95% ⑤Permits>1,500K ⑥CPI<2.5%, Fed→3.0%
❌ None met → crisis confidence 70%+

Subjective assessment:
2026Q2: 15% (5-30%)
2026Q4: 35% (15-60%)
2027Q1: 50% (25-75%)
2027Q2: 65% (30-85%)
2027Q3: 60% (30-80%)

The only certainty: we are uncertain. In 2020 Q1, models peaked at 30%. Actual within 1 month. Models always underestimate tail risks.

Five Dimensions of Mainstream Forecasting:

📊 Dim 1: LEI (OECD) — 10 indicators. Current below trend → slowdown in 6-9 months.
📊 Dim 2: Yield Curve (Fed/Goldman) — 10Y-2Y inversion 100% predicted recessions. Now normalized → mainstream sees reduced risk.
📊 Dim 3: Credit Spreads — HY 2.75% (near low) → "healthy." But spreads spike at LAST MOMENT before crisis (2008: 5% in July).
📊 Dim 4: Labor Market — Sahm + claims + nonfarm. Unemployment 4.2% decline → "soft landing." But LFPR 61.5% → true weakness hidden.
📊 Dim 5: Consumer — Confidence 49.5 crash but spending up → "savings buffer." Rarely tracks savings rate trend.

Consensus: GDP 2026: Fed 2.0%, Goldman 1.8%, JPM 1.7%. Recession prob 15-20%. Unemployment 4.3-4.6%.

Why mainstream more optimistic?: ①Traditional models "slow" during structural shifts ②AI positive contribution overestimated ③Savings 2.7% treated as "temporary" ④CRE risk "manageable"

Why my analysis more bearish?: ①Focus on savings rate — 2.7% = unsustainable (mainstream rarely emphasizes) ②Labor share — 4.4pp collapse = structural consumption decline (mainstream barely mentions) ③AI duality — creates AND replaces (mainstream emphasizes "creation") ④Consumption cliff model — absent from mainstream framework

Conclusion: Mainstream thinks in "traditional cycle" framework. I added "structural change" dimension. If traditional cycles continue, I'm wrong. If structural change dominates, mainstream is wrong.

I. DO THESE NOW

💰 Financial: Emergency fund 6 months / Optimize debt (credit card first) / Don't bet everything / 10-15% gold/bonds
💼 Career: Learn AI tools (free) / Build "AI + expertise" combo / AI-proof skills (creativity/communication/complex decisions) / Plan B (side income/network)
🧠 Mindset: Don't fall for "windfall" hype (90% is course-selling) / Don't fear "experts" (~50% accuracy) / Watch process data

II. WARNING SIGNALS (2+ simultaneously)
⚠️ Claims>220K / Savings<2.5% / Retail 2+ months decline / Confidence<45 / Spreads>4%
→ Stocks<50%, bonds 25-35%, lock fixed-rate loans

III. RECESSION CONFIRMED
🔴 Keep 30-40% cash / Don't panic sell / DCA when stocks down 20-30% / Use unemployment benefits / Keep skills updated

IV. HISTORICAL DATA
📈 Post-2008 S&P: 300%+ over 10yr / Post-COVID: 150%+ over 5yr
📉 But most people do opposite: buy high, sell low

V. "WINDALL" TRUTH
❌ AI course "experts" / "100K/month with AI" scams / AI gig platforms (below food delivery wages)
✅ Use AI to boost work efficiency / AI + professional skills = irreplaceable / Data-driven thinking

Mainstream predictions: OECD 27% high-risk automation / McKinsey 375M job transitions / Goldman 300M affected / IMF worsens inequality.

5 thinking dimensions: ①Tech substitution rate ②Employment elasticity ③Income distribution ④Productivity effect ⑤Policy response.

Truth about "windfalls": Course-selling "experts" / "100K/month with AI" / AI gig platforms with suppressed wages.

Real opportunities: AI + professional skills = irreplaceable / Learn free AI tools / Build "AI-proof" skills (creativity, communication, complex decisions).

I. Four-wave crisis transmission to ordinary people

💼 Wave 1: Unemployment → companies lay off → job search 2weeks→6months → stress
🏠 Wave 2: Asset decline → house price / stock / 401k drop → stop consuming
💳 Wave 3: Debt crisis → can't pay mortgage → foreclosure → credit ruined
🌍 Wave 4: Global contagion → USD strengthens → imports expensive → inflation / EM debt crisis

Key insight: Crisis is not "one day it happens." It transmits wave by wave over 6-18 months.

II. Will China be affected?

🇨🇳 Impact channels: ①Export decline (US recession→less Chinese goods) ②Capital outflow pressure ③Currency pressure (USD strong→RMB depreciation→imports expensive) ④Confidence transmission

🇨🇳 Policy toolkit (LARGER than US): ①Infrastructure investment ②Rate cuts ③Real estate rescue ④Consumption stimulus ⑤Fiscal deficit expansion ⑥SOE leverage ⑦FX management

🇨🇳 Ordinary people: Possible unemployment/wage cuts, slightly higher prices. Won't starve.

III. What about US?

🇺🇸 USD is world currency → can "print and export inflation" / Treasuries are global pricing anchor / Food stamps + unemployment + Medicaid buffer / Won't starve / But middle-class wealth could drop 30-50%

IV. Emerging Markets: Worst hit

🌍 USD-debt countries (Turkey/Argentina/Sri Lanka): USD appreciates → debt service surges → default → currency collapse → inflation 30-100%+ → REAL starvation risk

V. Core conclusions

💡 China: Impact exists. Government strong enough to rescue. But local gov + real estate debt high → tools have limits. People won't starve, but quality of life drops.
💡 US: Recession but internal absorption. USD privilege lets US export inflation to the world. People won't starve. Middle-class wealth drop 30-50%.
💡 EM: Worst hit. No ammo = can only be harvested. 30-100%+ inflation = REAL starvation risk.
💡 Global South: No FX reserves → no ammo → harvested.

I. Five forms of economic crisis

① Financial crisis (1929/2008/2023): Asset crash + bank failures. People: savings frozen / foreclosures / 401k halved
② Currency crisis (1997 Asia / 2015 CNY): Currency devaluation 30-80%. People: savings devalued / imports expensive / unemployment
③ Debt crisis (2010 Europe / Argentina): Can't repay = default. People: inflation 30-100%+
④ Stagflation (1970 Oil / 2022): Stagnation + high inflation. People: wages flat / prices soaring
⑤ Dollar hegemony crisis (1971 Nixon shock / ongoing): Dollar credit damaged. People: global restructuring

Most likely 2026-2027: ② Currency + ③ Debt + ④ Mild stagflation. Like "boiling frog" — slow deterioration, not sudden collapse.

II. Dollar Hegemony: Real Data (IMF COFER Jan 2026)

📊 USD still absolutely dominant:
• Global 56% FX reserves in USD (IMF COFER Jan 2026)
• 2000 peak 72% → 2025 56% (16-year, 16pp decline)
• Global 87% forex trade in USD (BIS 2013)
• 70%+ global debt denominated in USD
• US = 25% of global GDP, but USD = 56% of reserves

⚠️ But trend is clear: declining:
• 2022: USD reserve decline rate = 10x 20-year average (BNN Bloomberg: "faster than commonly acknowledged")
• 2016-2025: ~11% decline (FX-adjusted)
• 2008-2025: cumulative ~22% decline
• Euro ~20% / Renminbi ~3-4% (3rd reserve currency)

III. Forces accelerating de-dollarization (2024-2026)

① Post-2025 Trump tariffs: financial institutions rethinking USD reserve role
② Central banks: 3 consecutive years buying >1,000 tons gold/year
③ Russia+China Dec 2023: abandon USD bilateral settlement
④ China May 2023: $88B Russian oil in yuan
⑤ Saudi Arabia Jan 2023: 1st time in 48 years open to non-USD oil trade
⑥ BRICS Pay: alternative to SWIFT
⑦ China-Russia 2014: 1,500B yuan liquidity swap
⑧ Argentina 2023: buying Chinese imports in yuan

IV. Impact of dollar hegemony shock on ordinary people

🇺🇸 Short-term: US recession → safe haven → USD actually strengthens (historical pattern)
🇺🇸 Medium-term: Treasuries sold → dollar credit shaken → depreciation → imports expensive → inflation
🌍 Long-term: Multi-polar currency → USD below 50% → US "print to buy world" privilege ends → US middle-class living costs rise
🇨🇳 For China: RMB short-term pressure, but long-term benefits internationalization

V. Core conclusions

💡 USD hegemony won't collapse tomorrow, but trend is irreversible. 72% (2000) → 56% (2025) = 16pp in 22 years. At this rate, below 50% by ~2040.
💡 For US ordinary people: Short-term, strong USD makes overseas travel cheaper. Long-term, damaged hegemony = "print and buy world" privilege ends = living costs rise.
💡 For the world: EM worst hit short-term (USD回流 + debt crisis). Long-term, multi-polar system fairer but transition more turbulent.

Sources: IMF COFER (Wikipedia Jan 2026), BNN Bloomberg, BIS Triennial Survey, Wikipedia De-dollarization