MEMO FROM THE FUTURE: BERKSHIRE HATHAWAY
CEO Edition
BOARD STRATEGY SESSION June 2030
TO: Berkshire Hathaway Board of Directors
FROM: Greg Abel, CEO
DATE: June 2030
SUBJECT: AI Across the Berkshire Portfolio and the GEICO Inflection
OPENING
Berkshire Hathaway is a portfolio company: insurance, railroad, utilities, manufacturing, retail, finance. We've succeeded by finding undervalued, durable businesses and improving them slowly and methodically.
AI is now forcing us to rethink every business in the portfolio simultaneously. This memo proposes how to deploy AI across our most important assets—particularly GEICO and our insurance float—while managing the implications for capital allocation.
THE REALITY
GEICO's situation: - Market share: Down from 12% to 8.5% over the last 5 years - Customer acquisition cost: Up 40% in 2 years (digital aggregators winning) - Loss ratio: 96% (essentially breakeven on underwriting) - Float declining as customers leave for cheaper digital competitors
GEICO is in crisis. AI can fix it, but it requires fundamental rethinking.
The opportunity: AI-powered underwriting can reduce loss ratios 5-10 percentage points. AI-powered customer acquisition can cut CAC by 30%. Combined, GEICO becomes profitable again and competitive.
The challenge: This requires $5-10 billion in technology investment and a complete cultural reset. Warren never did this. Can we?
WHERE WE ARE
Current state: - Insurance float: $147 billion (earning 4-5% returns) - GEICO annual revenue: $27 billion, but essentially breakeven on underwriting - Insurance combined ratio: 103% (unprofitable) - Investments: $700 billion (mostly stocks and bonds) - Cash: $165 billion
The strategic tension: Berkshire's entire value proposition depends on insurance float being profitable. If GEICO stays unprofitable, we're managing decline, not growing. And insurance stocks are trading down because investors see GEICO as a problem, not a solution.
THE AI OPPORTUNITY
Opportunity 1: GEICO AI Transformation
The play: Modernize GEICO with AI-powered underwriting, pricing, and customer acquisition. Make GEICO profitable and competitive again.
How: - Deploy AI for underwriting: Analyze 1,000+ variables to predict claims risk better than competitors - Dynamic pricing: Real-time pricing based on current risk, not annual static models - Claims prediction: Identify high-risk policies upfront, reduce loss ratios 5-10 points - Customer acquisition: AI chatbots and personalization reduce acquisition costs 30%+ - Retention: AI predicts churn, enables targeted retention offers
Estimated impact: - Loss ratio improvement: 96% → 88-90% (4-6 points of margin improvement) - CAC reduction: 30-40% - Revenue growth: 5-8% annually (regain market share from digital competitors) - GEICO operating income: Currently ~$0 → $2-3 billion annually by 2035
Timeline: 2-3 years to full deployment
Investment: $5-7 billion in technology infrastructure
Profitability: 30%+ ROI
Opportunity 2: Generalize AI Across Insurance Portfolio
The play: Apply GEICO learnings across Berkshire's entire insurance portfolio (reinsurance, specialty, international).
How: - Build proprietary underwriting models that outperform the market - Create data moat: Berkshire's combined underwriting data (trillions in historical claims) is a competitive advantage - Dynamic pricing and claims management across all lines
Estimated impact: - Combined ratio improvement across portfolio: 2-4 points - Increased pricing power in specialty and reinsurance lines - Insurance float productivity increases 25-30%
Timeline: 3-4 years
Investment: $2-3 billion (incremental)
Opportunity 3: AI Across Berkshire Operating Companies
The play: Deploy AI across railroads, utilities, manufacturing, energy—everywhere in the portfolio.
How: - BNSF railroad: AI-powered logistics optimization, predictive maintenance, fuel efficiency - Berkshire Hathaway Energy: AI grid management, renewable optimization - Manufacturing: AI-powered predictive maintenance, supply chain optimization - Clayton Homes: AI-powered underwriting and customer acquisition
Estimated impact: - 5-15% cost reduction across operating companies - Better asset utilization - Improved profitability on $100+ billion in invested capital
Timeline: 2-5 years (depends on business)
Investment: $2-3 billion
MY RECOMMENDATION
Full-portfolio AI transformation. Start with GEICO (most urgent, highest ROI), expand to broader insurance portfolio, then deploy across operating companies.
This requires accepting that Berkshire is becoming a different kind of company. We're not just buying and holding; we're actively modernizing portfolio companies with technology.
EXECUTION PLAN
Phase 1: GEICO Transformation (2030-2033)
- Invest $5-7 billion in AI/tech infrastructure
- Modernize underwriting, pricing, customer acquisition
- Target: Loss ratio 88-90%, revenue growth 5-8%
Phase 2: Insurance Portfolio Expansion (2032-2035)
- Expand AI playbook across reinsurance, specialty, international
- Build proprietary underwriting models
- Target: 2-4 point combined ratio improvement
Phase 3: Operating Company Modernization (2031-2035)
- Deploy AI across railroad, energy, utilities, manufacturing
- Target: 5-15% cost reduction across portfolio
FINANCIAL IMPLICATIONS
By 2035:
- GEICO operating income: $2-3 billion (from ~$0 today)
- Insurance combined ratio: 99-100% (breakeven to slight profit, vs. 103% today)
- Insurance float: $180-200 billion (growing)
- Float productivity: 5-7% returns (vs. 4-5% today)
- Berkshire book value growth: 10-12% annually (vs. 7-8% historically)
Stock price target: $800,000-900,000 by 2035 (from $610,000 today).
Greg
Confidential — Board of Directors Only