Tech Sector Collapse? Why AI Is Crashing Stocks And What's Next
Source: The David Lin Report | Date: February 26, 2026
Investment Research Summary: Tech Sector Collapse & AI Impact
Investment Thesis
AI disruption is creating a fundamental profit rotation from legacy software companies to real economy businesses, while macro positioning remains heavily overweight US assets despite better opportunities in global markets, gold, and commodities.
Sentiment
NEUTRAL (on US tech/stocks) | BULLISH (on gold, commodities, non-US assets)
Time Horizon
MEDIUM-TERM (3-12 months, with secular trends extending longer)
Key Takeaways
- Software sector cannibalization is a profit rotation, not economic destruction - AI agents replacing legacy software (e.g., IBM's 30% crash) transfers profits to real economy companies rather than destroying aggregate value
- Portfolio positioning is dangerously concentrated - Most investors are "all in" on US bonds/stocks while being radically underweight war-resilient assets (gold, commodities) and global diversification
- 6040 portfolio is flat YTD while alternatives crush - Emerging markets, foreign stocks, and commodities significantly outperforming S&P 500 in 2025
- Dollar remains structurally overvalued - Near historic highs with typical correction being 25-30% decline; ongoing dedollarization by institutional investors favors non-US assets
- Systematic approach and humility beat conviction - Market humility and rule-based strategies outperform overconfident concentrated bets in current volatile environment
Market Views
- Gold: Currently $5,200 with "all clear" for continued upward trend after flow normalization; Western investor squeeze on supply/demand continues
- US Dollar: Elevated near historic highs, expect 25-30% structural decline over 10-15 year dedollarization cycle
- Bonds: Typical underperformer in war/inflation environments; most portfolios dangerously overweight
- Emerging Markets: Korean equities showing strong earnings growth with low P/E ratios compared to developed markets
- Geopolitical risk premium: Largest US military buildup since 2003 Iraq invasion around Iran underpriced in markets
Assets Discussed
- IBM - BEARISH (down 30% in 2 weeks, 13% in one day on Anthropic COBOL AI news; emblematic of software disruption)
- Gold - BULLISH (at $5,200, flows normalized, central bank buying continues, structural uptrend intact)
- Commodities - BULLISH (war trade, inflation hedge, outperforming S&P 500 YTD)
- US Stocks (S&P 500) - NEUTRAL/UNDERPERFORM (flat YTD, overvalued relative to global alternatives)
- US Bonds - BEARISH (overweight in portfolios, underperform in war/commodity bull markets)
- Emerging Market Equities - BULLISH (crushing S&P 500 YTD, cheaper valuations, AI infrastructure beneficiaries)
- Korean Stocks - BULLISH (strong earnings growth, low P/E ratios, compute infrastructure plays)
- Australian Dollar (vs USD) - BULLISH (strong economy, tightening cycle vs US easing)
- Bitcoin/Crypto - NEUTRAL (down 50%, high beta to risk assets, less diversifying than gold but not zero value)
Risk Factors
- AI productivity paradox: Cannot sustain both widening profit margins AND topline growth long-term (income earners = spenders); unemployment risk if margin expansion dominates
- War trade timing: Geopolitical escalation well-publicized but timing unpredictable; being strategically positioned matters more than tactical timing
- US exceptionalism continuation: 250-year trend of US dominance may persist despite structural concerns; diversification is hedge, not conviction bet
Notable Quotes
"The typical outperformers are gold and commodities and the typical underperformers are bonds in particular. And when you look at how most portfolios are positioned, they're basically all in on bonds and hold no gold and commodities."
"Software companies are leeches to real economy businesses and they're being disrupted and that's a good thing for the real economy... the pie continues to grow just fine. Really all the conversation is about who exactly takes what share of the pie."
Related Charts
Auto-generated summary.
