The software industry is experiencing a technological change (AI) that will permanently alter pricing, similar to fracking's impact on oil & gas in 2014. This is happening alongside excessive leverage in private credit.

Software represents 23% of direct lending but only 1% of US companies and 7% of public companies - a massive concentration risk from the 'gold rush' to finance software buyouts.

Public software companies (NASDAQ, S&P, Russell 2000) have reasonable debt levels (0.5-1x leverage), but private direct lending has 20x leverage - these companies cannot generate cash flow to reposition for AI.

Capital availability will dry up for loan extensions in 2027-28, leading to 'extend and pretend' scenarios where loans won't get paid interest.

Software lending needs conservative multiples (4x EBITDA, not 10x) with higher spreads. The broader economy will be fine - this is a sector-specific issue, not systemic.

Current 23% software exposure in direct lending is too high and will create good opportunities for disciplined lenders.

Public and private equity will buy distressed software assets at pennies on the dollar. Private equity has the upside; private credit doesn't.

Not buying software now. Focused on hard-asset lending: concrete, commercial real estate, aircraft, maritime assets, turbines - businesses with low obsolescence and good recovery value.

Bullish on the economy, lending against hard assets at 66% LTV with full recovery in defaults, unlike software with near-zero recovery.

This explains why industrials/materials are up 25% while software is down 25-40%.