The cynical heuristic: each 5–7 year theme is a state program with a stock market attached, and the trade is to “invest where the tax dollars go.” Pawlicker88’s list: dot-com 1996–2001, biotech and defense 2002–2008, oil and gas 2009–2014, green energy 2015–2021, AI 2021–2026. The periodization is sloppy — shale was a private capex boom and the dot-com bubble was retail money, not subsidy — but the recent half holds: ARRA and then the IRA for clean energy, CHIPS Act plus export controls plus sovereign AI budgets for semiconductors, and the war budgets for defense. The state picks the sector, the sector prices the decade of flows, then the flows stop and the multiple goes with them — solar −44% and clean energy −34% over the last 5 years while semis did +324% (Hard assets for the table).
Why it’s worth keeping despite the sloppiness: the fiscal flow is announced, budgeted and slow — the rare theme signal you can read in a law before it’s in the price. The tell for the end is the same: subsidies sunset, tariffs get renegotiated, the “strategic” label moves on. Cousin of Capital rotates, it doesn't leave (where the money goes next) and of AI circular financing (what the current flow looks like from inside).