Markets
The feedback loop
The tower is not abstract: a margin call decides which mine opens, a strait closing sets a derivative price. Sourced evidence of the coupling both ways.
Layer 3 is the fastest, most abstract layer, but it is not detached from the physical base. The coupling runs both ways and both are measurable: decisions taken in the markets shape the real world, and shocks in the real world shake the markets. For some channels the direction or size of the effect is contested in the peer-reviewed literature; those rows are given with both views.
Top-down, L3 decisions shape the real world
Commodity index investment grew from $15 billion in 2003 to over $200 billion in 2008. Tang & Xiong measure the correlation between the returns of indexed commodities and oil rising from near zero before 2004 to about 0.5, with no such rise for off-index commodities, consistent with financial flows moving spot prices.
Irwin & Sanders find no causal effect of index investment on price levels or volatility: in Granger-causality tests index positions do not lead price changes, and the 2008 spike coincided with tight fundamentals (low stocks, biofuel demand, a weak dollar).
Across 151 systemic banking crises since 1970 the median output loss is 30% of trend (pre-crisis) GDP over the crisis window; the median rise in public debt is 12% of GDP and the median direct fiscal (recapitalisation) cost 6.7% of GDP. Reinhart & Rogoff find real government debt rises 86% on average in the three years after a systemic crisis, with unemployment up 7 percentage points.
The market-set cost of capital governs which projects and mines get built. The user-cost elasticity of business investment is estimated at -0.5 to -1: a 10% rise in the cost of capital lowers the desired capital stock by 5-10%. In monetary-VAR estimates a 1 percentage-point rise in the policy rate, transmitted through bond yields, cuts business fixed investment by 2-5% within two years.
In Sharpe & Suarez's survey of CFOs, most firms report that a 1 percentage-point change in interest rates would not change their investment plans; investment appears more sensitive to demand and cash flow than to the cost of capital.
In 2010 Spread Networks spent $300 million to lay 1,331 km of fibre on a straighter New York-Chicago path, cutting the round trip from 16 ms to 13 ms. Within three years microwave-tower networks along the same corridor beat that fibre (one way 4.1-4.7 ms). Millions in capital and physical plant are spent for fractions of a millisecond.
Bottom-up, physical shocks shake the markets
Hamilton shows 9 of the 10 US recessions since World War II were preceded by a sharp rise in the crude oil price, with a large oil contribution to the 1973-75, 1980, 1990-91 and 2007-09 downturns. In his non-linear specification a doubling of the oil price is associated with a 2-3% fall in US GDP after 4-6 quarters. The 2022 energy shock repeated the pattern.
Kilian and Barsky & Kilian argue the oil price is largely endogenous: most 1970s increases were driven by global demand, not supply, and Bernanke, Gertler & Watson find much of the recession comes from the monetary-policy response rather than oil itself.
The 2021 Ever Given grounding closed the Suez Canal for six days; $9.6 billion of trade per day was held up and close to 400 ships queued. In the Red Sea crisis from late 2023, Suez transits fell 50-70%; the Shanghai-Europe container spot rate tripled to quadrupled in early 2024 and war-risk insurance for the Red Sea rose from 0.05-0.1% to 0.5-1% of hull value.
After Russia announced a grain-export ban on 5 August 2010 following drought and fires, CBOT wheat futures rose 70% between June and August 2010 (from $4.25 to $8.00 per bushel). The 2012 US Midwest drought pushed CBOT corn futures to a record $8.30 per bushel in August 2012 (up 60% from June) and soybeans to $17.70.
Biophysical economics argues that declining energy return on investment (EROI) and resource depletion put a rising drag on growth and so on long-run real asset returns: global oil-and-gas EROI fell from 30:1 in the 1990s to 15-18:1 in the 2010s, and a societal EROI floor of 5:1-11:1 is proposed below which discretionary economic activity contracts.
Mainstream growth economics finds no robust historical link from resource scarcity to returns: real commodity prices show no long-run upward trend, substitution and technical change have repeatedly offset depletion, and the resource share of GDP is small enough to limit aggregate effects even of large price moves. Nordhaus's work and the long record of failed scarcity forecasts (the 1980 Simon-Ehrlich wager) are cited against a binding physical ceiling.
Abstraction creates no matter; but a weightless number can decide which mine opens, which refinery is built and what millions pay for bread. The loop is closed and runs both ways, which is why there is no such thing as “just finance”.
Layer 3 methodology · §10 →