Research Strategy
“Who already holds which ground in complexity economics — and where is the intersection nobody occupies?”
Editorial layer over the CEcon landscape · 76 researchers · 5,300+ papers
A research strategy is a map of occupied territory. Complexity economics is studied in roughly eight schools across a dozen institutions; a small subset of its people have actually run money, and what that subset produced clusters tightly around liquid-market microstructure. The gaps are not in the methods — they are in where the methods have been pointed. This page is the editorial layer over the CEcon lab landscape (76 researchers, 5,300+ papers ingested via Semantic Scholar); the lanes on the Research tab are positioned inside the white space mapped here.
Every sub-discipline placed by its value to investing and trading against how thoroughly practitioners have already mined it. Filled burgundy dots are the white space the research lanes claim; hollow dots are occupied ground. Hover or tap any dot for the schools behind it. Scores are editorial judgments, stated so they can be argued with.
Climate & physical transmission
Value 8.7/10 · Mined 0.7/10Where · Iowa State (electricity ABM) · Oxford INET energy group · Zurich climate-finance — none price-facing; desks do it privately — Tesfatsion, Farmer’s energy group, Battiston
Value · Physical state variables are genuinely exogenous causal instruments — the identification quality causal financial ML demands and almost never gets.
Mined · Publicly near-empty: academics stop at policy, desks publish nothing.
Claimed by · Research Lanes I–III and VII — the core of the climate-grids-balance-sheets program, extended to the full firm population by the Observable Economy lane.
| Discipline | Value | Mined | Why it matters to investing & trading | Where |
|---|---|---|---|---|
| Market microstructure & impact | 9.2 | 9.3 | Directly monetized for three decades — impact laws and order-flow dynamics are live inputs to execution and alpha at every systematic desk. | CFM & École Polytechnique · Oxford Mathematical Institute · Boston University · Palermo |
| Causal financial ML | 8.8 | 7.3 | Backtest-overfitting discipline and causal identification — the current standard for whether quant research is believable at all. | Khalifa University · ADIA Lab · Cornell ORIE |
| Climate & physical transmission | 8.7 | 0.7 | Physical state variables are genuinely exogenous causal instruments — the identification quality causal financial ML demands and almost never gets. | Iowa State (electricity ABM) · Oxford INET energy group · Zurich climate-finance — none price-facing; desks do it privately |
| Tail risk & extreme value | 8.3 | 8.3 | Tail hedging is a productized category with a public convexity track record. | NYU Tandon · Universa · EVT groups |
| Market ecology | 7.5 | 1 | Strategies as species, markets as ecosystems — if made empirical, it predicts crowding and regime shifts, which is directly monetizable. | Oxford INET — theory only, no empirical site |
| Leverage cycles & credit | 7.3 | 5.4 | Collateral rates as the hidden state variable of booms and busts — learned on an MBS desk, still underused outside credit. | Yale · Ellington Management · US Office of Financial Research |
| Technology forecasting | 6.9 | 2.3 | Wright’s-law cost curves called the solar and battery declines that consensus missed — directly relevant to thematic and energy-transition books. | Oxford INET · Santa Fe Institute |
| Performativity & conventions | 6.5 | 0.6 | Knowing which conventions constitute prices is meta-alpha — it tells you which anomalies are structural and which will be arbitraged. | Edinburgh · Sciences Po / Mines Paris · LSE |
| Bubble & crash diagnostics | 6.2 | 6.3 | LPPLS survives as one input among many on vol desks; contested empirical record caps its standalone value. | ETH Zurich · SUSTech · Financial Crisis Observatory |
| Production & supply-chain networks | 6 | 3.2 | Shock propagation through input-output networks — proven relevant by COVID, tradable in single names, rarely traded systematically. | MIT · Oxford INET · Paris 1 |
| Ergodicity economics | 5.5 | 1.6 | Time-average reasoning bears directly on position sizing and leverage — Kelly logic with a physics pedigree. | London Mathematical Laboratory · SFI orbit |
| Adaptive markets | 5.4 | 5.7 | Evolution replacing efficiency — an influential frame with mixed fund results. | MIT Laboratory for Financial Engineering · AlphaSimplex |
| Financial networks & systemic risk | 5 | 3.9 | Risk-management value more than alpha — contagion topology tells you what breaks, not what to buy. | CSH Vienna · University of Zurich · Bank of England · ECB · OFR |
| Heterogeneous-agent finance | 4.6 | 3.5 | Switching and herding models reproduce the stylized facts — explanatory value high, predictive value modest. | Amsterdam CeNDEF · Brandeis |
| Economic complexity (ECI) | 4.3 | 1.9 | Capability-based growth prediction — relevant to EM sovereign macro on multi-year horizons, too slow for most books. | Harvard Growth Lab · Toulouse · Rome |
| Distributive incidence | 3.7 | 0.4 | Low direct trading value — its currency is research capital: the inequality frame SFI and ADIA fund, with market data nobody else brings. | SFI · Columbia/Barnard · energy-poverty groups · the Abu Dhabi winter school itself |
| Agent-based macro | 2.6 | 1 | Policy experiments, not signals — the furthest school from prices. | Sant’Anna Pisa · Milan Cattolica · Bielefeld · Kiel · Genoa |
Everyone who trades studies liquid-market microstructure; everyone who studies physical systems does not trade. The strategy is the unoccupied intersection: physically driven, fully observable small markets — traded in the Armstrong book, published through CEcon, framed by conventions and incidence for SFI and Abu Dhabi. Not a new method; the field’s existing methods, pointed somewhere they have never been pointed, by someone holding both the book and the pen.
The underlying dataset — 76 researchers tiered core/adjacent/peripheral with Semantic Scholar graphs, 5,300+ papers tagged across eleven domains, temporal and cross-domain gap annotations — lives in the CEcon platform built with Michael Ralph at complexity-economics.org. This page is the strategy read on top of that scaffolding; the landscape and researcher graphs there are the evidence base for the map here. complexity-economics.org / landscape →