Research and writing

Questions across scales.

From the collective behavior of neural networks to the economics of growth. Research, working papers and authored arguments, with their collaborators and original publications.

A study of scale and growth

Minimum
Viable Scale

Extinction and Escape under Increasing Returns

AuthorsPresident Javier Milei & Demian Reidel

July 2026 · Working paper

The paper separates the scale needed to sustain capital, the conditions needed to meet required returns, and the starting points from which growth is optimal.

Explore the paper

Minimum Viable Scale · Javier Milei, Demian Reidel · July 2026

Research and publications

Economics. Physics.
A continuing inquiry.

Scale and growth

When can an economy sustain itself, and when should it expand?

2026

Public summary manuscript

When regulation kills growth: Summary for Davos

Javier Milei, Demian Reidel

21 January 2026

Increasing returns, minimum viable labor scale and stationary outcomes. An earlier, separately titled manuscript in the same research programme.

Money across borders

Prices, exchange rates and the behavior of international financial markets.

2006

Doctoral dissertation

Essays in international economics and finance

Demian Axel Reidel

Harvard University, 2006

Doctoral research in international economics and finance, supervised by Kenneth S. Rogoff.

2004

Working paper and preserved slides

A Biased View of PPP

Demian Reidel, Jan Szilagyi

Harvard, 22 September 2004

Purchasing power parity asks how exchange rates relate to differences in prices between countries. With Jan Szilagyi, I examined how sectors and small samples can bias estimates of the speed of exchange rate adjustment. A dated 26 slide excerpt survives in university course materials.

Networks and randomness

My early physics work studied collective behavior, memory and mathematical methods for systems that evolve with randomness.

1997

Journal article

Statistical properties of random asymmetrical neural networks

E. N. Miranda, D. H. Zanette, D. Reidel

Physica A 241(3–4), 481–492 (1997) · DOI: 10.1016/S0378-4371(97)00178-7

What can a network do before it has been trained? We studied fully connected artificial neurons with randomly assigned couplings and thresholds. As the model’s parameters change, it shifts from one stable outcome to two possible stable outcomes. The analysis connects collective neural behavior with phase transitions in statistical physics.

1996

Journal article

Space-time transformations within the path-integral approach to stochastic processes

C. D. Batista, G. Drazer, D. Reidel, H. S. Wio

Physical Review E 54(1), 86–91 (1996) · DOI: 10.1103/PhysRevE.54.86

A change of variables can turn a difficult random process into a simpler problem. We explored transformations of space and time within the path integral method, using strongly damped motion in a changing harmonic potential as the worked example.

1994

Physics thesis

Redes neuronales: almacenamiento de secuencias y propiedades estadísticas

Demian Reidel

Instituto Balseiro, 1994

My physics thesis studied how neural networks store sequences and how their statistical properties arise from their many interacting parts. Supervised by Enrique Miranda and Damián Zanette at Instituto Balseiro.

Authored writing