Can financial crashes be anticipated using physics?
summary
Claims that conformal field theory diagnostics flag endogenous bubble crashes by tracking a rising anomalous dimension derived from log-price correlators. Method combines Sornette’s log-periodic power-law intuition, multi-agent scaling, and a 1D non-equilibrium CFT fit; applied to the 50 most liquid S&P names around the recent inflationary drawdown. Result: the anomalous-dimension trend rises sharply before crashes, outperforming autocorrelation, skewness and kurtosis signals (which fail or reverse sign).
AI-written summary · about
chapters (21)
- 0:00Introduction
- 1:10What are bubbles and crashes
- 2:33Alternative models
- 4:19Exogenous vs endogenous crashes
- 6:00Irrational exuberance
- 7:48Bubbles
- 9:21Early warning signals
- 13:24Network effects
- 13:58Symmetry groups
- 15:39Symmetry in finance
- 16:43Selfsimilarity
- 20:33Groups
- 22:25Options
- 26:23Liftshield script
- 27:37Dynamics
- 29:24Log periodic power law
- 31:08Its messy to fit
- 31:59Power loss scaling
- 32:29Nonlinearity
- 33:36Other signals
- 34:09Conclusion
related talks
34mFrom Black Scholes to Black Holes
22mThe Bitcoin Power Law Network Physics
01h 03mSampling Crash Course (continued)
30mAI vs Physics in Portfolio Optimisation
24mMachine learning and financial time series analysis
35mEasy Bayesian regularization for fitting financial time series and curves
38mLearning credit networks with applications in financial industry
21mPredicting Bond Prices: Regime Changes