Models

Models

1. VAR

   • Single asset class
   • Bell curve
   • Normal market conditions
   • Exacerbate diversification
   • Separate lending and risk
     • Fundamental rules of banking: default risk is inevitable liability of business
   • Turner Review
     • ST observation period.
     • Network risks-> self enforcing cycles.
     • Paradox: Systemic risk highest when measured risk lowest
       • Take on more risk
     • Risk vs uncertainty and systemic vs idiosyncratic
   • Apotheosis of 50 years of FE.
   • Complete fallacy-> blind faith in models.
   • Net risk less than risk of either two investments.
   • Endogenous risk e.g. uncertainty.

2. Gaussian Copula

   • Simple elegant ease accuracy ubiquitous
   • Spark of mathematical legerdemain
   • CDS prices-> implicit assumption price risk correctly.
   • No look at underlying
   • Constant correlation rather than mercurial.
     • Taleb- charlatanism.
   • Black boxes.
   • Regulators risk measures derived from model
     • Legitimacy
   • Conformity-> compounded risk.

3. Robhini

   • Not black swan-> white swan events remarkable predictable.
   • Enthusiastically embraced 20% yearly increase in prices.

4. Frank Knight

   • Uncertainty cannot be priced

5. Fama

   • Markets are utterly rational and efficient.
   • Random walk became conventional wisdom.

6. My View

   • Religious faith in models
   • Financial wizards carried away in euphoria
   • During market upheaval 'all correlations go to one'.
   • Exacerbated by one stop shop

7. Stiglitz

   • 1987 'Once in a lifetime crash-> one in every billion years.
     • 10 years later LCTM.
   • Financial markets do not learn-> people running models do not look at history.

8. Rajan

   • Linkages between markets and institutions.
     • Exposed to large systemic shocks.
     • Fat tail risks.
       • Keynes
         • Future inherently unpredictable-> no mode.
     • Euphoria leads to SS.

9. PA Problem

   • Moral hazard.
     • Myopic destructive incentives

10. Copycat behaviour e.g. Asch visual experiment feel left out.

    • Crazy ones those not in market.

11. Cassidy

    • Market strong self-equalibvrating tendencies.
      • GD historic anomaly-> error of MP converting recession into cataclysm
    • Disaster Myopia
    • Representative Heuristic
    • Rational irrationality
    • Prisoners dilemma
    • Drunks staggering down the street
    • Synchronous lateral excitation

12. Samuelson

    • Physics of avalanches
      • No equilibrium relaitons-> economic dislocations.
      • Displace snow in path which adds weight and force

13. Mansharamani

    • LCTM never had a monthly loss greater than 3% prior to 1998.
      • Likelihood 1 in 50 million.
    • Over/undershooting no unique equilibrium no self correcting self reinforcing.

14. Low cross regional correlation.

15. LCH

    • Take no account of most important determinant-> self control.