ºÝºÝߣ

ºÝºÝߣShare a Scribd company logo

John Rutledge, Claremont Graduate University February 14, 2012

?Download as PPT, PDF?
?
John Rutledge
John Rutledge

This lecture deals with recent developments in an area I call Far-From-Equilibrium economics and applies the work to both the recent financial crisis and optimal portfolio theory.

1 of 17
Download to read offline
Dr. John Rutledge Claremont Graduate University February 14, 2012 Far From Equilibrium Economics: Network Failure, Credit Crisis, and Optimal Portfolios
Our Fragile Island Economy Capital $200 trillion GDP/Work $15 trillion
Earth¡¯s Thin, Fragile Crust
Hayek Neural Networks Price system as information network Neural network (Sensory Order) System of telecommunications Price as a kind of symbol Prices coordinate actions of different people Solves division of knowledge problem One of the great triumphs of the human mind Division of knowledge the really central problem of economics Only the most essential information is passed on¡­only to those who need it Perfect markets, equilibrium are tautologies Change in knowledge disrupts equilibrium Economic problem, rapid adaptation to change Suggests 2 states (regimes), phase transition Price clearing vs. Non-price clearing regimes As in Edward Lorenz (1963) weather model Lorenz Attractor Lorenz Equations Deterministic Non-periodic flows
Evolutionary Economics Early work: Malthus-Darwin (1869), Marx (1867), Marshall, Veblen (1898 Why is Economics¡­? ), Schumpeter, Hayek, von Mises Nelson, Winter (1982). Carrier of information, knowledge, rules as genes. New models of agent behavior: behavioral, experimental, neuro New models of interaction: complex adaptive systems, dissipative structures, directed graph theory, percolation theory, de Donder, Schrodinger, Prigogine, Haken, Mandelbrot, Lorenz, Barabasi, Arthur Empirical work: Farmer, Lillo, Geanakoplos, Lo, Allen, Gale, Chen, Cont Synthesizers: Schweitzer, Sornette, Foster, Witt, Dopfer, Metcalfe, Potts, Hodgson Results: both agent behavior and interactions lead to cascading failures, power law signatures of far from equilibrium phase transitions between regimes Applications: Macroeconomics: financial crises, contagion, recession, depression Finance: Portfolio Theory
Credit Crises: Cascading Network Failures Network Architecture Network Core Core Shift Cascading Failure System Regression Savonarolla Effect Non-Price Clearing regime
Debt, Deflation Theory of Great Depressions Irving Fisher (1933) Cycle theory creed --Booms and Depressions (1932) ¡° Equilibrium¡­seldom reached and never long maintained.¡± p. 339 Disequilibrium¡­delicately poised¡­beyond certain limits instability ensues¡­breaking of many debtors constitutes a crash, after which no coming back to original equilibrium (sound familiar?) Two dominant factors, over-indebtedness (debt disease) to start with and deflation (dollar disease) following soon after. (debt is a network property) Debt starters¡­new opportunities¡­easy money cause of over-borrowing. Crises 1837, 1873, 1893 (2001?, 2007?) Public psychology for going into debt¡­lure of future income¡­hope of immediate capital gain¡­reckless promotions¡­scandals, frauds¡­always real basis
Fisher¡¯s Key Result: Liquidating Debt Defeats Itself Debt liquidation leads to distress selling¡­All fluctuations come about through a fall of prices Liquidation defeats itself¡­the more debtors pay the more they owe¡­the very effort of individuals to lessen the burden of their debts increases it because of the mass effect of the stampede to liquidate in swelling each dollar owed 1929 debt greatest known to that time¡­by 1933 liquidation reduced debt by 20% but increased the dollar 75%...real debt increased 40% Vicious spiral many years¡­universal bankruptcy¡­natural way out of depression¡­needless cruel bankruptcy, unemployment starvation¡­political revolution¡­reflation Controlling price level new importance¡­those in drivers¡¯ seat will be held to a new accountability
Portfolio Theory MPT, CAPM, APT Return distribution Fixed mean, variance, covariance Efficient frontier, optimal portfolio Risk free rate, market portfolio Asset allocation industry, Pension Act of 2006 MPT Does an Adequate Job in Normal Market Conditions MPT Fails During Credit Market Crises Credit markets fail, banks call loans Volatility jumps, clusters Correlations converge on 1.0 Asset prices fall across the board Optimal investors have little cash Forced asset sales at deep discounts Contagion between portfolio, businesses Makes financial crises worse
Credit Crises: Correlations Break Ddown Correlations are transitory by nature Correlations are not fundamental parameters of nature Secular increase in correlations Correlations approach 1.0 in times of crisis Diversifications benefits not realized Optimal portfolios lead to forced selling
Evolutionary Economics and Portfolio Theory Random distribution rules out big changes (exponentially small), mean, std dev Actual time series frequent big changes Power Law: N(x) = x ¨C ? ??? N(x) is number of movements of size x DJIA data Log N(x) = -3.96x -3.3 R 2 =.97 Random OK small, hopeless for large changes P(3%) = random 718 per century, actual 780. P(6%) = random 1 per century, actual 57 P(8%) = random 1 per 10 6 centuries, actual 11 P(10%) = power law predicts 6, actual 8. Fat tails, clustered volatility signatures Implications: non-equilibrium, phase transition between two states (regimes) Equilibrium regime Non-equilibrium regime (cascading failure)
Credit Crisis ¨C 2 Period Model Credit Crisis Non-price credit rationing Reduction in Lending Reported interest rate falls Unreported shadow price (R cc ) rises Opportunity cost of cash soars Optimal portfolio - credit crisis Opportunity cost of cash > r(E) All cash corner solution But 2 period model unsatisfying Does not allow return to equilibrium state View as statistical problem? Opportunity cost of cash distribution R(credit crunch) = 20%, r(equil) = 2% P(credit crunch) = 0.3 Expected value = 7.4% Standard deviation = 8.4% (no risk free asset) Correlation with risk asset? (certainly < 0) Conclusion: 2 period model unsatisfying
Credit Crisis - Optimal Investor 3 period model
Credit Crisis ¨C Cash-Rich Investor 3 period model
Summary Opportunity cost of cash Higher than risk-free rate on average Much higher than risk free rate during credit crises There is no risk-free (zero variance) asset Negatively correlated (-0.8%) with other risky asset returns Casualties: Cash-rich portfolios are optimal with credit crisis regimes Corner solutions possible/likely Tobin separation theorem no longer holds There is no longer a unique market portfolio of risky assets Extensions: Forced liquidation of business operating assets Collateral (Geanakoplos), credit channel, employment and output effects Empirical tests using mezzanine data Separate return distribution into equilibrium, non-equilibrium regimes Hypothesis: Equilibrium regime distribution Gaussian; non-equilibrium regime power law, fat tails Recognize that phase transition (network failure), not standard deviation, is key risk-return tradeoff Incorporate phase transition risk into portfolio theory
Next Project: Value Investing Intrinsic Value, Intrinsic Risk Tradeoff
Dr. John Rutledge Claremont Graduate University February 14, 2012 Far From Equilibrium Economics: Network Failure, Credit Crisis, and Optimal Portfolios

More Related Content

John Rutledge, Claremont Graduate University February 14, 2012

  • 1. Dr. John Rutledge Claremont Graduate University February 14, 2012 Far From Equilibrium Economics: Network Failure, Credit Crisis, and Optimal Portfolios
  • 2. Our Fragile Island Economy Capital $200 trillion GDP/Work $15 trillion
  • 4. Hayek Neural Networks Price system as information network Neural network (Sensory Order) System of telecommunications Price as a kind of symbol Prices coordinate actions of different people Solves division of knowledge problem One of the great triumphs of the human mind Division of knowledge the really central problem of economics Only the most essential information is passed on¡­only to those who need it Perfect markets, equilibrium are tautologies Change in knowledge disrupts equilibrium Economic problem, rapid adaptation to change Suggests 2 states (regimes), phase transition Price clearing vs. Non-price clearing regimes As in Edward Lorenz (1963) weather model Lorenz Attractor Lorenz Equations Deterministic Non-periodic flows
  • 5. Evolutionary Economics Early work: Malthus-Darwin (1869), Marx (1867), Marshall, Veblen (1898 Why is Economics¡­? ), Schumpeter, Hayek, von Mises Nelson, Winter (1982). Carrier of information, knowledge, rules as genes. New models of agent behavior: behavioral, experimental, neuro New models of interaction: complex adaptive systems, dissipative structures, directed graph theory, percolation theory, de Donder, Schrodinger, Prigogine, Haken, Mandelbrot, Lorenz, Barabasi, Arthur Empirical work: Farmer, Lillo, Geanakoplos, Lo, Allen, Gale, Chen, Cont Synthesizers: Schweitzer, Sornette, Foster, Witt, Dopfer, Metcalfe, Potts, Hodgson Results: both agent behavior and interactions lead to cascading failures, power law signatures of far from equilibrium phase transitions between regimes Applications: Macroeconomics: financial crises, contagion, recession, depression Finance: Portfolio Theory
  • 6. Credit Crises: Cascading Network Failures Network Architecture Network Core Core Shift Cascading Failure System Regression Savonarolla Effect Non-Price Clearing regime
  • 7. Debt, Deflation Theory of Great Depressions Irving Fisher (1933) Cycle theory creed --Booms and Depressions (1932) ¡° Equilibrium¡­seldom reached and never long maintained.¡± p. 339 Disequilibrium¡­delicately poised¡­beyond certain limits instability ensues¡­breaking of many debtors constitutes a crash, after which no coming back to original equilibrium (sound familiar?) Two dominant factors, over-indebtedness (debt disease) to start with and deflation (dollar disease) following soon after. (debt is a network property) Debt starters¡­new opportunities¡­easy money cause of over-borrowing. Crises 1837, 1873, 1893 (2001?, 2007?) Public psychology for going into debt¡­lure of future income¡­hope of immediate capital gain¡­reckless promotions¡­scandals, frauds¡­always real basis
  • 8. Fisher¡¯s Key Result: Liquidating Debt Defeats Itself Debt liquidation leads to distress selling¡­All fluctuations come about through a fall of prices Liquidation defeats itself¡­the more debtors pay the more they owe¡­the very effort of individuals to lessen the burden of their debts increases it because of the mass effect of the stampede to liquidate in swelling each dollar owed 1929 debt greatest known to that time¡­by 1933 liquidation reduced debt by 20% but increased the dollar 75%...real debt increased 40% Vicious spiral many years¡­universal bankruptcy¡­natural way out of depression¡­needless cruel bankruptcy, unemployment starvation¡­political revolution¡­reflation Controlling price level new importance¡­those in drivers¡¯ seat will be held to a new accountability
  • 9. Portfolio Theory MPT, CAPM, APT Return distribution Fixed mean, variance, covariance Efficient frontier, optimal portfolio Risk free rate, market portfolio Asset allocation industry, Pension Act of 2006 MPT Does an Adequate Job in Normal Market Conditions MPT Fails During Credit Market Crises Credit markets fail, banks call loans Volatility jumps, clusters Correlations converge on 1.0 Asset prices fall across the board Optimal investors have little cash Forced asset sales at deep discounts Contagion between portfolio, businesses Makes financial crises worse
  • 10. Credit Crises: Correlations Break Ddown Correlations are transitory by nature Correlations are not fundamental parameters of nature Secular increase in correlations Correlations approach 1.0 in times of crisis Diversifications benefits not realized Optimal portfolios lead to forced selling
  • 11. Evolutionary Economics and Portfolio Theory Random distribution rules out big changes (exponentially small), mean, std dev Actual time series frequent big changes Power Law: N(x) = x ¨C ? ??? N(x) is number of movements of size x DJIA data Log N(x) = -3.96x -3.3 R 2 =.97 Random OK small, hopeless for large changes P(3%) = random 718 per century, actual 780. P(6%) = random 1 per century, actual 57 P(8%) = random 1 per 10 6 centuries, actual 11 P(10%) = power law predicts 6, actual 8. Fat tails, clustered volatility signatures Implications: non-equilibrium, phase transition between two states (regimes) Equilibrium regime Non-equilibrium regime (cascading failure)
  • 12. Credit Crisis ¨C 2 Period Model Credit Crisis Non-price credit rationing Reduction in Lending Reported interest rate falls Unreported shadow price (R cc ) rises Opportunity cost of cash soars Optimal portfolio - credit crisis Opportunity cost of cash > r(E) All cash corner solution But 2 period model unsatisfying Does not allow return to equilibrium state View as statistical problem? Opportunity cost of cash distribution R(credit crunch) = 20%, r(equil) = 2% P(credit crunch) = 0.3 Expected value = 7.4% Standard deviation = 8.4% (no risk free asset) Correlation with risk asset? (certainly < 0) Conclusion: 2 period model unsatisfying
  • 13. Credit Crisis - Optimal Investor 3 period model
  • 14. Credit Crisis ¨C Cash-Rich Investor 3 period model
  • 15. Summary Opportunity cost of cash Higher than risk-free rate on average Much higher than risk free rate during credit crises There is no risk-free (zero variance) asset Negatively correlated (-0.8%) with other risky asset returns Casualties: Cash-rich portfolios are optimal with credit crisis regimes Corner solutions possible/likely Tobin separation theorem no longer holds There is no longer a unique market portfolio of risky assets Extensions: Forced liquidation of business operating assets Collateral (Geanakoplos), credit channel, employment and output effects Empirical tests using mezzanine data Separate return distribution into equilibrium, non-equilibrium regimes Hypothesis: Equilibrium regime distribution Gaussian; non-equilibrium regime power law, fat tails Recognize that phase transition (network failure), not standard deviation, is key risk-return tradeoff Incorporate phase transition risk into portfolio theory
  • 16. Next Project: Value Investing Intrinsic Value, Intrinsic Risk Tradeoff
  • 17. Dr. John Rutledge Claremont Graduate University February 14, 2012 Far From Equilibrium Economics: Network Failure, Credit Crisis, and Optimal Portfolios

Editor's Notes

  • #8: Fisher, Josiah Gibbs. (dissertation supervisor, Fisher financed Gibbs Collected Works Samuelson (1998, p. 1376). ¡°Perhaps most relevant of all for the geneses if Foundations , Edwin Bidwell Wilson was at Harvard. Wilson was the great (Josiah) Willard Gibbs¡¯s last (and, essentially only) prot¨¦g¨¦ at Yale . He was a mathematician, a mathematical physicist, a mathematical statistician, a mathematical economist, a polymath who had done first-class work in many fields of the natural and social sciences. I was perhaps his only disciple ¡­ I was vaccinated early to understand that economics and physics could share the same formal mathematical theorems.¡± Gibbs products, coined term statistical mechanics (explain laws of thermo using stat properties of ensembles of particles, with Boltzmann, Maxwell), invented vector calculus, Gibbs free energy. 1 st US PHD in engineering from Yale in 1863 Einstein¡­¡±greatest mind in American history¡± (letters to Poincare, Hilbert, Boltzmann, Mach, le Chatalier, van der Waals, Planck