Definition of the Subject
Empirical studies show that there are at least some components in future asset returns that are predictable using information that is currentlyavailable. When the linear time series models are employed in prediction, the accuracy of the forecast does not depend on the current return or theinitial condition. In contrast, with nonlinear time series models, properties of the forecast error depend on the initial value or the history. Theeffect of the difference in initial values in a stable nonlinear model, however, usually dies out quickly as the forecast horizon increases. For bothdeterministic and stochastic cases, the dynamic system is chaos if a small difference in the initial value is amplified at an exponential rate. Ina chaotic nonlinear model, the reliability of the forecast can decrease dramatically even for a moderate forecast horizon. Thus, the knowledgeof the sensitive dependence on initial conditions in a particular financial time series offers...
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsAbbreviations
- Global Lyapunov exponent:
-
A global stability measure of the nonlinear dynamic system. It is a long‐run average of the exponential growth rate of infinitesimally small initial deviation and is uniquely determined in the ergodic and stationary case. In this sense, this initial value sensitivity measure does not depend on the initial value. A system with positive Lyapunov exponents is considered chaotic for both deterministic and stochastic cases.
- Local Lyapunov exponent:
-
A local stability measure based on a short‐run average of the exponential growth rate of infinitesimally small initial deviations. Unlike the global Lyapunov exponent, this initial value sensitivity measure depends both on the initial value and the horizon for the average calculation. A smaller local Lyapunov exponent implies a better performance at the point of forecast.
- Noise amplification:
-
In a stochastic system with the additive noise, the effect of shocks can either grow, remain, or die out with the forecast horizon. If the system is nonlinear, this effect depends both on the initial value and size of the shock. For a chaotic system, the degree of noise amplification is so high that it makes the forecast almost identical to the iid forecast within the next few steps ahead.
- Nonlinear impulse response function:
-
In a stochastic system with the additive noise, the effect of shocks on the variable in subsequent periods can be summarized in impulse response functions. If the system is linear, the impulse response does not depend on the initial value and its shape is proportional to the size of shocks. If the system is nonlinear, however, the impulse response depends on the initial value, or the history, and its shape is no longer proportional to the size of shocks.
Bibliography
Abhyankar A, Copeland LS, Wong W (1995) Nonlinear dynamics in real-time equitymarket indices: evidence from the United Kingdom. Econ J 105:864–880
Abhyankar A, Copeland LS, Wong W (1997) Uncovering nonlinear structure inreal-time stock-market indexes: The S&P 500, the DAX, the Nikkei 225, and the FTSE-100. J Bus Econ Stat 15:1–14
Andersen TB, Bollerslev T, Diebold FX, Labys P (2003) Modeling and forecastingrealized volatility. Econometrica 71:579–625
Bailey BA, Ellner S, Nychka DW (1997) Chaos with confidence: Asymptotics andapplications of local Lyapunov exponents. In: Cutler CD, Kaplan DT (eds) Fields Institute Communications, vol 11. American Mathematical Society,Providence, pp 115–133
Barnett WA, Gallant AR, Hinich MJ, Jungeilges J, Kaplan D, Jensen MJ (1995)Robustness of nonlinearity and chaos tests to measurement error, inference method, and sample size. J Econ Behav Organ27:301–320
Barnett WA, Serletis A (2000) Martingales, nonlinearity, and chaos. J Econ DynControl 24:703–724
Bask M, de Luna X (2002) Characterizing the degree of stability of non-lineardynamic models. Stud Nonlinear Dyn Econom 6:3
Bollerslev T (1986) Generalized autoregressive conditional heteroskedasticity.J Econ 31:307–327
Brock WA, Dechert WD, Scheinkman JA, LeBaron B (1996) A test for independencebased on the correlation dimension. Econ Rev 15(3):197–235
Brock WA, Hommes CH (1998) Heterogeneous beliefs and routes to chaos in a simple asset pricing model. J Econ Dyn Control 22:1235–1274
Brock WA, Lakonishok J, LeBaron B (1992) Simple technical trading rules andthe stochastic properties of stock returns. J Financ 47:1731–1764
Campbell JY, Lo AW, MacKinlay AC (1997) The Econometrics of Financial Markets.Princeton University Press, Princeton
Campbell JY, Shiller R (1988) The dividend-price ratio and expectations offuture dividends and discount factors. Rev Financ Stud 1:195–228
Chan KS, Tong H (2001) Chaos: A Statistical Perspective. Springer, NewYork
Cheng B, Tong H (1992) On consistent nonparametric order determination andchaos. J Royal Stat Soc B 54:427–449
Cheung YW, Chinn MD, Pascual AG (2005) Empirical exchange rate models of thenineties: Are any fit to survive? J Int Money Financ 24:1150–1175
Christoffersen PF, Diebold FX (2006) Financial asset returns,direction-of-change forecasting, and volatility dynamics. Management Sci 52:1273–1287
Dechert WD, Gençay R (1992) Lyapunov exponents as a nonparametricdiagnostic for stability analysis. J Appl Econom 7:S41–S60
Diebold FX, Nason JA (1990) Nonparametric exchange rate prediction? J IntEcon 28:315–332
Ding Z, Granger CWJ, Engle RF (1993) A long memory property of stock marketreturns and a new model. J Empir Financ 1:83–106
Eckmann JP, Kamphorst SO, Ruelle D, Ciliberto S (1986) Liapunov exponents fromtime series. Phys Rev A 34:4971–4979
Eckmann JP, Ruelle D (1985) Ergodic theory of chaos and strange attractors.Rev Mod Phys 57:617–656
Engle RF (1982) Autoregressive conditional heteroskedasticity with estimatesof the variance of UK inflation. Econometrica 50:987–1008
Fama E (1970) Efficient capital markets: Review of theory and empirical work.J Financ 25:383–417
Fama E, French K (1988) Permanent and temporary components of stock prices. J Political Econ 96:246–273
Fama E, French K (1988) Dividend yields and expected stock returns. J FinancEcon 22:3–5
Fan J, Yao Q (2003) Nonlinear Time Series: Nonparametric and ParametricMethods. Springer, New York
Fan J, Yao Q, Tong H (1996) Estimation of conditional densities andsensitivity measures in nonlinear dynamical systems. Biometrika 83:189–206
Fernándes-Rodríguez F, Sosvilla-Rivero S, Andrada-Félix J (2005) Testingchaotic dynamics via Lyapunov exponents. J Appl Econom 20:911–930
Frank M, Stengos T (1989) Measuring the strangeness of gold and silver ratesof return. Rev Econ Stud 56:553–567
Gallant AR, Rossi PE, Tauchen G (1993) Nonlinear dynamic structures.Econometrica 61:871–907
Gençay R (1996) A statistical framework for testing chaotic dynamics viaLyapunov exponents. Physica D 89:261–266
Gençay R (1998) The predictability of security returns with simpletechnical trading rules. J Empir Financ 5:347–359
Gençay R (1999) Linear, non-linear and essential foreign exchange rateprediction with simple technical trading rules. J Int Econ 47:91–107
Giannerini S, Rosa R (2001) New resampling method to assess the accuracy ofthe maximal Lyapunov exponent estimation. Physica D 155:101–111
Grassberger P, Procaccia I (1983) Estimation of the Kolmogorov entropy from a chaotic signal. Phys Rev A 28:2591–2593
Hall P, Wolff RCL (1995) Properties of invariant distributions and Lyapunovexponents for chaotic logistic maps. J Royal Stat Soc B 57:439–452
Hommes CH, Manzan S (2006) Comments on Testing for nonlinear structure andchaos in economic time series. J Macroecon 28:169–174
Hong Y, Lee TH (2003) Inference on predictability of foreign exchange ratesvia generalized spectrum and nonlinear time series models. Rev Econ Stat 85:1048–1062
Hsieh DA (1989) Testing for nonlinear dependence in daily foreign exchangerates. J Bus 62:339–368
Hsieh DA (1991) Chaos and nonlinear dynamics: application to financialmarkets. J Financ 46:1839–1877
Koop G, Pesaran MH, Potter SM (1996) Impulse response analysis in nonlinearmultivariate models. J Econom 74:119–147
Kuan CM, Liu T (1995) Forecasting exchange rates using feedforward andrecurrent neural networks. J Appl Econom 10:347–364
LeBaron B (1992) Some relation between the volatility and serial correlationsin stock market returns. J Bus 65:199–219
Lin WL (1997) Impulse response function for conditional volatility in GARCHmodels. J Bus Econ Stat 15:15–25
Linton OB (2008) Semiparametric and nonparametric ARCH modelling. In: AndersonTG, Davis RA, Kreiss JP, Mikosch T (ed) Handbook of Financial Time Series. Springer, Berlin
Lo AW, MacKinlay AC (1988) Stock market prices do not follow random walks:evidence from a simple specification test. Rev Financ Stud 1:41–66
Lu ZQ, Smith RL (1997) Estimating local Lyapunov exponents. In: Cutler CD,Kaplan DT (eds) Fields Institute Communications, vol 11. American Mathematical Society, Providence,pp 135–151
Maasoumi E, Racine J (2002) Entropy and predictability of stock marketreturns. J Econom 107:291–312
Mayfield ES, Mizrach B (1992) On determining the dimension of real time stockprice data. J Bus Econ Stat 10:367–374
McCaffrey DF, Ellner S, Gallant AR, Nychka DW (1992) Estimating the Lyapunovexponent of a chaotic system with nonparametric regression. J Am Stat Assoc 87:682–695
Meese R, Rogoff K (1983) Exchange rate models of the seventies. Do they fitout of sample? J Int Econ 14:3–24
Nelson DB (1990) Conditional heteroskedasticity in asset returns: A newapproach. Econometrica 59:347–370
Nychka D, Ellner S, Gallant AR, McCaffrey D (1992) Finding chaos in noisysystem. J Royal Stat Soc B 54:399–426
Pesaran MH, Timmermann A (1995) Predictability of stock returns: robustnessand economic significance. J Financ 50:1201–1228
Pesin JB (1977) Characteristic Liapunov exponents and smooth ergodic theory.Russ Math Surv 32:55–114
Poterba JM, Summers LH (1988) Mean reversion in stock prices: evidence andimplications. J Financ Econ 22:27–59
Potter SM (2000) Nonlinear impulse response functions. J Econ Dyn Control24:1425–1446
Qi M (1999) Nonlinear predictability of stock returns using financial andeconomic variables. J Bus Econ Stat 17:419–429
Qi M, Maddala GS (1999) Economic factors and the stock market: a newperspective. J Forecast 18:151–166
Racine J (2001) On the nonlinear predictability of stock returns usingfinancial and economic variables. J Bus Econ Stat 19:380–382
Richardson M, Stock JH (1989) Drawing inferences from statistics based onmultiyear asset returns. J Financ Econ 25:323–348
Rosenstein MT, Collins JJ, De Luca CJ (1993) A practical method forcalculating largest Lyapunov exponents from small data sets. Physica D 65:117–134
Scheinkman JA, LeBaron B (1989) Nonlinear dynamics and stock returns. J Bus62:311–337
Schittenkopf C, Dorffner G, Dockner EJ (2000) On nonlinear, stochasticdynamics in economic and financial time series. Stud Nonlinear Dyn Econom 4:101–121
Shintani M (2006) A nonparametric measure of convergence towards purchasingpower parity. J Appl Econom 21:589–604
Shintani M, Linton O (2003) Is there chaos in the world economy? A nonparametric test using consistent standard errors. Int Econ Rev 44:331–358
Shintani M, Linton O (2004) Nonparametric neural network estimation ofLyapunov exponents and a direct test for chaos. J Econom 120:1–33
Taylor SJ (1986) Modelling Financial Time Series. Wiley, New York
Tjostheim D, Auestad BH (1994) Nonparametric identification of nonlinear timeseries: Selecting significant lags. J Am Stat Assoc 89:1410–1419
Tschernig R, Yang L (2000) Nonparametric lag selection for time series. J Time Ser Analysis 21:457–487
Tschernig R, Yang L (2000) Nonparametric estimation of generalized impulseresponse functions. Michigan State University, unpublished
Valkanov R (2003) Long-horizon regressions: theoretical results andapplications. J Financ Econom 68:201–232
Whang YJ, Linton O (1999) The asymptotic distribution of nonparametricestimates of the Lyapunov exponent for stochastic time series. J Econom 91:1–42
White H (1988) Economic prediction using neural networks: the case of IBMstock returns. Proceedings of the IEEE International Conference on Neural Networks 2. The Institute of Electrical and Electronics Engineers, San Diego,pp 451–458
White H, Racine J (2001) Statistical inference, the bootstrap, andneural-network modeling with application to foreign exchange rates. IEEE Trans Neural Netw 12:657–673
Wolf A, Swift JB, Swinney HL, Vastano JA (1985) Determining Lyapunov exponentsfrom a time series. Physica D 16:285–317
Wolff RCL (1992) Local Lyapunov exponent: Looking closely at chaos. J RoyalStat Soc B 54:353–371
Wolff R, Yao Q, Tong H (2004) Statistical tests for Lyapunov exponents ofdeterministic systems. Stud Nonlinear Dyn Econom 8:10
Yao Q, Tong H (1994) Quantifying the influence of initial values on non-linearprediction. J Royal Stat Soc Ser B 56:701–725
Yao Q, Tong H (1994) On prediction and chaos in stochastic systems. PhilosTrans Royal Soc Lond A 348:357–369
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 Springer-Verlag
About this entry
Cite this entry
Shintani, M. (2009). Financial Forecasting, Sensitive Dependence. In: Meyers, R. (eds) Encyclopedia of Complexity and Systems Science. Springer, New York, NY. https://doi.org/10.1007/978-0-387-30440-3_209
Download citation
DOI: https://doi.org/10.1007/978-0-387-30440-3_209
Publisher Name: Springer, New York, NY
Print ISBN: 978-0-387-75888-6
Online ISBN: 978-0-387-30440-3
eBook Packages: Physics and AstronomyReference Module Physical and Materials ScienceReference Module Chemistry, Materials and Physics