Forecasting Financial Markets

Nonlinear Dynamics

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 1
 Overview
¾ Fractal distributions
¾ ARFIMA models
¾ Chaotic systems
¾ Phase space
¾ Correlation integrals
¾ Lyapunov exponents

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 2
 Fractal Distributions
¾ Problems with Gaussian theory of financial
markets
ƒ Non-normal distribution of returns
• Fat tails
• Peaked
¾ Pareto (1897)
ƒ Found that proportion of people owning huge
amounts of wealth was far higher than predicted by
(log) normal distribution
ƒ “Fat-tails”
ƒ Many examples in nature

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 3
 Pareto-Levy Distributions
¾ Levy (1935) generalized Pareto’s law
ƒ Described family of fat-tailed, high-peak pdf’s
¾ Pareto-Levy density functions
ƒ Ln[f(t)] = iδt - γ|t|α(1+iβ(t/|t|)tan(απ/2)
ƒ Parameters
• α is measure of peakedness
• β is measure of skewness (range +/- 1)
• γ is scale parameter
• δ is location parameter of the mean

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 4
 Characteristics of Pareto-Levy
¾ α is fractal dimension of probability space
ƒ α=1/H
ƒ 0<α<2
• If α = 2, (β = 0, γ = δ = 1) distribution is Normal
ƒ EMH: α = 2; FMH: 1 < α < 2
ƒ Self-similarity
• If distribution of daily returns has α = a, so will
distribution of 5-day returns
ƒ Variance undefined for 1 <= α < 2
ƒ Mean undefined for α < 1

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 5
 Undefined Variance
¾ Example: Volatility in the S&P 500 Index
S&P 500 Index Roling 12 Month Volatility

200%

150%

100%

50%

0%
76

78

80

82

84

86

88

90

92

94

96

98
-

-
ar

ar

ar

ar

ar

ar

ar

ar

ar

ar

ar

ar
M

Copyright © 1999 – 2006 Investment Analytics M

Forecasting Financial Markets – Nonlinear Dynamics Slide: 6
 ARFIMA Models
¾ Generalized ARIMA models
ƒ ARFIMA(p,d,q)
• Fractional differencing parameter d = H - 0.5
¾ Models fractal Brownian motion
ƒ Short memory effects
ƒ Long memory effects

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 7
 ARFIMA(0, d, 0)
¾ No short memory effects
¾ Long memory depends on parameter d
ƒ 0 < d < 0.5: black noise
ƒ -0.5 < d < 0: pink noise
ƒ D = 0: white noise
ƒ D = 1: brown noise

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 8
 ARFIMA(0, d, 0)
¾ d < 0.5
ƒ {yt} is stationary
ƒ Represented as infinite MA process
∞
yt = ∑ψ k ε t − k
k =0

(k + d − 1)!
ψk =
k!(d − 1)!

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 9
 ARFIMA(0, d, 0)
¾ d > - 0.5
ƒ {yt} is invertible
ƒ Represented as infinite AR process

∞
yt = ∑ π k yt − k
k =1

(k − d − 1)!
πk =
k!(d − 1)!
Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 10
 ARMA(0, d, 0)
¾ Covariance (−1) k (−2d )!
γk =
(k − d )!(− k − d )!
¾ Correlation
(− d )! 2 d −1
ρk ~ k as k → ∞
(d − 1)!

¾ Partial Autocorrelation
d
φ kk =
k −d
Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 11
 ARFIMA(1, d, 0)
¾ Process: (1 - αB)∆dyt = εt
ƒ Combines long and short term memory processes
¾ Correlation function

(− d )!(1 + α ) k 2 d −1
ρk = ×
− α ) Ffunction
(d − 1is)!the(1Hypergeometric
ƒ F(a,b;C,z)
2
(1,1 + d ;1 − d ;α )
¾ Example: AR(1) vs. ARFIMA(1,d, 0)
ƒ AR(1): a = 0.711
ƒ ARFIMA(1, d, 0): d = 0.2, a = 0.5

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 12
 AR(1) vs. ARFIMA(1, d, 0)
0.8

0.7

0.6
Correlation

0.5

0.4

0.3
AR(1)
0.2 ARFIMA(1,d,0)

0.1

0.0
0 5 10 15 20 25
Lag

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 13
 Estimating ARFIMA Models
¾ Step 1
ƒ Estimate d in ARIMA(0, d, 0) model ∆dyt = εt
• Use R/S analysis to estimate d
¾ Step 2
ƒ Define ut = ∆dyt
ƒ Use box-Jenkins analysis to fit model αBut = βBεt
¾ Step 3
ƒ Define xt = (βB)-1(αB) yt
¾ Step 4: estimate d in model ∆dxt = εt
¾ Repeat steps 2-4 until parameters converge

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 14
 Chaotic Systems
¾ Charaterized by:
ƒ Fractal dimension
ƒ Sensitivity to initial conditions
¾ Phase space
ƒ Scatter plot of system variables
ƒ Allows for visual inspection for patterns
¾ Attractors
ƒ Region in phase space where solutions lie
ƒ Can have Euclidean or fractal dimension

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 15
 Example: Logistic-Delay Function
¾ Equation: Xt = aXt-1 (1-Xt-2)
¾ Attractor dimension
ƒ Depends on constant a
• Point attractor (spiral) for smaller values of a
• Limit cycle (ellipse) as a increases

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 16
 Logistic Function: a = 1.8
The Logistic Function

0.7
0.6

0.5
0.4

0.3
0.2

0.1

0.0

Phase Space of the Logistic Function

0.70

0.60

0.50

0.40
Xt-1

0.30

0.20

0.10

0.00
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70

Xt

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 17
 Logistic Function: a = 2.2
The Logistic Function

1.0

0.8

0.6

0.4

0.2

0.0

Phase Space of the Logistic Function

1.00
0.90
0.80
0.70
0.60
Xt-1

0.50
0.40
0.30
0.20
0.10
0.00
0.00 0.20 0.40 0.60 0.80 1.00

Xt

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 18
 Fractal (“Strange”) Attractors
¾ Example: Henon attractor
¾ Equations
ƒ xt+1 = 1+yt - axt2
ƒ yt+1 = bxt
¾ Phase portrait shows strange attractor
ƒ Fractal dimension 1.2
ƒ 1 < D < 2 indicates presence of 2 variables in system

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 19
 Henon Attractor
Phase Portrait of the Henon Attractor

0.5
0.4
0.3
0.2
0.1
0.0
-1.5 -1.0 -0.5 -0.1 0.0 0.5 1.0 1.5
-0.2
-0.3
-0.4
-0.5

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 20
 Strange Attractors in the Capital
Markets?
¾ Examine phase portrait of financial time series
ƒ Fractal attractor would indicate chaotic
(i.e. deterministic) process.
ƒ Dimension of attractor would indicate # of system
variables
¾ Problem:
ƒ What is dimensionality of phase space?

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 21
 Constructing Phase Space
¾ Recall Henon attractor
ƒ Phase space constructed using scatterplot of two
variables X and Y
¾ Reconstruct phase space
ƒ Use scatterplot of Xt and Xt-1
ƒ Generates same map
ƒ Note: constructed using one variable, no equations

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 22
 Reconstructed Phase Portrait
for Henon Attractor
Reconstucted Phase Portrait

1.5

1.0

0.5
Xt-1

0.0
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5
-0.5

-1.0

-1.5

Xt

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 23
 Phase Space Dimensionality
¾ Ruelle: reconstructed vs. actual phase space
ƒ Same fractal dimension
ƒ Same Lyapunov spectrum
¾ Takens(1981):
ƒ Can reconstruct phase space by lagging time series for each
dimension
¾ Problem: what time lag to use?
ƒ i.e. What dimension is attractor?
• Need to embed it in higher dimension than its own
• Dimension of attractor does not change when embedded in higher
dimension (e.G. A plane in 3-D space still has 2-D)

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 24
 Determining Embedding
Dimensionality
¾ Wolf: mt = Q
• M = embedding dimension
• T = time lag
• Q = mean orbital period
¾ Example
ƒ If period is 48 iterations we require:
• 2 points lagged 24 iterations in 2-D space
• 3 points lagged 16 iterations in 3-D space

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 25
 Fractional Dimensionality
of Phase Space
¾ Time series
ƒ One variable
ƒ Dimensionality < 2
¾ Phase space
ƒ Includes all variables in system
ƒ Dimensionality depends on complexity of system
• May be > 3-D

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 26
 Correlation Dimension
¾ Correlation integral
ƒ Grassberger & Procaccia (1983)
ƒ Measures probability that pair of points in attractor are
within distance R of one another
ƒ Approximates fractal dimension
N
1
Cm ( R ) = 2
N
∑ Z (R− | X
i , j =1
i − X j |)
i≠ j

ƒ Z(x) = 1 if x > 0; 0 otherwise

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 27
 Estimating the Correlation Integral
¾ As R increases, Cm(r) should increase in
proportion to RD
ƒ Cm(R) ~ RD
¾ Log[Cm(R) ] = const + Dlog(R)
¾ Procedure
ƒ Measure Cm(r) for increasing values of R
ƒ Log-log plot of Cm(r) vs R
ƒ OLS estimate of slope is correlation dimension D for
embedding dimension m

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 28
 Correlation Integral of
the Henon Attractor
Correlation Dimension

0.0
-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4
-0.2

-0.4

y = 1.2014x - 0.5762
Log[Cm(R)]

-0.6
R2 = 0.999
-0.8

-1.0

-1.2

-1.4

-1.6

-1.8
Log(R)

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 29
 BDS Test for Randomness
¾ Brock, Dechert, Scheinkman (1987)
ƒ Lag time series {yt, t = 1, . . , T} in N lagged series
• Reconstruct n-dimensional phase space a la Takens
ƒ CN(R,T) → C1(R)N as T→ ∞
ƒ BDS test statistic
0 .5
T
WN ( R, T ) =| C N ( R, T ) − C1 ( R, T ) N | ×
σ N ( R, T )
• σN(r,t) is the SD of the correlation integrals
• W ~ No(0,1)
• For large W, reject the hypothesis that series is random
– Note will detect both linear and non-linear, so typically use AR(1)
residuals to filter out linear effects

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 30
 BDS Test of Financial Markets
Series Dimension W

Dow (20 day returns) 6 28.72

Yen (daily) 6 116.05
S&P 500 (weekly) 6 23.89

ƒ All tests based on AR(1) residuals of above series

ƒ Sources: Hsieh(1989), LeBaron(1990), Peters (1993)

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 31
 Lab: Estimating the Correlation
Dimension for the S&P 500 Index
¾ Monthly S&P index returns (AR(1) residuals)
ƒ Cycle estimated at 42 months from R/S analysis
¾ Estimate correlation dimension
ƒ Use embedding dimensions m = 5 to 10
ƒ Lags = Int[42 / m]
¾ Chart log[Cm(r)] vs log(r)
• M = 5 to 10
¾ Regression analysis
ƒ Estimate phase space dimensionality D
• OLS estimate of slope in log(r) = Dlog[Cm(r)]

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 32
 Solution: Estimating the Correlation
Dimension for the S&P 500 Index
Correlation Index of the S&P500 Index
0.0
-1.14 -1.12 -1.10 -1.08 -1.06 -1.04 -1.02 -1.00

Log(R)
-0.5

Log[C(r)]
-1.0

-1.5

-2.0

5 6 7 8 9 10 -2.5

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 33
 Solution: Estimating the Correlation
Dimension for the S&P 500 Index
S&P 500 Index - Estimated Fractal Dimension

5.50

5.00
Correlation Dim ension

4.50 y = -0.1078x 2 + 2.08x - 4.8576

R2 = 0.9684
4.00

3.50

3.00

2.50

2.00
5 6 7 8 9 10
Em bedding Dim ension

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 34
 Solution: Estimating the Correlation
Dimension for the S&P 500 Index
5 6 7 8 9 10
DEST 2.86 3.82 4.15 5.12 5.10 5.14
SE 0.029 0.043 0.017 0.058 0.029 0.045
t 98.45 88.13 247.26 88.19 174.43 114.02
p 0.000 0.000 0.000 0.000 0.000 0.000
R2 99.90% 99.87% 99.98% 99.87% 99.97% 99.92%

ƒ Fractal dimension estimate Stablizes around 5.17

ƒ Concurs with LeBaron & Scheinkman (1986)
• Daily stock returns had fractal dimension between 5 and 6
ƒ Interpretation
• 5 or 6 dynamic variables determine S&P index process
• Extremely complex system, impossible to estimate

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 35
 Other Studies of Fractal Dimension
¾ Peters (1991)
ƒ Criticized LeBaron Study
• Data insufficiency - would require 106 data points to estimate fractal
dimension reliably
• Use of returns not appropriate for study of non-linear effects
ƒ Used inflation-adjusted prices over 40 year period
¾ Findings
ƒ Equity Index Est. Dimension
US (S&P500) 2.33
Japan 3.05
Germany 2.41
UK 2.94

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 36
 Lyapunov Exponents
¾ Measure of sensitivity to initial conditions
ƒ How rapidly nearby points in phase space diverge
(+ve) or converge (-ve)
ƒ One exponent for each dimension of phase space
• Linear dimension grows at rate 2L1t
• Area grows at rate 2(L1 + L2 )t etc.
¾ Equation Lyapunov exponent for ith dim. pi(t)
⎡1 ⎛ pi (t ) ⎞⎤
Li = Lim ⎢ Log 2 ⎜⎜ ⎟⎟⎥
t →∞ ⎣ t ⎝ pi (0) ⎠⎦
Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 37
 Lyapunov Exponents and Attractors
¾ Point attractors
ƒ 3 negative exponents
¾ Limit cycles
ƒ 2 negative, one zero exponent
• 2 dimensions that converge
¾ 3-D strange attractors
ƒ One positive, one zero, one negative
• Positive exponent shows sensitivity to initial conditions
• Negative exponent causes diverging point to remain in
range of attractor

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 38
 Lyapunov Exponents and the
Capital Markets
¾ Strange attractor?
ƒ Positive exponent due to technical factors or
sentiment
ƒ Negative exponent due to fundamental value
• Brings prices back into “reasonable” range

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 39
 Lyapunov Exponents and Time Series
¾ Find largest positive Lyapunov exponent L+
ƒ Measured in bits per day
ƒ Means we lose L+ bits of predictive power / day
¾ Example: L+ = 0.1
ƒ We lose 0.1 bits of predictive power / day
ƒ Suppose we can measure today’s conditions to 1 bit
precision
ƒ Information will lose all value after 1 / 0.1 = 10 days

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 40
 Estimating the Largest
Lyapunov Exponent
¾ Wolf’s algorithm
ƒ Measures divergence of nearby points in
reconstructed phase space
ƒ Indicates how rate of divergence scales over fixed
intervals of time
ƒ Should converge to L+ if appropriate embedding
dimension m and time lag t are chosen

1 m ⎛ L' (t j +1 ) ⎞
L = ∑ Log 2 ⎜
+ ⎟
t j =1 ⎜ L(t ) ⎟
⎝ j ⎠
Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 41
 Largest Lyapunov Exponents of
International Equity Markets
Equity Lyapunov Indicated
Market (bit / month) Cycle (months)

S&P500 0.0241 42
UK 0.0280 36
Japan 0.0228 44
Germany 0.0168 60

Source: Peters (1991)

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 42
 Conclusions
¾ Long memory process
ƒ Confirmed by two independent methods of analysis
• R/S and Lyapunov
¾ Equity and bond markets - nonlinear systems
ƒ Aperiodic cycles
• E.g. Average cycle length 42 months in S&P 500 index
ƒ Strange attractors
• Fractal attractor dimension 2.33 (5.17)
ƒ Fractional noise short term (technical factors?)
ƒ Chaotic long term (fundamental analysis?)
¾ Currency markets have no cycle - black noise

Copyright © 1999 – 2006 Investment Analytics Forecasting Financial Markets – Nonlinear Dynamics Slide: 43