Scribd
Upload
51 views

Financial Time Series Models

This document discusses models for measuring and forecasting volatility in financial time series data. It introduces the concepts of volatility clustering, where periods of high and low price volatility tend to group together. The Autoregressive Conditional Heteroscedasticity (ARCH) and Generalized ARCH (GARCH) models are presented as statistical methods to capture volatility clustering. The ARCH model specifies the conditional variance as a function of lagged squared errors. The GARCH model generalizes this by including lags of both the variance and squared errors. An example analysis using these models on stock market return data is described to illustrate capturing time-varying volatility.

Uploaded by

Abdi Hiir
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
51 views

Financial Time Series Models

This document discusses models for measuring and forecasting volatility in financial time series data. It introduces the concepts of volatility clustering, where periods of high and low price volatility tend to group together. The Autoregressive Conditional Heteroscedasticity (ARCH) and Generalized ARCH (GARCH) models are presented as statistical methods to capture volatility clustering. The ARCH model specifies the conditional variance as a function of lagged squared errors. The GARCH model generalizes this by including lags of both the variance and squared errors. An example analysis using these models on stock market return data is described to illustrate capturing time-varying volatility.

Uploaded by

Abdi Hiir
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 11

Financial Time Series

Models
Lecture 16
Basic Econometrics-II
Instructor: Shahid Akbar
Measuring Volatility
 In Finance, volatility is the degree of variation of a trading price series
over time.
 It is usually measured by the standard deviation of logarithmic
returns.
 Volatility Clustering: Financial time series, such as stock prices,
exchange rates, inflation rates, etc., often exhibit the phenomenon of
volatility clustering, that is, periods in which their prices show wide
swings for an extended time period followed by periods in which there
is relative calm.
Illustrative Example
 Figure 22.6 gives logs of the monthly U.S./U.K. exchange rate (dollars
per pound), Y*t=logYt or the period 1971–2007, for a total of 444
monthly observations.

 As you can see from this figure, there are considerable ups and downs
in the exchange rate over the sample period.
Continue ……….
 As you can observe, the relative changes (dY*t = Y*t− Y*t−1) in the
U.S./U.K. exchange rate show periods of wide swings for some time
periods and periods of rather moderate swings in other time periods,
thus exemplifying the phenomenon of volatility clustering.

 Now the Question is how to model/ statistically measure such Volatility


Clustering?
Autoregressive Conditional ,

Heteroscedasticity (ARCH) Model


 In order to model Volatility Clustering (Engle, 1982) introduced
ARCH Model defined as
 Yt = X t  + t , t N ( 0 ,  t2 )

 Then  t2   0   1  t2 1 ARCH (1) Model

 Generally,
 t
2
  0   1  t2 1   2  t2 2  . . . . . . . . .   p  t2 p
p
 t
2
  0  
i 1
 i  t2 i
 Which is called ARCH(p) model.
Generalized Autoregressive Conditional
Heteroscedasticity (GARCH) Model
 But GARCH is much more flexible, much more capable of matching a
wide variety of patterns of financial volatility
 ARCH had volatility depending on lagged errors squared only
 However, GARCH adds to this lags of volatility itself, i.e.

 t
2
= 0 +  1 GARCH(1,1)
------- t-1 +  2 t  1
2 2

 In general,
t2 = 0 +1t-1
2
+ 2 t-2
2
+.......+ p t-p
2
+1t21  2t22  ......  qt2q

 Is a GARCH(p,q) model.
Procedure: ARCH Model
 let us consider the k-variable linear regression model:

 The variance of ut at time t is dependent on the squared disturbance at


time (t − 1), thus giving the appearance of serial correlation.
 Of course, the error variance may depend not only on one lagged term
of the squared error term but also on several lagged squared terms as
follows,

 Under H0:α1=α2=···=αp=0, If there is no autocorrelation in the error


variance, we have var(ut) = α0, and we do not have the ARCH effect.
Continue …….
 Since σ2t do not directly observe, Engle has shown that running the
following regression can easily test the preceding null hypothesis:

 Hence under null hypothesis (No ARCH effect) one can use the usual
F-test or alternatively the following test-statistic,
Illustrative Example
 Figure 22.8 presents monthly percentage change in the NYSE (New
York Stock Exchange) Index for the period 1966–2002.

 It is evident from this graph that the percent price changes in the
NYSE Index exhibit considerable volatility. Especially the wide swing
around the 1987 crash in stock prices.
Continue ………
 To capture the volatility in the stock return seen in the figure, let us consider a
very simple model

where Yt=percent change in the NYSE stock index and u t=random error term.
From the data, we obtained the following OLS regression:
Yˆ t = 0.00574
t = (3.36)
d = 1.4915
Now obtain the residual from this regression and estimate ARCH (1) model
as,

 Since the lagged squared disturbance term is statistically significant (p


value of about 0.000), it seems the error variances are correlated; that
is, there is an ARCH effect.
Thank You

You might also like