Univariate Time Series Modelling and Forecasting PDF

Chapter 6 Univariate time series modelling and forecasting ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 1  Univariate Time Series Models Where we attempt to predict returns using only information contained in thei...

Chapter 6 Univariate time series modelling and forecasting ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 1  Univariate Time Series Models Where we attempt to predict returns using only information contained in their past values. Some Notation and Concepts A Strictly Stationary Process A strictly stationary process is one where P{yt1  b1,..., ytn  bn} = P{yt1 + m  b1,..., ytn + m  bn} i.e. the probability measure for the sequence {yt} is the same as that for {yt+m}  m. A Weakly Stationary Process If a series satisfies the next three equations, it is said to be weakly or covariance stationary 1. E(yt) =  , t = 1,2,..., 2. E ( yt −  )( yt −  ) =    2 3. E ( yt1 − )( yt 2 − ) =  t 2 − t1  t1 , t2 ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 2  Univariate Time Series Models (cont’d) So if the process is covariance stationary, all the variances are the same and all the covariances depend on the difference between t1 and t2. The moments E ( yt − E ( yt ))( yt + s − E ( yt + s )) =  s , s = 0,1,2,... are known as the covariance function. The covariances, s, are known as autocovariances. However, the value of the autocovariances depend on the units of measurement of yt. It is thus more convenient to use the autocorrelations which are the autocovariances normalised by dividing by the variance:   = s , s = 0,1,2,... 0 s If we plot s against s=0,1,2,... then we obtain the autocorrelation function or correlogram. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 3  A White Noise Process A white noise process is one with (virtually) no discernible structure. A definition of a white noise process is E ( y ) =  t Var ( yt ) =  2  2 if t = r  t −r =  0 otherwise Thus the autocorrelation function will be zero apart from a single peak of 1 at s = 0. s  approximately N(0,1/T) where T = sample size We can use this to do significance tests for the autocorrelation coefficients by constructing a confidence interval. 1 For example, a 95% confidence interval would be given by .196 . If T the sample autocorrelation coefficient, s , falls outside this region for any value of s, then we reject the null hypothesis that the true value of the coefficient at lag s is zero. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 4  Joint Hypothesis Tests We can also test the joint hypothesis that all m of the k correlation coefficients are simultaneously equal to zero using the Q-statistic developed by Box and m Q = T  k2 Pierce: k =1 where T = sample size, m = maximum lag length The Q-statistic is asymptotically distributed as a  m2. However, the Box Pierce test has poor small sample properties, so a variant has been developed, called the Ljung-Box statistic: m  k2 Q = T (T + 2)  ~  m2 k =1 T −k This statistic is very useful as a portmanteau (general) test of linear dependence in time series. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 5  An ACF Example Question: Suppose that a researcher had estimated the first 5 autocorrelation coefficients using a series of length 100 observations, and found them to be (from 1 to 5): 0.207, -0.013, 0.086, 0.005, -0.022. Test each of the individual coefficient for significance, and use both the Box- Pierce and Ljung-Box tests to establish whether they are jointly significant. Solution: A coefficient would be significant if it lies outside (-0.196,+0.196) at the 5% level, so only the first autocorrelation coefficient is significant. Q=5.09 and Q*=5.26 Compared with a tabulated 2(5)=11.1 at the 5% level, so the 5 coefficients are jointly insignificant. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 6  Moving Average Processes Let ut (t=1,2,3,...) be a sequence of independently and identically distributed (iid) random variables with E(ut)=0 and Var(ut)=2 , then yt =  + ut + 1ut-1 + 2ut-2 +... + qut-q is a qth order moving average model MA(q). Its properties are E(yt)=; Var(yt) = 0 = (1+12 +  22 +...+ q2 )2 Covariances  ( +   s +1 1 +   s+2 2 +... +   q q−s ) 2 for s = 1,2,..., q s = s  0 for s  q ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 7  Example of an MA Problem 1. Consider the following MA(2) process: X t = u t +  1u t −1 +  2 u t − 2 where ut is a zero mean white noise process with variance  2. (i) Calculate the mean and variance of Xt (ii) Derive the autocorrelation function for this process (i.e. express the autocorrelations, 1, 2,... as functions of the parameters 1 and 2). (iii) If 1 = -0.5 and 2 = 0.25, sketch the acf of Xt. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 8  Solution (i) If E(ut)=0, then E(ut-i)=0  i. So E(Xt) = E(ut + 1ut-1+ 2ut-2)= E(ut)+ 1E(ut-1)+ 2E(ut-2)=0 Var(Xt) = E[Xt-E(Xt)][Xt-E(Xt)] but E(Xt) = 0, so Var(Xt) = E[(Xt)(Xt)] = E[(ut + 1ut-1+ 2ut-2)(ut + 1ut-1+ 2ut-2)] = E[ u t2 +  12 u t2−1 +  22 u t2− 2 +cross-products] But E[cross-products]=0 since Cov(ut,ut-s)=0 for s0. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 9  Solution (cont’d) So Var(Xt) = 0= E [ u t +  1 u t −1 +  2 u t − 2 ] 2 2 2 2 2 =  +1  + 2  2 2 2 2 2 = (1 +  1 +  2 ) 2 2 2 (ii) The acf of Xt. 1 = E[Xt-E(Xt)][Xt-1-E(Xt-1)] = E[Xt][Xt-1] = E[(ut +1ut-1+ 2ut-2)(ut-1 + 1ut-2+ 2ut-3)] = E[(  1u t −1 +  1 2 u t − 2 )] 2 2 =  1  2 +  1 2  2 = ( 1 +  1 2 ) 2 ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 10  Solution (cont’d) 2 = E[Xt-E(Xt)][Xt-2-E(Xt-2)] = E[Xt][Xt-2] = E[(ut + 1ut-1+2ut-2)(ut-2 +1ut-3+2ut-4)] = E[(  2 u t2− 2 )] =  2 2 3 = E[Xt-E(Xt)][Xt-3-E(Xt-3)] = E[Xt][Xt-3] = E[(ut +1ut-1+2ut-2)(ut-3 +1ut-4+2ut-5)] =0 So s = 0 for s > 2. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 11  Solution (cont’d) We have the autocovariances, now calculate the autocorrelations:  0 =  0 = 1 0 1 ( 1 +  1 2 ) 2 ( 1 +  1 2 ) 1 = = =  0 (1 +  12 +  22 ) 2 (1 +  12 +  22 ) 2 ( 2 ) 2 2 2 = = =  0 (1 +  12 +  22 ) 2 (1 +  12 +  22 )  3 =  3 = 0 0  s =  s = 0s  2 0 (iii) For 1 = -0.5 and 2 = 0.25, substituting these into the formulae above gives 1 = -0.476, 2 = 0.190. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 12  ACF Plot Thus the acf plot will appear as follows: 1.2 1 0.8 0.6 0.4 acf 0.2 0 0 1 2 3 4 5 6 -0.2 -0.4 -0.6 s ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 13  Autoregressive Processes An autoregressive model of order p, an AR(p) can be expressed as y t =  +  1 y t −1 +  2 y t − 2 +... +  p y t − p + u t Or using the lag operator notation: Lyt = yt-1 Liyt = yt-i p y t =  +   i y t −i + u t i =1 p or y t =  +   i L y t + u t i i =1 or  ( L) y t =  + u t where  ( L) = 1 − (1 L + 2 L2 +... p Lp ). ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 14  The Stationary Condition for an AR Model The condition for stationarity of a general AR(p) model is that the roots of 1 − 1z −  2 z2 −...− p z p = 0 all lie outside the unit circle. A stationary AR(p) model is required for it to have an MA() representation. Example 1: Is yt = yt-1 + ut stationary? The characteristic root is 1, so it is a unit root process (so non- stationary) Example 2: Is yt = 3yt-1 – 2.75yt-2 + 0.75yt-3 +ut stationary? The characteristic roots are 1, 2/3, and 2. Since only one of these lies outside the unit circle, the process is non-stationary. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 15  Wold’s Decomposition Theorem States that any stationary series can be decomposed into the sum of two unrelated processes, a purely deterministic part and a purely stochastic part, which will be an MA(). For the AR(p) model,  ( L) y t = u t , ignoring the intercept, the Wold decomposition is y t =  ( L)u t where,  ( L) = (1 − 1 L − 2 L2 −...− p Lp ) −1 ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 16  The Moments of an Autoregressive Process The moments of an autoregressive process are as follows. The mean is given by 0 E ( yt ) = 1 − 1 − 2 −... −  p The autocovariances and autocorrelation functions can be obtained by solving what are known as the Yule-Walker equations: 1 = 1 + 12 +... +  p −1 p  2 = 11 + 2 +... +  p − 2 p     p =  p −11 +  p − 22 +... +  p If the AR model is stationary, the autocorrelation function will decay exponentially to zero. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 17  Sample AR Problem Consider the following simple AR(1) model yt =  + 1 yt −1 + ut (i) Calculate the (unconditional) mean of yt. For the remainder of the question, set =0 for simplicity. (ii) Calculate the (unconditional) variance of yt. (iii) Derive the autocorrelation function for yt. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 18  Solution (i) Unconditional mean: E(yt) = E(+1yt-1) = +1E(yt-1) But also So E(yt)=  +1 ( +1E(yt-2)) =  +1  +12 E(yt-2)) E(yt) =  +1  +12 E(yt-2)) =  +1  +12 ( +1E(yt-3)) =  +1  +12  +13 E(yt-3) ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 19  Solution (cont’d) An infinite number of such substitutions would give E(yt) =  (1+1+12 +...) + 1y0 So long as the model is stationary, i.e. 1 < 1, then 1 = 0. So E(yt) =  (1+1+12 +...) =  1 − 1 (ii) Calculating the variance of yt: yt = 1 yt −1 + ut From Wold’s decomposition theorem: y t (1 − 1 L) = u t y t = (1 − 1 L) −1 u t y t = (1 + 1 L + 1 L2 +...)u t 2 ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 20  Solution (cont’d) So long as 1  1 , this will converge. yt = ut + 1ut −1 + 1 ut −2 +... 2 Var(yt) = E[yt-E(yt)][yt-E(yt)] but E(yt) = 0, since we are setting  = 0. Var(yt) = E[(yt)(yt)] ( )( ) = E[ ut + 1ut −1 + 12ut −2 +.. ut + 1ut −1 + 12ut −2 +.. ] = E[(ut 2 + 12ut −12 + 14ut −2 2 +... + cross − products)] = E[(ut + 1 ut −1 + 1 ut −2 +...)] 2 2 2 4 2 =  u2 + 12 u2 + 14 u2 +... =  u (1 + 1 + 1 +...) 2 2 4 =  2 u (1 − 12 ) ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 21  Solution (cont’d) (iii) Turning now to calculating the acf, first calculate the autocovariances: 1 = Cov(yt, yt-1) = E[yt-E(yt)][yt-1-E(yt-1)] Since a0 has been set to zero, E(yt) = 0 and E(yt-1) = 0, so 1 = E[ytyt-1] 1 = E[(ut + 1ut −1 + 1 ut −2 +...)(u t −1 + 1u t − 2 + 1 u t −3 +...) ] 2 2  u = E[ 1 t −1 2 +  1 3 u t − 2 +... + cross − products] 2 = 1 + 1  + 1  +... 2 3 2 5 2 1 2 = (1 − 12 ) ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 22  Solution (cont’d) For the second autocorrelation coefficient, 2 = Cov(yt, yt-2) = E[yt-E(yt)][yt-2-E(yt-2)] Using the same rules as applied above for the lag 1 covariance 2 = E[ytyt-2] = E[(ut + 1ut −1 + 1 ut − 2 +...)(ut −2 + 1ut −3 + 1 ut −4 +...) ] 2 2 = E[ 1 2 u t − 2 2 + 1 4 u t −3 2 +... + cross − products] = 12 2 + 14 2 +... = 1  (1 + 1 + 1 +...) 2 2 2 4 12 2 = (1 − 12 ) ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 23  Solution (cont’d) If these steps were repeated for 3, the following expression would be obtained 13 2 3 = (1 − 12 ) and for any lag s, the autocovariance would be given by 1s  2 s = (1 − 12 ) The acf can now be obtained by dividing the covariances by the variance: ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 24  Solution (cont’d) 0 0 = =1 0    2 2    1 2   1    2   2   (1 −  1 )  (1 −  1 ) 1   2   1 =  = = 1 2 = = =  12   0   0  2   2   2   2   (1 −  1 )  (1 −  1 )     3 = 1 3 … s = 1s ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 25  The Partial Autocorrelation Function (denoted kk) Measures the correlation between an observation k periods ago and the current observation, after controlling for observations at intermediate lags (i.e. all lags < k). So kk measures the correlation between yt and yt-k after removing the effects of yt-k+1 , yt-k+2 , …, yt-1. At lag 1, the acf = pacf always At lag 2, 22 = (2-12) / (1-12) For lags 3+, the formulae are more complex. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 26  The Partial Autocorrelation Function (denoted kk) (cont’d) The pacf is useful for telling the difference between an AR process and an ARMA process. In the case of an AR(p), there are direct connections between yt and yt-s only for s p. So for an AR(p), the theoretical pacf will be zero after lag p. In the case of an MA(q), this can be written as an AR(), so there are direct connections between yt and all its previous values. For an MA(q), the theoretical pacf will be geometrically declining. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 27  ARMA Processes By combining the AR(p) and MA(q) models, we can obtain an ARMA(p,q) model:  ( L) yt =  +  ( L)u t where  ( L) = 1 − 1 L − 2 L2 −...− p Lp and  ( L) = 1 + 1L +  2 L2 +... +  q Lq or y t =  + 1 y t −1 +  2 y t − 2 +... +  p y t − p +  1u t −1 +  2 u t − 2 +... +  q u t − q + u t with E (ut ) = 0; E (ut ) =  ; E (ut u s ) = 0, t  s 2 2 ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 28  The Invertibility Condition Similar to the stationarity condition, we typically require the MA(q) part of the model to have roots of (z)=0 greater than one in absolute value. The mean of an ARMA series is given by  E ( yt ) = 1 − 1 − 2 −...−p The autocorrelation function for an ARMA process will display combinations of behaviour derived from the AR and MA parts, but for lags beyond q, the acf will simply be identical to the individual AR(p) model. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 29  Summary of the Behaviour of the acf for AR and MA Processes An autoregressive process has a geometrically decaying acf number of spikes of pacf = AR order A moving average process has Number of spikes of acf = MA order a geometrically decaying pacf ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 30  Some sample acf and pacf plots for standard processes The acf and pacf are not produced analytically from the relevant formulae for a model of that type, but rather are estimated using 100,000 simulated observations with disturbances drawn from a normal distribution. ACF and PACF for an MA(1) Model: yt = – 0.5ut-1 + ut 0.05 0 1 2 3 4 5 6 7 8 9 10 -0.05 -0.1 -0.15 acf and pacf -0.2 -0.25 -0.3 acf -0.35 pacf -0.4 -0.45 Lag ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 31  ACF and PACF for an MA(2) Model: yt = 0.5ut-1 - 0.25ut-2 + ut 0.4 0.3 acf pacf 0.2 0.1 acf and pacf 0 1 2 3 4 5 6 7 8 9 10 -0.1 -0.2 -0.3 -0.4 Lags ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 32  ACF and PACF for a slowly decaying AR(1) Model: yt = 0.9yt-1 + ut 1 0.9 acf 0.8 pacf 0.7 0.6 acf and pacf 0.5 0.4 0.3 0.2 0.1 0 1 2 3 4 5 6 7 8 9 10 -0.1 Lags ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 33  ACF and PACF for a more rapidly decaying AR(1) Model: yt = 0.5yt-1 + ut 0.6 0.5 acf pacf 0.4 acf and pacf 0.3 0.2 0.1 0 1 2 3 4 5 6 7 8 9 10 -0.1 Lags ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 34  ACF and PACF for a more rapidly decaying AR(1) Model with Negative Coefficient: yt = -0.5yt-1 + ut 0.3 0.2 0.1 0 1 2 3 4 5 6 7 8 9 10 acf and pacf -0.1 -0.2 -0.3 -0.4 acf -0.5 pacf -0.6 Lags ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 35  ACF and PACF for a Non-stationary Model (i.e. a unit coefficient): yt = yt-1 + ut 1 0.9 acf pacf 0.8 0.7 acf and pacf 0.6 0.5 0.4 0.3 0.2 0.1 0 1 2 3 4 5 6 7 8 9 10 Lags ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 36  ACF and PACF for an ARMA(1,1): yt = 0.5yt-1 + 0.5ut-1 + ut 0.8 0.6 acf pacf 0.4 acf and pacf 0.2 0 1 2 3 4 5 6 7 8 9 10 -0.2 -0.4 Lags ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 37  Building ARMA Models - The Box Jenkins Approach Box and Jenkins (1970) were the first to approach the task of estimating an ARMA model in a systematic manner. There are 3 steps to their approach: 1. Identification 2. Estimation 3. Model diagnostic checking Step 1: - Involves determining the order of the model. - Use of graphical procedures - A better procedure is now available ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 38  Building ARMA Models - The Box Jenkins Approach (cont’d) Step 2: - Estimation of the parameters - Can be done using least squares or maximum likelihood depending on the model. Step 3: - Model checking Box and Jenkins suggest 2 methods: - deliberate overfitting - residual diagnostics ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 39  Some More Recent Developments in ARMA Modelling Identification would typically not be done using acf’s. We want to form a parsimonious model. Reasons: - variance of estimators is inversely proportional to the number of degrees of freedom. - models which are profligate might be inclined to fit to data specific features This gives motivation for using information criteria, which embody 2 factors - a term which is a function of the RSS - some penalty for adding extra parameters The object is to choose the number of parameters which minimises the information criterion. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 40  Information Criteria for Model Selection The information criteria vary according to how stiff the penalty term is. The three most popular criteria are Akaike’s (1974) information criterion (AIC), Schwarz’s (1978) Bayesian information criterion (SBIC), and the Hannan-Quinn criterion (HQIC). AIC = ln( 2 ) + 2k / T k SBIC = ln( ˆ 2 ) + ln T T 2k HQIC = ln( ˆ 2 ) + ln(ln( T )) T where k = p + q + 1, T = sample size. So we min. IC s.t. p  p, q  q SBIC embodies a stiffer penalty term than AIC. Which IC should be preferred if they suggest different model orders? – SBIC is strongly consistent but (inefficient). – AIC is not consistent, and will typically pick “bigger” models. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 41  ARIMA Models As distinct from ARMA models. The I stands for integrated. An integrated autoregressive process is one with a characteristic root on the unit circle. Typically researchers difference the variable as necessary and then build an ARMA model on those differenced variables. An ARMA(p,q) model in the variable differenced d times is equivalent to an ARIMA(p,d,q) model on the original data. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 42  Exponential Smoothing Another modelling and forecasting technique How much weight do we attach to previous observations? Expect recent observations to have the most power in helping to forecast future values of a series. The equation for the model St =  yt + (1-)St-1 (1) where  is the smoothing constant, with 01 yt is the current realised value St is the current smoothed value ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 43  Exponential Smoothing (cont’d) Lagging (1) by one period we can write St-1 =  yt-1 + (1-)St-2 (2) and lagging again St-2 =  yt-2 + (1-)St-3 (3) Substituting into (1) for St-1 from (2) St =  yt + (1-)( yt-1 + (1-)St-2) =  yt + (1-) yt-1 + (1-)2 St-2 (4) Substituting into (4) for St-2 from (3) St =  yt + (1-) yt-1 + (1-)2 St-2 =  yt + (1-) yt-1 + (1-)2( yt-2 + (1-)St-3) =  yt + (1-) yt-1 + (1-)2 yt-2 + (1-)3 St-3 ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 44  Exponential Smoothing (cont’d) T successive substitutions of this kind would lead to  T  St =   (1 −  ) yt −i  + (1 −  ) S0 i T  i =0  since 0, the effect of each observation declines exponentially as we move another observation forward in time. Forecasts are generated by ft+s = St for all steps into the future s = 1, 2,... This technique is called single (or simple) exponential smoothing. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 45  Exponential Smoothing (cont’d) It doesn’t work well for financial data because – there is little structure to smooth – it cannot allow for seasonality – it is an ARIMA(0,1,1) with MA coefficient (1-) - (See Granger & Newbold, p174) – forecasts do not converge on long term mean as s→ Can modify single exponential smoothing – to allow for trends (Holt’s method) – or to allow for seasonality (Winter’s method). Advantages of Exponential Smoothing – Very simple to use – Easy to update the model if a new realisation becomes available. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 46  Forecasting in Econometrics Forecasting = prediction. An important test of the adequacy of a model. e.g. - Forecasting tomorrow’s return on a particular share - Forecasting the price of a house given its characteristics - Forecasting the riskiness of a portfolio over the next year - Forecasting the volatility of bond returns We can distinguish two approaches: - Econometric (structural) forecasting - Time series forecasting The distinction between the two types is somewhat blurred (e.g, VARs). ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 47  In-Sample Versus Out-of-Sample Expect the “forecast” of the model to be good in-sample. Say we have some data - e.g. monthly FTSE returns for 120 months: 1990M1 – 1999M12. We could use all of it to build the model, or keep some observations back: A good test of the model since we have not used the information from 1999M1 onwards when we estimated the model parameters. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 48  How to produce forecasts Multi-step ahead versus single-step ahead forecasts Recursive versus rolling windows To understand how to construct forecasts, we need the idea of conditional expectations: E(yt+1  t ) We cannot forecast a white noise process: E(ut+s  t ) = 0  s > 0. The two simplest forecasting “methods” 1. Assume no change : f(yt+s) = yt 2. Forecasts are the long term average f(yt+s) = y ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 49  Models for Forecasting Structural models e.g. y = X + u yt = 1 +  2 x2t +  +  k xkt + ut To forecast y, we require the conditional expectation of its future E ( yt t −1 ) = E (1 +  2 x2t +  +  k xkt + ut ) value: = 1 +  2 E (x2t ) +  +  k E (xkt ) But what are ( x 2t ) etc.? We could use x2 , so E ( yt ) = 1 +  2 x2 +  +  k xk = y !! ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 50  Models for Forecasting (cont’d) Time Series Models The current value of a series, yt, is modelled as a function only of its previous values and the current value of an error term (and possibly previous values of the error term). Models include: simple unweighted averages exponentially weighted averages ARIMA models Non-linear models – e.g. threshold models, GARCH, bilinear models, etc. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 51  Forecasting with ARMA Models The forecasting model typically used is of the form: p q f t , s =  + i f t ,s −i +  j ut + s − j i =1 j =1 where ft,s = yt+s , s 0; ut+s = 0, s > 0 = ut+s , s  0 ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 52  Forecasting with MA Models An MA(q) only has memory of q. e.g. say we have estimated an MA(3) model: yt =  + 1ut-1 +  2ut-2 +  3ut-3 + ut yt+1 =  +  1ut +  2ut-1 +  3ut-2 + ut+1 yt+2 =  +  1ut+1 +  2ut +  3ut-1 + ut+2 yt+3 =  +  1ut+2 +  2ut+1 +  3ut + ut+3 We are at time t and we want to forecast 1,2,..., s steps ahead. We know yt , yt-1,..., and ut , ut-1 ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 53  Forecasting with MA Models (cont’d) ft, 1 = E(yt+1  t ) = E( +  1ut +  2ut-1 +  3ut-2 + ut+1) =  +  1ut +  2ut-1 +  3ut-2 ft, 2 = E(yt+2  t ) = E( +  1ut+1 +  2ut +  3ut-1 + ut+2) =  +  2ut +  3ut-1 ft, 3 = E(yt+3  t ) = E( +  1ut+2 +  2ut+1 +  3ut + ut+3) =  +  3ut ft, 4 = E(yt+4  t ) =  ft, s = E(yt+s  t ) =  s4 ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 54  Forecasting with AR Models Say we have estimated an AR(2) yt =  + 1yt-1 +  2yt-2 + ut yt+1 =  +  1yt +  2yt-1 + ut+1 yt+2 =  +  1yt+1 +  2yt + ut+2 yt+3 =  +  1yt+2 +  2yt+1 + ut+3 ft, 1 = E(yt+1  t ) = E( +  1yt +  2yt-1 + ut+1) =  +  1E(yt) +  2E(yt-1) =  +  1yt +  2yt-1 ft, 2 = E(yt+2  t ) = E( +  1yt+1 +  2yt + ut+2) =  +  1E(yt+1) +  2E(yt) =  +  1 ft, 1 +  2yt ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 55  Forecasting with AR Models (cont’d) ft, 3 = E(yt+3  t ) = E( +  1yt+2 +  2yt+1 + ut+3) =  +  1E(yt+2) +  2E(yt+1) =  +  1 ft, 2 +  2 ft, 1 We can see immediately that ft, 4 =  +  1 ft, 3 +  2 ft, 2 etc., so ft, s =  +  1 ft, s-1 +  2 ft, s-2 Can easily generate ARMA(p,q) forecasts in the same way. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 56  How can we test whether a forecast is accurate or not? For example, say we predict that tomorrow’s return on the FTSE will be 0.2, but the outcome is actually -0.4. Is this accurate? Define ft,s as the forecast made at time t for s steps ahead (i.e. the forecast made for time t+s), and yt+s as the realised value of y at time t+s. Some of the most popular criteria for assessing the accuracy of time series forecasting techniques are: N 1 MSE = N t =1 ( yt + s − f t , s ) 2 N 1 MAE is given by MAE = N t =1 yt + s − f t , s 1 N yt +s − ft ,s Mean absolute percentage error: MAPE= 100  N t =1 yt +s ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 57  How can we test whether a forecast is accurate or not? (cont’d) It has, however, also recently been shown (Gerlow et al., 1993) that the accuracy of forecasts according to traditional statistical criteria are not related to trading profitability. A measure more closely correlated with profitability: 1 N % correct sign predictions =  zt + s N t =1 where zt+s = 1 if (yt+s. ft,s ) > 0 zt+s = 0 otherwise ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 58  Forecast Evaluation Example Given the following forecast and actual values, calculate the MSE, MAE and percentage of correct sign predictions: Steps Ahead Forecast Actual 1 0.20 -0.40 2 0.15 0.20 3 0.10 0.10 4 0.06 -0.10 5 0.04 -0.05 MSE = 0.079, MAE = 0.180, % of correct sign predictions = 40 ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 59  What factors are likely to lead to a good forecasting model? “signal” versus “noise” “data mining” issues simple versus complex models financial or economic theory ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 60  Statistical Versus Economic or Financial loss functions Statistical evaluation metrics may not be appropriate. How well does the forecast perform in doing the job we wanted it for? Limits of forecasting: What can and cannot be forecast? All statistical forecasting models are essentially extrapolative Forecasting models are prone to break down around turning points Series subject to structural changes or regime shifts cannot be forecast Predictive accuracy usually declines with forecasting horizon Forecasting is not a substitute for judgement ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 61  Back to the original question: why forecast? Why not use “experts” to make judgemental forecasts? Judgemental forecasts bring a different set of problems: e.g., psychologists have found that expert judgements are prone to the following biases: – over-confidence – inconsistency – recency – anchoring – illusory patterns – “group-think”. The Usually Optimal Approach To use a statistical forecasting model built on solid theoretical foundations supplemented by expert judgements and interpretation. ‘Introductory Econometrics for Finance’ © Chris Brooks 2013 62