#generate facebook data
fb<-getSymbols("FB",auto.assign =FALSE,from = "2011-09-01", to = "2021-09-01") #as my train
## 'getSymbols' currently uses auto.assign=TRUE by default, but will
## use auto.assign=FALSE in 0.5-0. You will still be able to use
## 'loadSymbols' to automatically load data. getOption("getSymbols.env")
## and getOption("getSymbols.auto.assign") will still be checked for
## alternate defaults.
##
## This message is shown once per session and may be disabled by setting
## options("getSymbols.warning4.0"=FALSE). See ?getSymbols for details.
head(fb)
## FB.Open FB.High FB.Low FB.Close FB.Volume FB.Adjusted
## 2012-05-18 42.05 45.00 38.00 38.23 573576400 38.23
## 2012-05-21 36.53 36.66 33.00 34.03 168192700 34.03
## 2012-05-22 32.61 33.59 30.94 31.00 101786600 31.00
## 2012-05-23 31.37 32.50 31.36 32.00 73600000 32.00
## 2012-05-24 32.95 33.21 31.77 33.03 50237200 33.03
## 2012-05-25 32.90 32.95 31.11 31.91 37149800 31.91
plot(fb)
fbClose<-fb$FB.Close
plot(fbClose)
#Standard Garch Model,
#arma order=1,1: look one day back, use yesterday's stock price and error term to
#analyze today's price
fb1<-ugarchspec(variance.model=list(model="sGARCH",garchOrder=c(1,1)),
mean.model = list(armaOrder=c(1,1)),distribution.model = "std")
fbGarch1<-ugarchfit(spec=fb1,data=fbClose)
## Warning in .sgarchfit(spec = spec, data = data, out.sample = out.sample, :
## ugarchfit-->warning: solver failer to converge.
fbGarch1
##
## *---------------------------------*
## * GARCH Model Fit *
## *---------------------------------*
##
## Conditional Variance Dynamics
## -----------------------------------
## GARCH Model : sGARCH(1,1)
## Mean Model : ARFIMA(1,0,1)
## Distribution : std
##
## Convergence Problem:
## Solver Message:
#Information Criteria: Akaike 4.4656
#armaOrder=2,2: look two days back
fb2<-ugarchspec(variance.model=list(model="sGARCH",garchOrder=c(1,1)),
mean.model = list(armaOrder=c(2,2)),distribution.model = "std")
fbGarch2<-ugarchfit(spec=fb2,data=fbClose)
fbGarch2
##
## *---------------------------------*
## * GARCH Model Fit *
## *---------------------------------*
##
## Conditional Variance Dynamics
## -----------------------------------
## GARCH Model : sGARCH(1,1)
## Mean Model : ARFIMA(2,0,2)
## Distribution : std
##
## Optimal Parameters
## ------------------------------------
## Estimate Std. Error t value Pr(>|t|)
## mu 37.744650 0.339608 111.1420 0.000000
## ar1 0.030485 0.002284 13.3501 0.000000
## ar2 0.972865 0.002307 421.7466 0.000000
## ma1 0.926465 0.000153 6044.7362 0.000000
## ma2 -0.045453 0.000622 -73.0815 0.000000
## omega 0.010934 0.006102 1.7918 0.073164
## alpha1 0.054384 0.008584 6.3358 0.000000
## beta1 0.944616 0.009278 101.8134 0.000000
## shape 4.159885 0.312259 13.3219 0.000000
##
## Robust Standard Errors:
## Estimate Std. Error t value Pr(>|t|)
## mu 37.744650 0.092673 407.287 0.000000
## ar1 0.030485 0.001379 22.112 0.000000
## ar2 0.972865 0.001387 701.166 0.000000
## ma1 0.926465 0.000135 6883.389 0.000000
## ma2 -0.045453 0.001630 -27.884 0.000000
## omega 0.010934 0.009516 1.149 0.250570
## alpha1 0.054384 0.014750 3.687 0.000227
## beta1 0.944616 0.017323 54.531 0.000000
## shape 4.159885 0.359502 11.571 0.000000
##
## LogLikelihood : -5010.874
##
## Information Criteria
## ------------------------------------
##
## Akaike 4.2960
## Bayes 4.3182
## Shibata 4.2960
## Hannan-Quinn 4.3041
##
## Weighted Ljung-Box Test on Standardized Residuals
## ------------------------------------
## statistic p-value
## Lag[1] 0.8723 0.3503
## Lag[2*(p+q)+(p+q)-1][11] 5.6336 0.7213
## Lag[4*(p+q)+(p+q)-1][19] 9.5885 0.5379
## d.o.f=4
## H0 : No serial correlation
##
## Weighted Ljung-Box Test on Standardized Squared Residuals
## ------------------------------------
## statistic p-value
## Lag[1] 0.005205 0.9425
## Lag[2*(p+q)+(p+q)-1][5] 0.153267 0.9956
## Lag[4*(p+q)+(p+q)-1][9] 0.411361 0.9992
## d.o.f=2
##
## Weighted ARCH LM Tests
## ------------------------------------
## Statistic Shape Scale P-Value
## ARCH Lag[3] 0.1505 0.500 2.000 0.6981
## ARCH Lag[5] 0.2575 1.440 1.667 0.9513
## ARCH Lag[7] 0.3858 2.315 1.543 0.9875
##
## Nyblom stability test
## ------------------------------------
## Joint Statistic: 3.3744
## Individual Statistics:
## mu 0.0008058
## ar1 0.0063391
## ar2 0.0222204
## ma1 0.6716270
## ma2 0.5580412
## omega 0.7575060
## alpha1 0.3187442
## beta1 0.4689022
## shape 0.6610613
##
## Asymptotic Critical Values (10% 5% 1%)
## Joint Statistic: 2.1 2.32 2.82
## Individual Statistic: 0.35 0.47 0.75
##
## Sign Bias Test
## ------------------------------------
## t-value prob sig
## Sign Bias 0.45744 0.6474
## Negative Sign Bias 0.08927 0.9289
## Positive Sign Bias 0.46011 0.6455
## Joint Effect 0.81277 0.8464
##
##
## Adjusted Pearson Goodness-of-Fit Test:
## ------------------------------------
## group statistic p-value(g-1)
## 1 20 34.61 0.015542
## 2 30 49.97 0.009098
## 3 40 60.07 0.016662
## 4 50 66.66 0.047336
##
##
## Elapsed time : 1.050828
plot(fbGarch2,which='all')
##
## please wait...calculating quantiles...
#Information Criteria: Akaike 4.4547
#armaOrder=3,3: look three days back
fb3<-ugarchspec(variance.model=list(model="sGARCH",garchOrder=c(1,1)),
mean.model = list(armaOrder=c(3,3)),distribution.model = "std")
fbGarch3<-ugarchfit(spec=fb3,data=fbClose)
fbGarch3
##
## *---------------------------------*
## * GARCH Model Fit *
## *---------------------------------*
##
## Conditional Variance Dynamics
## -----------------------------------
## GARCH Model : sGARCH(1,1)
## Mean Model : ARFIMA(3,0,3)
## Distribution : std
##
## Optimal Parameters
## ------------------------------------
## Estimate Std. Error t value Pr(>|t|)
## mu 33.304800 0.972333 3.4252e+01 0.000000
## ar1 1.780682 0.000231 7.7116e+03 0.000000
## ar2 -0.795481 0.000141 -5.6416e+03 0.000000
## ar3 0.015211 0.000077 1.9817e+02 0.000000
## ma1 -0.826075 0.019480 -4.2406e+01 0.000000
## ma2 0.007276 0.024951 2.9161e-01 0.770585
## ma3 -0.000680 0.019170 -3.5466e-02 0.971708
## omega 0.011332 0.006200 1.8277e+00 0.067588
## alpha1 0.053956 0.008580 6.2883e+00 0.000000
## beta1 0.945044 0.009278 1.0186e+02 0.000000
## shape 4.121346 0.309715 1.3307e+01 0.000000
##
## Robust Standard Errors:
## Estimate Std. Error t value Pr(>|t|)
## mu 33.304800 0.481970 6.9101e+01 0.000000
## ar1 1.780682 0.000781 2.2811e+03 0.000000
## ar2 -0.795481 0.000339 -2.3483e+03 0.000000
## ar3 0.015211 0.000010 1.5766e+03 0.000000
## ma1 -0.826075 0.020850 -3.9620e+01 0.000000
## ma2 0.007276 0.026451 2.7507e-01 0.783264
## ma3 -0.000680 0.019259 -3.5303e-02 0.971838
## omega 0.011332 0.009551 1.1865e+00 0.235409
## alpha1 0.053956 0.014595 3.6969e+00 0.000218
## beta1 0.945044 0.017062 5.5388e+01 0.000000
## shape 4.121346 0.351802 1.1715e+01 0.000000
##
## LogLikelihood : -5003.884
##
## Information Criteria
## ------------------------------------
##
## Akaike 4.2917
## Bayes 4.3188
## Shibata 4.2917
## Hannan-Quinn 4.3016
##
## Weighted Ljung-Box Test on Standardized Residuals
## ------------------------------------
## statistic p-value
## Lag[1] 0.8888 0.3458
## Lag[2*(p+q)+(p+q)-1][17] 4.4502 1.0000
## Lag[4*(p+q)+(p+q)-1][29] 12.8892 0.7590
## d.o.f=6
## H0 : No serial correlation
##
## Weighted Ljung-Box Test on Standardized Squared Residuals
## ------------------------------------
## statistic p-value
## Lag[1] 0.002839 0.9575
## Lag[2*(p+q)+(p+q)-1][5] 0.132137 0.9967
## Lag[4*(p+q)+(p+q)-1][9] 0.388335 0.9993
## d.o.f=2
##
## Weighted ARCH LM Tests
## ------------------------------------
## Statistic Shape Scale P-Value
## ARCH Lag[3] 0.1263 0.500 2.000 0.7223
## ARCH Lag[5] 0.2521 1.440 1.667 0.9527
## ARCH Lag[7] 0.3761 2.315 1.543 0.9882
##
## Nyblom stability test
## ------------------------------------
## Joint Statistic: 3.3081
## Individual Statistics:
## mu 0.002986
## ar1 0.012905
## ar2 0.013183
## ar3 0.013375
## ma1 0.524620
## ma2 0.408995
## ma3 0.488961
## omega 0.775343
## alpha1 0.292935
## beta1 0.431370
## shape 0.562630
##
## Asymptotic Critical Values (10% 5% 1%)
## Joint Statistic: 2.49 2.75 3.27
## Individual Statistic: 0.35 0.47 0.75
##
## Sign Bias Test
## ------------------------------------
## t-value prob sig
## Sign Bias 0.5503 0.5821
## Negative Sign Bias 0.1615 0.8717
## Positive Sign Bias 0.4758 0.6342
## Joint Effect 1.0054 0.7999
##
##
## Adjusted Pearson Goodness-of-Fit Test:
## ------------------------------------
## group statistic p-value(g-1)
## 1 20 36.65 0.008774
## 2 30 45.55 0.025970
## 3 40 55.97 0.038353
## 4 50 62.12 0.098714
##
##
## Elapsed time : 1.048802
#Information Criteria: Akaike 4.4573
#armarOrder=2,2 is the optimal model
#predict the stock price in ten days
fbPredict<-ugarchboot(fbGarch2,n.ahead=10,method=c("Partial","Full")[1])
plot(fbPredict,which=2)
#the plot shows the upper and lower limits of the stock price estimation
#compare with actual value
fb1<-getSymbols("FB",auto.assign =FALSE,from = "2021-09-01", to = "2021-09-11") #as my train
fbClose1<-fb1$FB.Close
plot(fbClose1)
