Advanced Forecasing Techniques and Models: Time-Series Forecass Shor Examples Series using Risk Simulaor For more informaion please visi: www.realopionsvaluaion.com or conac us a: admin@realopionsvaluaion.com
Forecasing Time-Series Analysis File Name: Forecasing Time-Series Analysis Locaion: Modeling Toolki Forecasing Time Series Analysis Brief Descripion: This sample model illusraes how o run ime-series analysis forecass, which ake ino accoun hisorical base values, rends and seasonaliies o projec he fuure Requiremens: Modeling Toolki, Risk Simulaor The hisorical sales revenue daa are locaed in he Time-series Daa workshee in he model. The daa are quarerly sales revenue from Q1 2000 o Q4 2004. The daa exhibi quarerly seasonaliy, which means ha he seasonaliy is 4 (here are 4 quarers in 1 year or 1 cycle). Time-series forecasing decomposes he hisorical daa ino he baseline, rend, and seasonaliy, if any. The models hen apply an opimizaion procedure o find he alpha, bea, and gamma parameers for he baseline, rend, and seasonaliy coefficiens, and hen recompose hem ino a forecas. In oher words, his mehodology firs applies a backcas o find he bes-fiing model and bes-fiing parameers of he model ha minimizes forecas errors, and hen proceeds o forecas he fuure based on he hisorical daa ha exis. This of course assumes ha he same baseline growh, rend, and seasonaliy hold going forward. Even if hey do no say, when here exiss a srucural shif (e.g., company goes global, has a merger, spin-off, ec.) he baseline forecass can be compued and hen he required adjusmens can be made o he forecass. Procedure To run his model, simply: 1. Selec he hisorical daa (cells H11:H30). 2. Selec Risk Simulaor Forecasing Time-Series Analysis. 3. Selec Auo Model Selecion, Forecas 4 Periods and Seasonaliy 4 Periods (Figure 1). Noe ha you can selec Creae Simulaion Assumpions only if an exising Simulaion Profile exiss. If no, click on Simulaion New Simulaion Profile, and hen run he ime-series forecas per seps 1 o 3 above bu remember o check he Creae Simulaion Assumpions box. Model Resuls Analysis For your convenience, he analysis Repor and Mehodology workshees are included in he model. A fied char and forecas values are provided in he repor as well as he error measures and a saisical
summary of he mehodology (Figure 2). The Mehodology workshee provides he saisical resuls from all eigh ime-series mehodologies. Several differen ypes of errors can be calculaed for ime-series forecas mehods, including he meansquared error (MSE), roo mean-squared error (RMSE), mean absolue deviaion (MAD), and mean absolue percen error (MAPE). The MSE is an absolue error measure ha squares he errors (he difference beween he acual hisorical daa and he forecas-fied daa prediced by he model) o keep he posiive and negaive errors from canceling each oher ou. This measure also ends o exaggerae large errors by weighing he large errors more heavily han smaller errors by squaring hem, which can help when comparing differen ime-series models. The MSE is calculaed by simply aking he average of he Error 2. RMSE is he square roo of MSE and is he mos popular error measure, also known as he quadraic loss funcion. RMSE can be defined as he average of he absolue values of he forecas errors and is highly appropriae when he cos of he forecas errors is proporional o he absolue size of he forecas error. MAD is an error saisic ha averages he disance (absolue value of he difference beween he acual hisorical daa and he forecas-fied daa prediced by he model) beween each pair of acual and fied forecas daa poins. MAD is calculaed by aking he average of he Error, and is mos appropriae when he cos of forecas errors is proporional o he absolue size of he forecas errors. MAPE is a relaive error saisic measured as an average percen error of he hisorical daa poins and is mos appropriae when he cos of he forecas error is more closely relaed o he percenage error han he numerical size of he error. This error esimae is calculaed by aking he average of Y Yˆ, where Y Y is he hisorical daa a ime, while Ŷ is he fied or prediced daa poin a ime using his imeseries mehod. Finally, an associaed measure is he Theil s U saisic, which measures he naivey of he model s forecas. Tha is, if he Theil s U saisic is less han 1.0, hen he forecas mehod used provides an esimae ha is saisically beer han guessing. Figure 3 provides he mahemaical deails of each error esimae.
Figure 1: Running a ime-series analysis forecas Figure 2: Time-series analysis resuls
Figure 3: Error compuaions