Advanced

Svensk arbetslöshetsdata: Hjälper barometerdata att prognostisera Sveriges arbetslöshet

Månsson, Kristofer (2008)
Department of Statistics
Abstract
The purpose of this essay is to investigate whether data from the Buisiness Tendancy Survey (BTS) is useful for forecasting the Swedish unemployment rate. The BTS is made by the National Institute of Economic Research (NIER) who makes a large business survey each quarter based on questions to approximately 7000 thousand Swedish firms. The unemployment rate is also a survey based measure made by Statistics Sweden. In the survey they ask a random sample of the working population (16-64 years) whether they are currently working or if they are unemployed. All data used in the essay is collected from Ecowin for the time period 1980:1 through 2007:1. The data is split into two different parts. The first is the in-sample period (1980:1-2001:4)... (More)
The purpose of this essay is to investigate whether data from the Buisiness Tendancy Survey (BTS) is useful for forecasting the Swedish unemployment rate. The BTS is made by the National Institute of Economic Research (NIER) who makes a large business survey each quarter based on questions to approximately 7000 thousand Swedish firms. The unemployment rate is also a survey based measure made by Statistics Sweden. In the survey they ask a random sample of the working population (16-64 years) whether they are currently working or if they are unemployed. All data used in the essay is collected from Ecowin for the time period 1980:1 through 2007:1. The data is split into two different parts. The first is the in-sample period (1980:1-2001:4) which is used to estimate ARIMA- and TFM-models with the Box-Jenkins methodology. To evaluate the estimated models the t-test is used to see if the parameters are significant and the Ljung-Box-test is used to see if the residuals are white-noise. When the adequate models are found forecasts for the horizons t+1, t+4 and t+8 are made for the out-of-sample period (2002:1-2007:1). The forecasts are then evaluated by ME and MSE which are measurements of the size of the forecasting errors. Thereafter the results from the different models are compared to see if the TFM-models are better than the ARIMA-model and the Random Walk model. (Less)
Please use this url to cite or link to this publication:
author
Månsson, Kristofer
supervisor
organization
year
type
M2 - Bachelor Degree
subject
keywords
arbetslöshetsprognos, konjunturbarometern, transferfunktionsmodell, Statistics, operations research, programming, actuarial mathematics, Statistik, operationsanalys, programmering, aktuariematematik
language
Swedish
id
1848869
date added to LUP
2008-01-23
date last changed
2011-06-01 12:43:54
@misc{1848869,
  abstract     = {The purpose of this essay is to investigate whether data from the Buisiness Tendancy Survey (BTS) is useful for forecasting the Swedish unemployment rate. The BTS is made by the National Institute of Economic Research (NIER) who makes a large business survey each quarter based on questions to approximately 7000 thousand Swedish firms. The unemployment rate is also a survey based measure made by Statistics Sweden. In the survey they ask a random sample of the working population (16-64 years) whether they are currently working or if they are unemployed. All data used in the essay is collected from Ecowin for the time period 1980:1 through 2007:1. The data is split into two different parts. The first is the in-sample period (1980:1-2001:4) which is used to estimate ARIMA- and TFM-models with the Box-Jenkins methodology. To evaluate the estimated models the t-test is used to see if the parameters are significant and the Ljung-Box-test is used to see if the residuals are white-noise. When the adequate models are found forecasts for the horizons t+1, t+4 and t+8 are made for the out-of-sample period (2002:1-2007:1). The forecasts are then evaluated by ME and MSE which are measurements of the size of the forecasting errors. Thereafter the results from the different models are compared to see if the TFM-models are better than the ARIMA-model and the Random Walk model.},
  author       = {Månsson, Kristofer},
  keyword      = {arbetslöshetsprognos,konjunturbarometern,transferfunktionsmodell,Statistics, operations research, programming, actuarial mathematics,Statistik, operationsanalys, programmering, aktuariematematik},
  language     = {swe},
  note         = {Student Paper},
  title        = {Svensk arbetslöshetsdata: Hjälper barometerdata att prognostisera Sveriges arbetslöshet},
  year         = {2008},
}