Forecasting is the process of predicting the future based on the past data. Typically, forecasting uses the statistical technique employing different methods such as time series, moving average, linear regression and exponential smoothing. The study uses the 15-year dataset of IBM (International Business Machine) revenues from 1999 to 2016 fiscal years. The study collects large dataset because of the larger the dataset, the better the accuracy of the results.
The researcher collects revenue data of the IBM between 1999 and 2016 from the Statista (2016) website, and the dataset used for the analysis is as follow:
Revenue ($Billion)
Different methods are used for the forecast. The linear regression, exponential smoothing and moving average are used for analysis.
Linear Regression
The linear regression is the forecasting technique that assists in enhancing the relationship between dependent and independent variables. The benefits of the linear regression is that it assists in providing accurate results if large data are obtained. The study uses the data in Table 1 to produce revenue forecast using the Linear regression technique. The output is revealed in fig 1.
Fig 1: Linear Regression Forecasting
SUMMARY OUTPUT
Regression Statistics
Multiple R
0,24
R Square
0,06
Adjusted R Square
0,00
Standard Error
5,34
Observations
18,00
ANOVA
df
SS
MS
F
Significance F
Regression
1,00
28,93
28,93
1,02
0,33
Residual
16,00
455,57
28,47
Total
17,00
484,50
Coefficients
Standard Error
t Stat
P-value
Lower 95%
Upper 95%
Lower 95,0%
Upper 95,0%
Intercept
1992,97
14,47
137,69
0,00
1962,28
2023,65
1962,28
2023,65
X Variable 1
0,16
0,16
1,01
0,33
-0,17
0,49
-0,17
0,49
Exponential Smoothing
The exponential smoothing is another forecasting technique used and the result is presented in Fig 2 and Table 2.
Fig 2: Exponential Smoothing Forecasting
Table 2: Forecast using Exponential Smoothing
Year
Revenue ($ Billion)
Forecast with Exponential Smoothing
1999
87,55
2000
88,4
2001
83,07
87,635
2002
81,19
87,867
2003
89,13
82,882
2004
96,29
81,984
2005
91,13
89,846
2006
91,42
95,774
2007
98,79
91,159
2008
103,63
92,157
2009
95,76
99,274
2010
98,87
102,843
2011
106,92
96,071
2012
104,51
99,675
2013
99,75
106,679
2014
92,8
104,034
2015
81,7
99,055
2016
79,9
91,69
2017
81,52
Moving Average
Moving Average is a technique that assists in forecasting the future revenues of the IBM, and the major benefits of the moving average are its simplicity. (Lucey 2002).
The output of the moving average forecasting method is revealed in table 3.
Table 3: Forecasting Using Moving Average
Year
Revenue ($Billion)
Moving Average
1999
87,55
2000
88,4
2001
83,07
2002
81,19
86,34
2003
89,13
84,22
2004
96,29
84,46333
2005
91,13
88,87
2006
91,42
92,18333
2007
98,79
92,94667
2008
103,63
93,78
2009
95,76
97,94667
2010
98,87
99,39333
2011
106,92
99,42
2012
104,51
100,5167
2013
99,75
103,4333
2014
92,8
103,7267
2015
81,7
99,02
2016
79,9
91,41667
2017
84,8
Fig 3: Moving Average Forecasting
Compare and Contrast the Forecasting Method
Overview of the linear regression, exponential smoothing and moving average show that all the three models can assist in forecasting future revenue for the IBM. Moreover, all the three forecast methods have the ability to produce accurate results if a large number are available. Moreover, both the exponential smoothing and moving average have the ability to produce short-term forecasting.
The different between three methods is that while the moving average relies on the past data of all the years for forecasting, the exponential smoothing relies on the recent data to predict the accurate results. Moreover, the moving average and exponential smoothing are not appropriate for the long-term forecasting, they are only appropriate for the short-term forecasting. As being revealed in Table 2 and 3, it is revealed that moving and exponential smoothing can only produce a year forecast for the IBM revealing that IBM is likely to record a decline in the 2017 revenue. As being revealed in Table 2 and Table 3, the IBM will produce a decline in the revenue in the year 2017 fiscal year. However, the linear regression analysis is the most appropriate for the long-term forecasting for the IBM. Unlike the moving average and exponential smoothing that produce short-term forecasting, the linear regression is appropriate for long-term forecasting. (Hidalgo, & Goodman, 2013).
Forecast best for IBM
The linear regression is the most appropriate method for forecasting of IBM revenue because it can assist the firm to achieve a long-term forecasting, which the company will need for a long-term financial planning.
Impact this forecast have on IBM from a Financial Metrics Standpoint
The forecast shows that IBM is likely to record a decline in the revenue in the next few years revealing that the company will record a loss of revenue. Moreover, the company will record a loss in the net profits revealing that IBM may not be able to meet their obligations to their shareholders.
Reference
Lucey T. (2002). Quantitative Technique. UK. BookPower.
Hidalgo, B., & Goodman, M. (2013). Multivariate or Multivariable Regression? American Journal of Public Health, 103(1), 39-40. doi:10.2105/ajph.2012.300897
Statista (2016). Revenue of IBM worldwide from 1999 to 2015 (in billion U.S. dollars) Statista Inc.
You’re 100% through this paper. Sign up to read the full paper.
Sign Up Now — Instant Access Already a member? Log inAlways verify citation format against your institution’s current style guide requirements.