Forecasting Values are coming same in R

Question

I have one sample data

Sno     period       year_quarter   country            city        sales_revenue
1       1/1/2009        2009-Q1     Argentina       Buenos Aires        3008
2       1/4/2009        2009-Q2     Argentina       Buenos Aires        3244
3       1/7/2009        2009-Q3     Argentina       Buenos Aires        8000
4       1/10/2009       2009-Q4     Argentina       Buenos Aires        8719
5       1/1/2010        2010-Q1     Argentina       Buenos Aires        3008
6       1/4/2010        2010-Q2     Argentina       Buenos Aires        3244
7       1/7/2010        2010-Q3     Argentina       Buenos Aires        78
8       1/10/2010       2010-Q4     Argentina       Buenos Aires        7379
9       1/1/2011        2011-Q1     Argentina       Buenos Aires        3735
10      1/4/2011        2011-Q2     Argentina       Buenos Aires        7339
11      1/7/2011        2011-Q3     Argentina       Buenos Aires        17240
12      1/10/2011       2011-Q4     Argentina       Buenos Aires        20465
13      1/1/2012        2012-Q1     Argentina       Buenos Aires        13134
14      1/4/2012        2012-Q2     Argentina       Buenos Aires        15039

I forecasted the three quarter i.e 2012 q3, 2012 q4 and 2013 q1 with the help of the ETS(A,N,N).Code for the prediction is as below

retail_data.xts<-xts(retail_data$sales_revenue, retail_data$period);
retail_data.ts <- as.ts(retail_data.xts);
retail_data.ets <- ets(retail_data.ts,model="ANN");
retail_data.fore <- forecast(retail_data.ets, h=4);
plot(retail_data.fore);

the outcome of the calculation is

 Point Forecast    Lo 80    Hi 80     Lo 95    Hi 95
15       14905.37 8925.968 20884.78 5760.6608 24050.09
16       14905.37 7202.071 22608.68 3124.1881 26686.56
17       14905.37 5798.868 24011.88  978.1739 28832.58
18       14905.37 4584.713 25226.04 -878.7150 30689.46

All the forecast values are the same.
Is it due to the small dataset or my approach is not good?
Need advice.

Answer 1

By using model = "ANN" you are fitting a simple exponential smoothing model with additive errors (A). See help(ets) for possible models or leave the model argument out for automatic model selection. Your model includes no trend and no seasonality (NN).

Mathematical details on the possible models are given in A state space framework for automatic forecasting using exponential smoothing methods as stated on the help page for ets. As explained on pages 441 and 442, the series level l_t is a linear function of the original time series Y_t. In a model without trend and seasonality (e.g. ANN) the forecasts F_{t+h} are not dependent on h, F_{t+h} = l_t. This is why the forecasts in the above example are the same for all horizons, only the confidence intervals widen with increasing h.

I guess a discussion on which model is appropriate would be OT here, but I think your approach using exponential smoothing is reasonable given the short time series.

Answer 2

Also, what I found is that if you have time series with smaller number of values, then you will get the same forecasts. I guess this is due to the fact that the model is unable to derive seasonality or trend components from the available time series. But when I included more past data, I got better forecasts.