Diesel Fuel and Its Prices

Let's take a look at diesel fuel prices. An overview of the data is presented in the

figure below.

Analysis of Price Behavior tells us the following:

The series is non-stationary (Augmented Dickey-Fuller Test - p-value = 0.39), but the

series of first differences is stationary (see the figure below).

Components of the STL decomposition are shown below.

The strength of the trend (Ft) and the strength of the seasonality (Fs) are as follows:

Ft = 0.9968656

Fs = 0.05954845

The time series exhibits a strong trend and weak seasonality. This indicates that the

data shows a clear direction of change over time but does not exhibit significant

periodic fluctuations.

Forecast for the Test Data is Presented in the figure below:

*

The table below shows the results of the prediction and the observed values:

Date actual forecast

2024-06-06 52.59 52.71646

2024-06-07 52.57 52.72052

2024-06-10 52.57 52.81903

2024-06-11 52.57 52.75053

2024-06-12 52.56 52.74745

2024-06-13 52.61 52.66242

2024-06-14 52.60 52.57854

2024-06-17 52.60 52.53943

2024-06-18 52.60 52.49269

2024-06-19 52.59 52.47355

2024-06-20 52.61 52.53264

2024-06-21 52.61 52.50237

2024-06-24 52.61 52.63939

2024-06-25 52.66 52.65885

2024-06-26 52.65 52.76961

The maximum forecast error is 0.473715%. Be cautious in evaluating the results,

as the first differences exhibit a very slight upward trend. This could lead to

a gradual increase in prices if there are no "cataclysms" in the market!