상세 컨텐츠

본문 제목

BDI & SCFI prediction (‘22/16W)

Forecast

by Adrian909 2022. 4. 17. 09:00

본문

반응형

 The BDI, SCFI predicted last week and the average actual BDI, SCFI the last week (4/11~14) are compared and tested.

 

1. Last week's forecasts

1-1. BDI, Actual Vs. Predicted

Date Last week Forecast Actual BDI MAPE(%) Accuracy(%)
‘22/15W(4/11~14) 2,034.748 2,067.75 1.6% 98.4%

 were achieved.

 1-2. SCFI, Actual Vs. Predicted   

Date Last week Forecast Actual SCFI MAPE(%) Accuracy(%)
‘22/15W(4/11~14) 4,247.188 4,228.65 0.4% 99.6%

 The prediction accuracy of BDI and SCFI was 98.4% and 99.6%, respectively. The predicted values (2,034.748) and actual values (2,067.75) of BDI, and predicted values (4,247.188) and actual values (4,228.65) of SCFI all achieved Trend and Accuracy. In the future, I plan to compare and test the predicted values two weeks later.


2.     BDI Outlook

 

2-1. ARIMA, Time Series Analysis

 

 To forecast the BDI, I used weekly data (428 pieces) from January 3, 2014 to April 14, 2022 to predict the future BDI for three weeks.

Date Forecast Trend(WoW)
‘22/16W(4/18~22) 2,187.904
‘22/17W(4/25~29) 2,225.078
‘22/18W(5/2~6) 2,236.571

  

 As shown in the table and figure above, BDI (as of 4/14: 2,067.75) is expected to stop falling and rebound.

In the 16th week (4/18-22), the forecast value is 2,187.904, which is expected to increase by 5.8% compared to 4/14, and this upward trend is expected to continue to 18W.

Ref) BDI & SCFI prediction (‘22/15W) (tistory.com)

 

2-2. logistic regression analysis

 

Logistic regression analysis was performed with the same data predicted by the ARIMA model previously.

 The expected trend for next week is 0.187, below the baseline of 0.5, so the direction of the market is predicted to decline. It is slightly higher than the previous week's forecast of 0.178, but unlike the ARIMA predicted earlier, the logistic regression model shows that next week will still decline.


3.     SCFI Outlook

3-1. ARIMA, Time Series Analysis

For the SCFI index forecast, I used weekly data (626 pieces) from October 16, 2009 to April 15, 2022 to predict the future three-week freight rate.

 

Date Forecast Trend(WoW)
‘22/16W(4/18~22) 4,215.044
‘22/17W(4/25~29) 4,211.921
‘22/18W(5/2~6) 4,206.928

 As shown in the table and figure above, the SCFI (as of 4/15: 4,238.65) still cannot stop its pace and is expected to continue to decline. The difference from the 15th week analysis (BDI & SCFI prediction (‘22/15W) (tistory.com)) is that the SCFI remains flat at the 17th week and then declines again at the 18th week. There are several issues that are considered a rebound signal, but if you look only statistically/mathematically, you can see that the downtrend is still very strong.

 

3-2. logistic regression analysis

 

 The result of the logistic regression analysis model of SCFI is also the same as the ARIMA value. With the forecast value of 0.38, we expect the SCFI to move down next week. However, the trend is slightly higher than last week's analysis value ((0.368), and it is estimated that the decline will be moderate to some extent.

 

Thanks.

728x90

'Forecast' 카테고리의 다른 글

BDI & SCFI prediction (‘22/18W)  (0) 2022.05.02
BDI & SCFI prediction (‘22/17W)  (0) 2022.04.23
BDI & SCFI prediction (‘22/15W)  (0) 2022.04.10
SCFI Prediction Using SVR Model  (0) 2022.04.09
BDI Prediction Using Logistic Regression Analysis  (0) 2022.04.08

관련글 더보기

댓글 영역