BDI & SCFI prediction (‘22/22W)

In the 22nd week, BDI and SCFI predicted values and actual values are compared and tested.

---

1.     Last Week’s forecasts

1-1.        BDI, Actual Vs. Predicted

Date	Last week Forecast	Actual BDI	MAPE(%)	Accuracy(%)
‘22/21W(5/23~27)	3,185.337	3,072.60	3.7%	96.3%

 

1-2.        SCFI, Actual Vs. Predicted

Date	Last week Forecast	Actual SCFI	MAPE(%)	Accuracy(%)
‘22/21W(5/23~27)	4,159.141	4,175.35	0.4%	99.6%

 The prediction accuracy of BDI and SCFI was 96.3% and 99.6%, respectively. BDI predicted values (3,185.337) and actual values (3,072.60), and SCFI predicted values (4,159.141) and actual values (4,175.35) all achieved high accuracy. The singular point is that 1) the downward trend of BDI was accurately predicted in this report. Even though the BDI reached this year's high, according to the econometric model, it was accurately predicted that there would be a correction following the recent rapid rise. 2) SCFI has risen for the second week in a row. Therefore, the direction of SCFI could not be accurately predicted in this study. Nevertheless, the very high accuracy is due to the very low rise. Still, it is very difficult to see a clear transition to an uptrend.

  The average BDI MAPE value for the last 9 predictions is 3.9% and the prediction rate is 96.1%, and the SCFI average MAPE value is 0.6% and the prediction rate is 99.4%.

---

2.     BDI forecast

2-1. ARIMA, Time series analysis

 For the BDI index forecast, I used weekly data (434 pieces) from January 3, 2014 to May 27, 2022 to predict the freight for three weeks in the future.

Date	Forecast	Trend(WoW)
‘22/22W(5/30~6.3)	2,937.708	↓
‘22/23W(6/6~10)	2,813.338	↓
‘22/24W(6/13~17)	2,726.316	↓

 As shown in the table and figure above, BDI (as of May 27th: 3,072.60) fell by -4.0% compared to last week (as of May 20th: 3,200.40). In particular, if you look at the daily, you can see the trend of the freight rates dropping sharply for two days on Thursday and Friday beyond the middle of the week.

As a result of the ARIMA model, the BDI is also expected to decline next week and continue to decline until 24 weeks.

Ref) BDI & SCFI prediction (‘22/21W) (tistory.com)

BDI & SCFI prediction (‘22/21W)

 

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 0.520, close to the baseline level of 0.5. Above 0.5, there is a slight slope to the rise, but the width is very low. In other words, even in logistic regression, the direction of BDI next week is predicted to be flat.

---

3.     SCFI Outlook

3-1. ARIMA, Time series analysis

 For the SCFI forecast, I used weekly data (632 pieces) from October 16, 2009 to May 27, 2022 to predict the future freight for three weeks.

Date	Forecast	Trend(WoW)
‘22/22W(5/30~6.3)	4,177.124	↗
‘22/23W(6/6~10)	4,172.313	↘
‘22/24W(6/13~17)	4,167.565	↘

 

 As shown in the table and figure above, the SCFI (as of May 27th: 4,175.35) rose for two consecutive weeks following last week (as of May 20th: 4,162.69). However, the rate of increase is very low at 0.3%. It is still a difficult section to be certain about the direction.

 

3-2. Logistic regression analysis

 The predicted value of SCFI's logistic regression analysis model was 0.393, which predicted the next week's SCFI's freight direction to decline. The results are different from those of the ARIMA model previously shown. Also, it is similar to last week's logistic value of 0.38, and it still seems very difficult to determine the direction of the freight rate.

 

3-3. SVR model

 Like the previous analysis (SVM), the SVM model predicted a continuous decline in freight rates. However, considering that the predicted value is gradually increasing while reducing the decline, the current section is expected to be a major turning point in the rise or fall of freight rates. Keep an eye on SCFI's fare direction.

 

Date	Forecast	Trend(WoW)
‘22/22W(5/30~6.3)	4,078.451	↘
‘22/23W(6/6~10)	4,051.972	→
‘22/24W(6/13~17)	4,033.873	→

Thanks.