Change search
ReferencesLink to record
Permanent link

Direct link
Nowcasting US GDP with Baltic Dry Index: A study investigating the use of the MIDAS Model
Södertörn University, School of Social Sciences.
2016 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

One of the conclusions made in the aftermath of the last financial crisis was that forecasting was failing in the context of predicting the current economic activity, which meant there were few efficient instruments to monitor the economy and thereby no early stage intervention which could have mitigated the severity of the crises. 

Due to these facts presented nowcasting, forecasting in short horizon, and the use of models that could combine different data frequencies like the Mixed Data Sampling (MIDAS) model gained a lot of attention.

This study investigates if the MIDAS model improves nowcasting and if the Baltic Dry Index (BDI) is a good indicator for the US GDP.

Before making any conclusions from the result, the characteristic of the BDI is covered and explained as to why it could reflect general growth.

The intention of using BDI as an indicator for US GDP was to find an indicator that may explain economic activity in a more accurate way due to the characteristics of the BDI.

The result was in line with previous empirical work and proved that the MIDAS model is superior in nowcasting in comparison to the least square model with flat aggregation defined as a benchmark model.

The rejection of the cointegration test for BDI may question the use of it as an indicator for US GDP at present time, this due to the extreme circumstances that currently affect the bulk dry market.

Place, publisher, year, edition, pages
2016. , 44 p.
National Category
URN: urn:nbn:se:sh:diva-30543OAI: diva2:944910
Subject / course
Social and Behavioural Science, Law
Available from: 2016-06-30 Created: 2016-06-30 Last updated: 2016-06-30Bibliographically approved

Open Access in DiVA

fulltext(2217 kB)39 downloads
File information
File name FULLTEXT01.pdfFile size 2217 kBChecksum SHA-512
Type fulltextMimetype application/pdf

By organisation
School of Social Sciences

Search outside of DiVA

GoogleGoogle Scholar
Total: 39 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 74 hits
ReferencesLink to record
Permanent link

Direct link