The shipping industry is vital in keeping the world connected: allowing nations to trade with each other, supporting global supply chains, and ensuring that goods flow seamlessly across borders.

One of the most critical factors for success in sea transport is effectively monitoring shipping freight rates: generally defined as the amount of money paid to a shipowner to move a unit of cargo between two ports. However, because the shipping industry spans multiple sectors, each with its own contract types and complexities, this definition can be interpreted differently depending on the situation.

In the dry bulk shipping sector, for example, ships carry commodities such as grain, metal ores, or coal; while in the tanker sector, they transport liquid goods such as oil. In both cases, the freight rate is simply defined as the price for hiring the ship. For container shipping, however, the calculation is more complex. Here, costs are divided between payments from charterers (who lease the ships) to the shipowners, and from shippers (who own the cargo) to the carriers (who transport the cargo).

This multi-layered cost structure in container shipping adds complexity not only for companies managing operations, but also for economists and regulators who need to forecast market trends and ensure fair practices in the industry.

To address these challenges, numerous studies over the past two decades have sought to improve the understanding of freight rate calculations through quantitative models. In their study, Liu’s team provide a comprehensive review of these studies, shedding new light on the major factors that drive freight rates in today’s shipping industry.

Through their analysis, the team aims to offer new empirical insights into the dynamic behaviours of freight rates and the key variables that impact quantitative model freight rates across different shipping sectors. By analysing and categorizing this past research, they hope that improved future models could help companies, regulators, and economists to predict and manage freight rates more effectively.

Liu’s team identified all relevant journal articles on shipping freight rates published between 2000 and 2019. They collected a total of 168 papers and organized them into four categories: papers analysing time series characteristics in shipping freight rates over time, those examining factors behind changes in rates as time series, studies predicting future freight rates, and finally, those investigating cross-sectional dynamics of freight rates as non-time series data.

The models used in these studies varied widely depending on the goals of the research. For example, Liu’s team identified five key factors specific to the container shipping sector, which differed from the factors commonly used for the dry bulk and tanker sectors.

Firstly, studies on container shipping often focus on trade volume and transport distance as demand indicators, rather than the prices of commodities being transported. This is due to the diverse types of cargo that container ships handle, compared with bulk shipping, which typically moves a single type of commodity, such as iron ore or coal.

Secondly, factors such as service frequency, connectivity, and port infrastructure are crucial for container shipping freight rates because shippers rely on these services. In contrast, dry bulk and tanker charterers manage the transport of cargo themselves, making these factors less critical in determining freight rates.

Thirdly, container freight rates are significantly affected by trade imbalances. When there is a mismatch between the volume of imports and exports between regions, empty containers must be repositioned, which is costly for shipping companies. This logistical challenge doesn’t affect bulk or tanker shipping to the same extent.

Fourthly, freight rates in the container sector are influenced by market competition. Mergers, acquisitions, and the formation of shipping alliances create competitive pressures that lower freight rates. The greater the number of companies vying for business, the more likely prices are to drop. This dynamic is less pronounced in bulk shipping, where long-term contracts often shield rates from short-term competitive pressures.

Finally, the rise of larger container ships has played a significant role in reducing freight rates. Bigger ships can carry more cargo, which reduces the per-unit cost of transporting goods. In contrast, bulk freight rates are more closely tied to charter prices and vessel sizes. This distinction between container shipping and bulk shipping has led to different modelling approaches across the two sectors.

On top of these factors, Liu’s team also considered how freight rates are naturally subject to change due to varying supply, demand, market conditions, and external factors such as political instability or economic crises.

Some researchers believe that freight rates tend to stabilize over time, returning to a normal level due to the competitive nature of the shipping industry. However, most studies treat freight rates as non-stationary, meaning significant shifts happen over periods, as volatility and uncertainty characterize the industry.

Freight rates also follow patterns that repeat over time, including short-term cycles lasting around three to four years and longer cycles lasting up to 20 years. These cycles reflect broader economic trends, such as periods of economic growth followed by recessions.

The team also noted that freight rates frequently experience periods of high volatility, where several large changes occur in quick succession. While models are useful for capturing these shifts, Liu’s team showed that simpler methods are gaining popularity because they can better handle the inherent complexity of shipping data.

Through their comprehensive review, the team has mapped the progress of quantitative models used to analyse shipping freight rates over the past 20 years. Their work highlights significant advancements in the field, while also pointing out the limitations of previous studies. Classic econometrics remains the most popular method, although the use of machine learning has increased slightly over the past decade in shipping research. Machine learning generally outperforms traditional methods in forecasting, and hybrid models have demonstrated success compared to original methods.

Building on these findings, future researchers may be able to develop more accurate models that benefit the entire shipping industry: potentially helping shipowners and carriers optimize their operations. Similarly, regulators would gain access to more reliable data, enabling them to oversee the industry and ensure that markets remain competitive and fair.

Ultimately, Liu’s team believes that the continued development of these models will be crucial in helping the global shipping industry navigate an increasingly complex and interconnected world.