Shipping Takes a Tumble
What’s dry bulk shipping? And what happened in the Panamax market in January 2019?
This article is aimed at all readers. I present Fractal’s analysis of a recent sharp fall in the shipping markets, and start by introducing the basics of dry bulk shipping.
If you’re part of the shipping industry, we present our analysis here.
What is dry bulk shipping?
I’m currently a Data Scientist at Fractal Logistics. We’re in the ‘dry bulk’ shipping industry: dry bulk is everything that is not ‘wet’ (oil) and is not yet manufactured (so not shipping containers like Maersk).
The main commodities in dry bulk are:
- coal
- iron ore
- grains
Grains support global food chains, coal supports global energy chains, and iron ore is critical to our steel-guzzling urban life. I therefore like to think that dry bulk props up our modern way of life.
Why are shipping prices volatile?
The price of shipping is v o l a t i l e. Day-on-day changes are regularly more than 5% and the market exhibits strong momentum: it is common for 20% changes in a two-week period.
The main causes of this volatility are:
- Unexpected geopolitical or natural events
China suddenly imposes tariffs on US grains and drought hits a harvest of Brazilian soybeans — this fundamentally changes trade flows. - Misallocation of shipping supply
Big ships are slow, there aren’t that many of them, and the world is big. When there is an under- or over-supply of ships in a region, prices will see a corresponding sharp and sustained rise (under-supply) or fall (over-supply).
What does Fractal do?
At Fractal we specialise in predicting misallocation of shipping supply.
Each day our servers crunch 3,000,000 vessel locations (‘AIS’ data) from 20,000 ships. We estimate whether a vessel has just entered Santos (Brazil) and is loading soybeans, or is fully laden with coal from Queensland (Australia) heading for Qingdao (China). From this data we derive a granular, real-time understanding of global shipping flows.
Based on this insight, we position our ships in areas where we predict an increase in shipping prices in the coming weeks.
What’s the point of this article?
At Fractal we focus on the ‘Panamax’ market: ships ranging from 66,000 to 100,000 deadweight tonnes. The average price per day to hire a Panamax ship is called the ‘Panamax 4TC’. In mid-October 2018, the 4TC was at a near five-year high in the $14,000s. Three months later, the 4TC hit its lowest level since June 2016.
In this article, we outline our view on three main questions:
- How does this fall compare with others?
- What was the fundamental cause of the fall?
- How can we explain the timing of the fall?
Summary answers:
- The second fastest fall since 2011 and the largest 50-day drop
- 3.5-year low in Capacity Utilisation
- i) no Gulf of Mexico season; ii) low supply in the Atlantic; iii) East Coast South America veiled the fundamentals; iv) Swine Flu hits Chinese soybean demand; v) sudden Chinese coal restrictions
1) How does this fall compare with others?
Summary: The second fastest fall since 2011 and the largest 50-day drop
Below we plot the Panamax 4TC Index since 2011 (black), alongside the 90-day rolling average (blue). The gap between the two provides a useful reference for recognising sharp falls. In grey, we highlight the nine major falls since 2011.
Below are the summary statistics of these falls. From 19th December 2018 to 4th February 2019, the 4TC fell continuously day-on-day from $12k to $4.5k — an average daily drop of 2.1%.
We use percentage change, not absolute change, so we can easily compare across drops which begin at different index levels. We also define the duration of a fall as the number of calendar days, not business days, as we think this is more representative of the adjusting fundamentals. This is especially true for the Christmas season when the index is not reported for a prolonged period.
We note that:
- Among the shorter duration falls (<=50 days), Jan 2019 is the largest percentage fall
- Overall, Jan 2019 is the second fastest daily fall
2) What was the fundamental cause of the fall?
Summary: 3.5-year low in Capacity Utilisation
‘Capacity Utilisation’ is a high-level metric to signal the extent of spare freight capacity. In a market with weak freight demand, ships tend to sail slower and are more likely to be idle — lowering capacity utilisation. When capacity utilisation is low, this signals spare capacity and exerts downward pressure on freight prices. The chart below shows capacity utilisation since 2015:
Just as the 4TC was hitting a near 5-year high in mid-October 2018, the 90-day mean capacity utilisation fell to a near 3-year low. By mid-December, the 90-day mean hit lows not seen since April 2015.
It is rare to see such a large disconnect between capacity utilisation and the spot freight market. On this basis, the market was poised for a sharp downward correction.
3) How can we explain the timing of the fall?
Summary: i) no Gulf of Mexico season; ii) Low supply in the Atlantic; iii) East Coast South America veiled the fundamentals; iv) Swine Flu hits Chinese soybean demand; v) sudden Chinese coal restrictions
We believe capacity utilisation set the stage for a market correction. However, we need to consider other factors to explain why the correction only materialised in late December.
i) The Gulf of Mexico export season that wasn’t
“Although the 25% tariff on Chinese imports of US soya beans was implemented on 6 July 2018, its effects have only really started to show since the start of the fourth quarter. Last year, 55% of US soya bean exports to China occurred between October and December…US exports to China have dropped by 98% since 1 September compared to the same period last season…China has taken just 2.8% of US soybean exports since Sep 1.”
— Peter Sand(BIMCO), 2019–01–16
‘Tonmiles’ is a common shipping measure of how far ships are sailing (miles) weighted by their capacity (tonnes). Exports from the Gulf of Mexico (USA), mostly soybeans to China, normally account for ~9% of world Panamax laden tonmiles.
The chart below plots expected GoM laden tonmiles, based on historical seasonal trends (blue), vs. actual laden tonmiles (black). In previous years, the exports from GoM pick up in September in line with the soybean harvest. In 2018, the GoM soybean export season was non-existent (black).
A missing exports season from GoM is the principle reason for the fall in capacity utilisation. From early October 2018, this non-existent season of exports from GoM placed immense latent downward pressure on the 4TC.
ii) Low supply in the Atlantic
However, we must consider freight supply to understand why this lack of demand didn’t translate into low Atlantic rates from October. Given that the tariff had been announced in July 2018, fewer vessels had repositioned in the Atlantic in Q4 2018. Below we plot the number of vessels in the North Atlantic by month for each year. The graph for 2018 (orange) shows a seasonally low number of ships in the North Atlantic during October and November.
This low Atlantic supply helped to support Atlantic shipping rates. However, by December 2018, supply was back in line with historical norms — piling downward pressure on the index.
iii) East Coast South America soybeans partly compensated in early Q4
Along with Gulf of Mexico, East Coast South America (ECSA) is the other major global provider of soybeans. From the beginning of October 2018, ECSA replaced GoM to meet China’s soybean demand. At the moment that shipments weren’t departing from GoM, there was a flurry of fixing in ECSA for soybeans to China.
This led to a (seasonally adjusted) sharp uptick in shipments from ECSA from mid-November, which somewhat compensated for the market’s underlying weakness.
However, standard voyages from ECSA to China are ~10 days shorter than voyages from GoM to China. Therefore, the same amount of soybean imports required fewer laden tonmiles. By December, the odds were still stacked against the index.
iv) Chinese soybean demand weakened from November
“Data from China’s General Administration of Customs shows that the country’s soybean imports in the final two months of 2018 plunged by 39% from the previous year.”
— Karen Braun (Reuters), 2019–01–16
In mid-October 2018, the first outbreak of African Swine Flu (ASF) was discovered in South China. Guangdong banned the transport of live pigs after an outbreak was discovered in neighbouring Hunan province. This appeared to spark latent fears of a broader spread of ASF, and Chinese soymeal futures started tumbling from highs in mid-October (soymeal is used to feed hogs).
The ‘soy crush margin’ measures excess demand for the products of soybeans (e.g. soymeal to feed hogs). As seen below, Chinese soy crush margins fell sharply from mid-October 2018, turning negative by early November.
Source: Clarksons Platou & Reuters
Falling crush margins meant Chinese demand for soybean imports was particularly weak from November 2018. Although ECSA soybean exports were strong in October 2018 from compensating for GoM’s non-season, weak Chinese demand in November created a corresponding drop in ECSA exports at the end of Q4 2018.
v) China restricted coal imports from mid-November
“China Customs from the 14th January showed that the country imported 10.2mt coal in December 2018, down 55% compared to 22.7mt in December 2017. This was put down to the strict customs clearance controls implemented during December.”
— Clarksons Platou, 2019–01–14
In mid-November 2018, China had reached its yearly quota for imports of coking coal. In a surprise move with no official announcement, Chinese Customs began implementing strict, lengthy customs controls on imported coking coal from mid-November.
‘Congestion’ is a common shipping measure for how long ships have to wait in a port before (off)loading their cargo. The strict Chinese import controls created lengthy congestion in Chinese ports from mid-November. The congestion is a symptom, not a cause, of the strict customs controls.
As coal imports into China constitute over 20% of global laden tonmiles, this was the straw that broke the 4TC’s back.
Sentiment crashed just before Christmas 2018. After Christmas, we experienced a sharp and painful market correction.
Thank you from everyone at Fractal Logistics. Feel free to check out our previous work quantifying the relationship between the Panamax markets.
Our inbox is always open 😎
research@fraclog.co
Want to join us playing chess from space? 🤓
recruiting@fraclog.co
