How Shipping Prices Behave
The forces behind commodity, freight and bunker prices: how much each major market shock actually reshaped them, the seasonal rhythms buried in each series, and an honest read on how far ahead prices are really foreseeable. For the full per-series decision canvas, see the Bunkers, Commodities and Freight desks.
How prices reacted to major shocks
For each market shock, the immediate move in the prices it plausibly touched — the change from the month before the event to the month after. This is the market's realized reaction with all concurrent forces included: co-movement around the date, not isolated causation. Routing shocks whose effect lands on freight rather than spot prices are covered elsewhere.
Black Sea grain deal collapses
Russia exits the grain deal; Black Sea shipping risk returns.
- CommodityMaize-22%
- CommodityWheat (US HRW)-9%
Black Sea Grain Initiative
Deal reopens Ukrainian grain corridors, easing agri prices.
- CommodityWheat (US HRW)-17%
- CommodityMaize-14%
- CommoditySoybeans-9%
Russia invades Ukraine
Invasion triggers an energy and grain shock; oil and gas spike.
- CommodityCoal (South African)+75%
- CommodityCoal (Australian)+59%
- CommodityNatural gas (Europe)+50%
- BunkerHouston MGO+47%
- CommodityDiesel/heating oil (NY Harbor No.2)+47%
- BunkerPort of Rotterdam MGO+45%
- BunkerPort of Singapore MGO+43%
- CommodityCrude oil (Dubai)+36%
China commodity restocking
Post-COVID Chinese demand drives iron ore and coal to records.
- CommodityCoal (Australian)+41%
- CommodityCoal (South African)+27%
- CommodityIron ore (cfr spot)+19%
- CommodityDiesel/heating oil (NY Harbor No.2)+12%
COVID-19 demand collapse
WHO declares a pandemic; global oil and freight demand collapses.
- CommodityCrude oil (WTI)-67%
- CommodityCrude oil (WTI, EIA spot)-67%
- CommodityCrude oil (Brent, EIA spot)-67%
- CommodityGasoline (NY Harbor, conventional)-62%
- CommodityCrude oil (Brent)-58%
- CommodityCrude oil (Dubai)-57%
- CommodityDiesel/heating oil (NY Harbor No.2)-46%
- CommodityCoal (South African)-29%
The calendar pattern in each price
A robust seasonal-trend decomposition of each series' monthly history. Bars show the typical seasonal multiplier by month — above the line means that month usually runs hot versus the de-seasonalised trend. Only series with a genuine calendar signal are shown.
Port of Rotterdam HSFO
Fujairah HSFO
Port of Los Angeles HSFO
Port of Singapore HSFO
Port of Los Angeles VLSFO
Fujairah VLSFO
Port of Singapore VLSFO
Port of Santos VLSFO
Where a forecast actually earns its place
We roll a strict out-of-sample backtest on every series and keep a directional 3-month projection only where the model genuinely beats a random walk (MASE below 1). Everything else is left as history and an uncertainty cone — because most prices are, honestly, near-random walks and a confident call would be a lie.
| Series | 3-mo projection | 80% range | Backtest skill |
|---|---|---|---|
BunkerPort of Los Angeles VLSFO | 1,012 | 917.1 – 1,106 | MASE 0.64 |
BunkerPort of Rotterdam VLSFO | 667.9 | 575.1 – 760.7 | MASE 0.76 |
BunkerPort of Los Angeles MGO | 1,467 | 1,288 – 1,646 | MASE 0.81 |
BunkerFujairah MGO | 1,225 | 1,041 – 1,410 | MASE 0.83 |
BunkerFujairah VLSFO | 713.7 | 449.6 – 977.9 | MASE 0.83 |
BunkerHouston MGO | 1,022 | 746.2 – 1,297 | MASE 0.83 |
BunkerPort of Singapore HSFO | 491.5 | 364.8 – 618.1 | MASE 0.84 |
BunkerPort of Singapore VLSFO | 694.7 | 544.7 – 844.6 | MASE 0.87 |
CommodityMaize | 182.1 | 165.3 – 198.8 | MASE 0.87 |
BunkerPort of Rotterdam HSFO | 584.5 | 471.7 – 697.4 | MASE 0.87 |
BunkerNew York-New Jersey VLSFO | 588.5 | 475.9 – 701.1 | MASE 0.90 |
BunkerPort of Rotterdam MGO | 1,063 | 777.8 – 1,348 | MASE 0.90 |
CommodityGasoline (NY Harbor, conventional) | 3.07 | 2.37 – 3.77 | MASE 0.90 |
BunkerPort of Santos VLSFO | 633.8 | 446.8 – 820.9 | MASE 0.91 |
BunkerHouston HSFO | 452.1 | 339.4 – 564.9 | MASE 0.92 |
CommodityCrude oil (Dubai) | 77.4 | 58.6 – 96.2 | MASE 0.93 |
CommodityCrude oil (WTI, EIA spot) | 74.6 | 52.7 – 96.5 | MASE 0.93 |
BunkerPort of Los Angeles HSFO | 755.6 | 652.4 – 858.8 | MASE 0.94 |
BunkerPort of Hong Kong VLSFO | 707.4 | 535.8 – 879 | MASE 0.94 |
BunkerPort of Singapore MGO | 989.2 | 750.3 – 1,228 | MASE 0.94 |
CommodityCrude oil (Brent, EIA spot) | 75.4 | 48.2 – 102.5 | MASE 0.94 |
BunkerNew York-New Jersey MGO | 1,143 | 823.6 – 1,463 | MASE 0.96 |
CommodityWheat (US HRW) | 293.2 | 269 – 317.3 | MASE 0.96 |
CommodityCrude oil (Brent) | 85 | 56.4 – 113.7 | MASE 0.96 |
CommodityNatural gas (Henry Hub, EIA spot) | 3.13 | 0 – 6.64 | MASE 0.97 |
CommodityNatural gas (US) | 3.07 | 0 – 6.47 | MASE 0.98 |
CommodityCrude oil (WTI) | 74.4 | 53.2 – 95.7 | MASE 0.99 |
BunkerPort of Hong Kong HSFO | 510.8 | 328.3 – 693.3 | MASE 0.99 |
A further 14 series show no directional edge over a random walk at this horizon. For those we deliberately publish no point forecast — only the history and its uncertainty cone.
Event reaction is the realized price change from the month before an event to the month after, for the series a shock could plausibly touch; concurrent market forces are included, so it is co-movement around the date, not proof of causation. Seasonality is a descriptive decomposition of past monthly history, not a forecast. The nowcast is gated on an out-of-sample backtest and shown only where it beats a naive baseline. Figures are analytical estimates for context, not investment advice. Market econometrics →