Data fundamentals · 7 min read
Port congestion as a trading signal: how vessel backups move markets
A queue of ships waiting outside a port is a physical fact that hasn't hit the tape yet. That lag — between what's happening on the water and what shows up in prices — is the whole opportunity.
Markets price information as it arrives. Port congestion is information that arrives early: the vessels backing up at anchor today are the freight rates, the inventory builds, and the earnings surprises of next week. By the time congestion shows up in a shipping company's quarterly report or a freight-rate index, the move is largely over. Reading the congestion directly — from the ships themselves — is how you get there first.
What "congestion" actually measures
Every commercial vessel broadcasts its position, speed, and status over AIS. From that raw feed you can derive three things that matter:
- Vessel count in the port's waters — how much traffic is present right now.
- Anchored ratio — what share of those vessels are sitting still at anchor rather than moving. A high anchored ratio is the signature of a backup: ships waiting for a berth, or laden tankers holding cargo offshore.
- Dwell time — how long vessels stay before departing. Rising dwell means the port is clearing slower than ships are arriving.
A single snapshot of these numbers is nearly useless, because ports have strong daily and weekly rhythms. What matters is the deviation from normal. A congestion score is only meaningful when it's measured against that port's own baseline for that hour of the day — which is why a raw "85/100" matters only once you know the port usually runs 50.
Advertisement
Why it leads price
The lag comes from physics and accounting. A backup at a container hub takes days to translate into higher spot freight rates, and weeks to show up in a carrier's reported numbers. A surge of laden crude tankers leaving Houston tightens domestic supply before the inventory draw prints in the EIA report. Anchored tankers building up offshore signal a supply glut before the futures curve fully reflects it. In every case the vessel data moves first, typically by five to ten days, because ships are slow and reporting is slower.
The direction of the read depends on the port. At a container gateway (LA/Long Beach), a backup lifts freight rates and is bullish for carriers like ZIM and Matson. At an energy export terminal (Houston, Sabine Pass), a surge in tanker activity signals strong export demand and tightening domestic supply — bullish crude or natural gas. At a floating-storage hub, a rising anchored ratio signals oversupply — bearish crude. Same data, opposite trade, depending on what flows through.
The failure modes (read this before you trade it)
- Coverage gaps look like signals. If AIS reception drops, vessel counts fall — and a coverage hole can masquerade as a port emptying out. Always separate "no ships" from "no data."
- Congestion can already be priced. If the correlated stock has already run, the move may be done. Check whether price has moved before you assume the signal is fresh.
- Persistent ≠ repeatable. A port that stays congested for a month isn't a new signal every day — it's one event. The edge is in the change, not the level.
- Classification is noisy. Vessel types are self-reported and sometimes wrong, so "tanker vs LNG carrier" splits are approximate.
How to turn it into a trade
The cleanest setups stack the congestion read with a corroborating signal. Houston tanker surge plus a compressing Brent-WTI spread is a stronger US-export read than either alone. A container backup plus rising freight rates is stronger than the vessel count by itself. And every entry needs an invalidation: the level or condition that says the thesis is wrong — usually "congestion reverts toward baseline within 48 hours."
What HarborSignal shows
Each port page shows live vessel count, anchored ratio, dwell time, and a congestion score with its percentile versus the port's own 30-day history — plus a lead-lag readout of how that port's congestion has tracked its primary ticker. When an anomaly clears the statistical threshold, it becomes a signal with a direction, an entry, an invalidation level, and a 7-day outcome score. The data is the same ground truth the freight desks watch — just on one screen, and free.