Risk-weighted shipping factor model
Spot
~$200k
Spot peak ~$400k
1yr TC
~$100k
+144% above avg
3yr TC
~$78k
Market expects decline
Spot / 3yr
~4.0x
Extreme backwardation
Max VLCC orderbook over the past 2–4 years. Ships ordered at peak orderbook deliver 2–3 years later and weigh on rates for 3–7 years afterwards. High value = market still absorbing heavy prior deliveries. Essential for 3yr returns in cyclical industries.
Current: 5yr VLCC ~$140M / newbuild ~$128M = 1.09x (sell signal). Buy signal: 5yr < newbuild. Source: Clarksons/VesselsValue.
TC curve shape
Strongly inverted — market pricing in sharp rate decline
Context
Backwardation
−80%
spot → 3yr decline
Spot vs mean
8.9×
vs $41k long-run
Spot/3yr (1yr: 10% / 3yr: 13%)
4.0x
Ext. backwardation
1yr/3yr (1yr: 30% / 3yr: 2%)
2.0x
Curve strongly inverted
E(1yr ret.)
—
E(3yr ret.)
—
Worst case
—
| Factor | 1yr weight | 3yr weight |
|---|---|---|
| 1yr TC | 35% | 12% |
| 1yr/3yr TC ratio | 30% | 2% |
| 3yr TC | 12% | 20% |
| Spot/3yr ratio | 10% | 13% |
| P/NAV + trend | 13% | 13% |
| Orderbook | 0% | 20% |
| Delivery overhang | 0% | 10% |
| S&P / newbuild ratio | 0% | 10% |
Curve insight: 3yr TC (~$50k) is well below 1yr TC (~$100k) — the market expects rates to fall sharply over the next 3 years. An inverted TC curve is one of the strongest cycle-peak signals in the dataset.
| Year | 1yr TC | 3yr TC | P/NAV | 1yr ret. | 3yr ret. | Score |
|---|---|---|---|---|---|---|
| 2001 | $28k | $26k | 0.80x | +55% | +120% | 78 |
| 2002 | $32k | $28k | 0.85x | +80% | +180% | 72 |
| 2009 | $22k | $24k | 0.65x | +25% | +40% | 80 |
| 2012 | $15k | $22k | 0.58x | +68% | +98% | 78 |
| 2018 | $26k | $25k | 1.08x | +135% | +56% | 71 |
| 2020 | $45k | $25k | 0.95x | +14% | +287% | 65 |
| 2021 | $28k | $28k | 0.65x | +74% | +164% | 81 |
| 2022 | $35k | $35k | 1.05x | +96% | +139% | 69 |
| 2004 | $75k | $52k | 2.00x | +5% | -20% | 12 |
| 2007 | $85k | $60k | 1.80x | -27% | -30% | 22 |
| 2015 | $52k | $40k | 1.60x | -46% | -56% | 38 |
| 2013 | $18k | $24k | 0.75x | -33% | -56% | 79 |
| 2023 | $48k | $45k | 1.15x | -22% | +114% | 60 |
| 2026 now | $100k | $50k | ~1.5x | ? | ? | 4 |
Note: Rows from 2007 onwards reflect the verified backtest (adj_close equity returns, FRO primary). Rows before 2007 are illustrative calibration anchors. The 2013 row shows an honest model miss — the signal was buy, but rates fell further into 2014–15. The cause: the model did not capture time-decay after the 2007–2008 ordering peak (orderbook 50% and 48%). The 2010–2013 delivery wave created lasting oversupply, and the market had to absorb the overhang for more than five years before structural normalisation. Conversely, 2018 should have scored higher — ten years after the peak, with an ageing fleet and a low orderbook, this is a classic bullish setup. A delivery-overhang variable has now been added to the model to capture this dynamic, weighted conservatively pending more cyclical data points. The 2023 row shows that a 1yr signal can be wrong mid-cycle even when the 3yr outcome is strongly positive. All observations with 1yr TC above $50k combined with P/NAV above 1.5x have yielded negative 3yr returns since 2000.
Geopolitical tension is double-edged at current levels. Further escalation can send spot higher short-term — but a diplomatic resolution, ceasefire or Hormuz normalisation will remove the geopolitical premium quickly and send rates sharply lower. Market is already pricing in significant risk premium.
↑ Escalation
↓ Resolution / normalisation
1yr TC vs. historical distribution
σ from mean
+2.6σ
Extreme — top 1%
% above mean
+144%
Mean: ~$41k/day
Volatility band
$18k – $57k
±1σ (68% range)
VLCC earnings vol.
68%
StdDev/Mean (Stopford)
Cycle length distribution (years)
Mean: 10.4yr · σ: 4.9yr · Range: 3–20+yr
Peaks last
weeks – years
Tanker peaks: sharp spikes
Troughs last
6 mth – 6 yr
Longer than peaks
Stopford: "There is no simple formula for predicting the shape of the next stage, far less the next cycle. Recoveries can stall and slump back in months or last five years. Peaks may last a month or a year." Tanker cycles are more volatile, with sharper spikes and deeper troughs than dry cargo — exit timing is critical. Rule of thumb ~7yr, but cycles range from 3 to 20+ years. The industry view of Euronav (CMB) is 5–10 years.
SELL / AVOID — score 70–100
1yr TC above $60k and/or spot/3yr above 2x. E(1yr): −5 to +5%. E(3yr): −15% to −60%.
TRIM / ACCUMULATE — score 38–69
1yr TC $30–60k. TC curve near flat. E(1yr): 0–30%. E(3yr): −10% to +55%.
BUY / STRONG BUY — score 0–37
1yr TC below $35k, spot below 3yr TC, normal/contango curve. E(1yr): +20–70%. E(3yr): +45–130%.
Dual-horizon model: 1yr (momentum) and 3yr (mean-reversion) with separate weight sets. See weight table for details. Spot removed as standalone factor. Non-linear scoring. Hist. avg 1yr TC 2000–2024: ~$41k. Not investment advice. Sources: DHT, Frontline, Baltic Exchange, Argus Media, Clarksons — March 2026.
Expected CSV format:
year,spot,tc1yr,tc3yr,tc5yr
2000,65,35,30,28
2001,40,28,26,27
...
Optional columns: ret1yr and ret3yr (actual return in %) — enables backtest scatter showing signal vs. actual outcome. NAV column (nav) can also be included (0.4–3.0).