Screen-print style poster of a risk gauge over an LNG carrier

CASE STUDIES / RISK · AUG 2​025

The same book: $46M or $64M at risk.

One LNG portfolio, one day, two accepted ways to measure the loss it could take. Measured against price history the answer is $46M; measured against a thousand simulated futures it is $64M. Which method you pick is itself a risk decision.

$64.5M

THE ONE-DAY LOSS SIMULATION SEES AT 95% CONFIDENCE · HISTORY ON THE SAME BOOK SHOWS $46.1M

View as slides (PDF) →

A small book with four different risks.

01 · THE BOOK

Four legs, twenty-four contractual positions, one cargo every two months from September 2025. US length feeds Germany, Qatari length feeds the UK, and every leg carries a different index: HH, TTF, Brent and NBP. Twelve paired journeys make up the book, worth $169.2M at the curve, and each one carries its own risk profile.

CONSTRUCTED EXAMPLE BOOK · INDEX VALUES OF 20 JULY 2​025 · HISTORY JAN 2​024 TO JUL 2​025 · 1,000 SIMULATIONS · PORTFOLIOS ARE NOT RE-OPTIMISED ACROSS PATHS

THE BOOK · 3.5 TO 3.9M MMBTU PER CARGO

FOB Long · Calcasieu Pass6 CARGOES · 110% HH + $1
DES Short · Brunsbuettel6 CARGOES · 95% TTF
FOB Long · Qatar6 CARGOES · 9% BRENT
DES Short · South Hook6 CARGOES · 85% NBP
Fleet2 VESSELS · 171K CBM EACH

The same book, two answers.

02 · TWO METHODS

Historical VaR replays how the book would have fared under every past day's price move. Monte Carlo runs it through a thousand simulated futures instead. Simulation reports about 40% more risk on value at risk and nearly 50% more on expected shortfall, because it contains scenarios the recent past never produced.

$0M$25M$50M$75M$100M46.164.5+39.7%VAR · 5%58.282.1+41.1%VAR · 1%51.476.3+48.4%SHORTFALL · 5%61.190.5+48.2%SHORTFALL · 1%HISTORICALMONTE CARLO · POTENTIAL ONE-DAY LOSS, $MSAME BOOK, SAME DAY · HISTORY JAN 2​024 TO JUL 2​025 VS 1,000 SIMULATIONS
EXACT ENGINE RESULTS · HISTORICAL $46.1M / $58.2M / $51.4M / $61.1M · MONTE CARLO $64.5M / $82.1M / $76.3M / $90.5M · VALUES ARE LOSSES

Risk is not spread evenly.

03 · PER CARGO

Computing VaR cargo by cargo is what makes the number actionable. The odd cargoes sail Calcasieu Pass to Brunsbuettel, the even ones Qatar to South Hook, and each carries a different delivery month and index pair. The spread between the cheapest and the most expensive cargo is nearly tenfold.

$0M$2M$4M$6M123456789101112CARGO 7 IS THE CHEAPEST TO CARRY:THE STUDY CITES NO INITIAL BALLAST LEGCARGO · ODD = CALCASIEU PASS TO BRUNSBUETTEL · EVEN = QATAR TO SOUTH HOOK1% VAR5% VAR · HISTORICAL, PER CARGO, $M
HISTORICAL VALUE AT RISK PER CARGO · READ FROM THE ENGINE CHART · THE THREE LARGEST ARE ALL QATAR TO SOUTH HOOK · THE SOURCE STUDY MAPS ODD CARGOES TO CALCASIEU PASS YET EXPLAINS CARGO 7 AS THE FIRST QATAR JOURNEY

Simulation is higher and flatter.

04 · CONSISTENCY

Put the two methods on the same scale and the difference in character shows. Historical VaR swings from cargo to cargo because it inherits whatever each index actually did in that month; the simulated model prices are less sensitive to realised volatility, so the estimates come out both larger and far more uniform.

$0M$5M$10M123456789101112SIMULATION IS HIGHER AND FLATTER · HISTORY SWINGS FROM CARGO TO CARGOCARGO · 5% VALUE AT RISK PER CARGO, $MHISTORICALMONTE CARLO
5% VAR PER CARGO, BOTH METHODS · MONTE CARLO VALUES READ FROM THE ENGINE CHART · 1,000 SIMULATIONS

Re-indexing moves risk, it does not delete it.

05 · THE LEVER

The three riskiest cargoes are all Qatar to South Hook, so the UK leg is re-indexed to 50% NBP plus $4: less index exposure, more of the price in a static dollar adder. Portfolio 5% VaR falls from $46.1M to $41.8M and expected shortfall from $51.4M to $46.8M, with profit essentially unchanged at +0.18%. One cargo gets riskier in the process: risk moves, it rarely disappears.

$0M$10M$20M$30M$40M$50M46.141.8PORTFOLIO4.73.6CARGO 125.05.0CARGO 44.14.9CARGO 6BEFOREAFTER · GREEN LOWER RISK, AMBER HIGHER5% VAR BEFORE AND AFTER RE-INDEXING · PROFIT HELD FLAT (+0.18%)
PROFIT $169.2M TO $169.5M · THE SOURCE STUDY QUOTES THE UK LEG AS 85% NBP IN THE PORTFOLIO OVERVIEW AND AS 80% NBP ON THE ADJUSTMENT PAGE; THE TARGET FORMULA IS UNAMBIGUOUS

06 · THE VERDICT

Choosing the method is itself a risk decision:
the same book reports $46M or $64M depending on the lens.

History tells you what the market has already done to a book like this. Simulation tells you what it could still do. A risk report that quotes one number without naming the lens is only half an answer.

What the two lenses each see.

07 · TAKEAWAYS

✓

The method moves the number by 40%

The same book on the same day reports a 5% one-day loss of $46.1M on history and $64.5M under simulation. Expected shortfall diverges even further, by roughly 48%. Picking the method is a risk decision in itself.

✓

Cargo-level VaR shows where risk actually sits

History swings from cargo to cargo because each leg carries different months and indices; simulation is flatter because model prices are less sensitive to realised volatility. Cargo 7 is the cheapest of all, which the study puts down to a missing initial ballast leg.

!

Re-indexing lowers portfolio risk, but not everywhere

Moving the UK leg to 50% NBP + $4 cuts portfolio 5% VaR from $46.1M to $41.8M at almost unchanged profit, yet one cargo gets riskier. The effect depends on the chosen window, index and adder: this is an illustration, not a pricing recommendation.

Modelled with X-LNG.

08 · THE TOOL

1

Your world goes in

Contracts, vessels, charter rates, prices, spot assumptions and constraints. The full book, not a slice.

2

One optimal plan comes out

Feasible, P&L-maximising and constraint-compliant, re-solved for every scenario in minutes.

3

Every number checks out

Each result can be recalculated by hand. Transparency your risk committee can audit.

CLOUD-BASED, ANY PORTFOLIO SIZE · BUILT AND ADVANCED DAILY BY ~25 MATHEMATICIANS, PHYSICISTS AND COMPUTER SCIENTISTS