War Room

War Room: Scenario Simulation

Every crisis meeting opens with the question nobody in the room can answer: what would this event actually do to my input costs, and to my book? The War Room answers it before the event happens. You describe a macro shock in your own words — or pick a curated preset loadout — and the platform propagates it through statistically measured market relationships built from years of real price data. Minutes later you have a ranked list of which commodities would be hit, how hard, and how much wider the range of outcomes becomes, plus a ready-to-use action memo you can plan a budget or a portfolio around. That is the whole point: it turns “we should think about that risk” into a document you can take into a budget review.

Browsing the preset loadouts is open to every signed-in user; staging and running scenarios has to be enabled for your organisation, and your account manager can arrange that. Scenarios you save are shared with your whole organisation by default. That is deliberate — the War Room doubles as a team stress-test library, so the scenario a risk colleague built last quarter is one click away from running against today's market, and the whole team stress-tests from the same playbook instead of everyone improvising their own. There is one exception: a pair-trade stress test exported from the statistical-arbitrage screen stays private to its creator, because it encodes one person's trade idea.

Running your first scenario

A run is a repeatable routine: describe an event in plain English, refine what the platform understood, run it, then read the five result tabs in order. Steps one to three get a run started and oriented, steps four to eight walk the tabs in reading order, and step nine covers saving, scheduling, comparing and exporting. Your first run takes minutes; every run after that is muscle memory.

1. Describe the shock — The War Room landing page opens on a Scenario prompt — a large natural-language box whose placeholder rotates through example what-ifs, along the lines of “What if a major exporter halts shipments for a month?”. Type your own worry into it and press Stage and the platform reads your sentence and turns it into concrete, editable shocks. If you would rather start from something proven, click any Preset loadout card instead. The same page also holds your organisation's Saved scenarios and a Recent runs table, so repeating a stress test or reopening a past result is one click.
2. Review and refine — The Stage scenario dialog is your editorial pass: the AI proposes, you decide. If you arrived by typing a prompt, a parse-confidence badge tells you how sure the platform is that it read your sentence correctly — when it is red, read every proposed shock line by line, because a misread prompt produces a confident answer to the wrong question. The Shocks list is fully editable: change a commodity, correct an implausible magnitude, tune each shock's duration and intensity to test a milder or harsher version of the same event, reorder them, or add another. Give the scenario a clear name and description, then press Run scenario — or Save as preset only to park it in the shared library for later.
3. Watch the run — The run page is laid out like a command bridge. Four progress dots fill left to right, one per analyst — Macro, then GameTheory, then Quant, then Strategist — with a status chip beside them reading queued, running, completed or failed, and an elapsed timer while the work is in progress. Nothing blocks you while it runs. When results arrive, a top strip gives you the ten-second executive read: how many commodities the shock reaches, the single largest projected move, how many of the biggest movers went up versus down, and how many markets would be tipped into a different mood. A large reach with several mood changes means this event reshapes your market landscape rather than just nudging prices — and it deserves the full tab-by-tab read below.
4. Read the ripple map — Open the Map tab and read it from the centre outward. The commodities you shocked pulse in the middle; each ring out is one further step of transmission. The inner ring is the exposure you already knew about — the outer rings are the reason you ran the scenario.
5. Check the producers' negotiation — On the Game tab, read the Stability dial before anything else: it tells you how much of an announced supply cut is worth planning against. Everything else on the tab explains why each producer picks the move it picks.
6. Study the range of futures — The Markets tab turns the headline into a planning range. Read each fan chart as one image rather than reaching for the single number at the top of the card — the width of the fan is the uncertainty your scenario creates, and that is usually the more actionable half of the result.
7. Turn results into action — The Strategy tab is the action memo: pre-structured hedge tickets, a short list of real-world tripwires to put on your watchlist, and the strategist's own warnings about its own recommendations. Read the warnings before acting on anything above them.
8. Audit the run — Spend two minutes on the Log tab and in the Intel drawer before you circulate anything. This is the audit trail — the place where you satisfy yourself that the chain from market data to conclusion holds link by link, and where you find the explanation when a result looks odd.
9. Share, re-run, and keep watch — Export the run as a PDF for anything you intend to act on, save the scenarios that matter so they are re-tested for you automatically, and use Compare to set a recent run of the same scenario beside the current one. A saved scenario stops being a one-off exercise and becomes a standing early-warning system.

Reading the results: the five tabs

The Map tab. The commodities you shocked pulse at the centre and concentric rings mark steps of transmission: the inner ring is hit through a single measured relationship, the outer rings are hit second-hand. Node size is the size of the projected move, green is up and red is down, and a dashed halo means that market's whole character would change, not just its price. Beneath the graph, the Commodity impacts table carries the precise figures — Δ Median is the headline, the most likely price change caused by the shock alone; baseline and shocked typical paths sit beside it; σ Mult says how much wider the range of outcomes becomes; VaR 95% and CVaR 95% are the worst-case pair; and a hop count says how far down the ripple each market sits. Procurement can multiply Δ Median by annual spend on that commodity for a first-cut budget delta. Traders should check whether both legs of a pair trade land on the same side of the shock — one green leg and one red leg means the scenario splits the hedge. This is the screen that turns “a shock over there” into “an exposure over here”: a gas disruption quietly lifting fertiliser two steps out is exactly the second-order cost the map exists to surface.

The Game tab. Read the Stability dial first. Green means the modelled deal between the largest producers would hold, so an announced cut is worth planning against in full. Red means it is fragile — producers have an incentive to quietly over-produce, announced cuts tend to leak away in practice, and you should plan against a milder outcome than the press release implies. The player-strategy cards show each producer's chosen move — cut, hold or expand — with a short rationale, and payoff figures that are a comparative score for ranking one move against another, never a profit forecast. On producer-quota scenarios the engine computes the negotiation itself and may replace the quota number you typed with the outcome the producers would realistically strike; on other scenarios the tab carries the analyst's strategic reading instead.

The Markets tab. The most affected commodities appear as fan-chart cards. Read each fan as one image: the dark core is where half of the simulated futures land, the pale outer fan is where nearly all of them land, and the gap between the dashed baseline median and the solid shocked median is what your shock alone did to the most likely path. Filters split winners from losers and isolate the markets that would change mood. Procurement should budget to the top edge of the shocked band rather than the median — the median is the likely story, the top edge is the one that breaks a budget. Traders should lay the band against their stop levels: if the dark core crosses your stop, the position does not survive the scenario, so size smaller or restructure before the event rather than after. One honest caveat: a Projection caution banner means the full simulation was not available and the figures on screen are directional estimates only — right for orientation, wrong for budgeting.

The Strategy tab. Hedge tickets are laid out like dealing slips — a long chip naming what to buy, a ratio in the middle, a short chip naming what to sell, a one-line rationale for why that pair offsets the scenario's damage, and three price tiles for entry, exit and stop: where the trade starts, where it ends well, and where it ends badly. Below the tickets sit three to five monitoring tripwires — freight rates, inventory levels, policy announcements — which are the difference between a scenario you filed away and a scenario you are standing guard over, and a red-bordered list of risk warnings. Be clear about what these are: AI-drafted decision support built on the engine's numbers, not instructions. Traders validate every ticket against their own book, liquidity and limits before an order goes in; procurement treats it as a specification to hand to whoever executes financial hedges.

The Log tab and the Intel drawer. The analyst timeline shows each analyst's one-line job, its status and its finished summary; the event log records every stage of the run from the moment it was queued to completion, or to the error if it failed. Alongside every tab, the Intel drawer keeps each analyst's full write-up one click away, together with the economies the scenario touches and any warnings the platform raised about how it read your prompt. Two minutes here means you can answer “where did this number come from?” in the meeting rather than after it — and that is precisely what lets a budget or a position rest on the result.

Keeping a scenario alive after the run. Export PDF produces a branded report in reader order — executive summary, the negotiation analysis, the quantitative assessment, the top movers with portfolio-level worst-case measures shown baseline against shocked — a deliberately coarse whole-book read, for comparing how much riskier the scenario makes everything together rather than as an absolute figure for your own positions — and the hedging recommendations with their tripwires and warnings. Every content page carries the footer “Confidential — Scenario simulation, not investment advice”, which sets expectations for every reader you send it to. Saved scenarios shared with your organisation are re-run for you automatically each morning after the daily data refresh; those runs carry a Scheduled chip and cannot be cancelled. And Compare lets you mark a recent run of the same scenario and read the two next to each other — worth doing, because every run rebuilds from live correlations and prices, so yesterday's answer and today's can genuinely differ.

What good looks like

• Describe the shock in plain English and press Stage — or click a preset loadout, which opens the staging dialog pre-filled.
• Stage it carefully: check the parse-confidence badge, tune each shock's magnitude, duration and intensity, then name it.
• Run it and watch the four analyst dots march from Macro to Strategist.
• Read the Map to see which commodities are hit, how hard, and how far down the ripple chain.
• Check the Game tab on producer-quota scenarios — who leads, what they agree, and whether the deal holds.
• Study the Markets fans for the realistic range of outcomes, not just the headline move.
• Read the Strategy tab for hedge tickets, tripwires and the strategist's own warnings.
• Audit the Log and the Intel drawer so every number has a traceable source before it leaves the room.
• Export the PDF, let the daily re-run stand watch, and compare runs as the market moves.
• The payoff: the crisis meeting happens before the crisis. When the real headline lands, you have already seen the ripple map, priced the worst case and drafted the protection — all that remains is to execute a plan you have rehearsed.