Scenario Modelling¶

Overview¶

Scenario planning in treasury involves creating multiple versions of cash flow forecasts based on different assumptions to prepare for uncertainty. Treasury teams use scenarios to stress-test their liquidity positions, model best/worst case outcomes, and make more robust financial decisions.

This matters because treasury must maintain adequate liquidity even when actual results differ significantly from base forecasts. By modeling scenarios like reduced collections, delayed payments, or economic downturns, teams can identify potential shortfalls before they occur and plan appropriate responses.

Key challenges include the manual effort required to create and maintain multiple scenarios, difficulty in applying systematic assumptions across forecast categories, and lack of tools that make scenario comparison intuitive.

For detailed ICP context and terminology, see fundamentals.md

Top Jobs & Desired Outcomes¶

Full history: jobs.md

1. Apply percentage-based assumptions while preserving patterns ✓¶

Desired Outcomes: - Minimize the risk of overestimating collections or underestimating outflows - Increase the ability to model scenarios quickly (e.g., -5% on collections) - Reduce manual effort in creating multiple forecast scenarios - Minimize the loss of weekly/daily patterns when adjusting forecast totals

Sources: Personio (2025-10-21, 2025-12-04, 2026-02-18), ON (2026-02-18, 2026-03-04) — Confirmed (5 sources, prototype tested live with ON)

"I would like to keep that rationale in there and not just layer on top like a blanket 2 million a week additional" — Tom, Personio

2. Validate indirect P&L budget assumptions using direct cash forecast ✓¶

Desired Outcomes: - Minimize the time required to sense-check controlling's budget against direct cash data - Increase confidence that FP&A's 5-year plan is realistic by cross-referencing with cash flow patterns - Reduce reliance on mental calculations when assessing whether budget assumptions hold - Increase ability to feed budget-level assumptions (e.g., +40% inventory) into the direct forecast and see the cash impact

Sources: Instacart (2025-07-01), ON (2026-03-04) — Confirmed (2 sources)

"It's very difficult for me to tell if it's realistic or not because it's just the consequence of a lot of other planning. So this is where I use Palm a lot just to do a sense check." — Lucia, ON

3. Compose multiple assumptions into named event scenarios ✓¶

Desired Outcomes: - Minimize the effort to model multi-faceted business events as coherent scenarios - Increase reusability of individual assumptions across different scenario combinations - Reduce the complexity of understanding combined impacts from multiple simultaneous changes

Sources: ON (2026-02-18, 2026-03-04), Personio (2026-02-18) — Confirmed (3 sources, LEGO-brick model)

"When you change one brick of the Lego... it's easier for us to judge if the change makes sense." — Lucia, ON

4. Define worst/best case scenarios from historical extremes ✓¶

Desired Outcomes: - Minimize the effort to calculate worst-case cash position from historical min/max by category group - Increase the ability to layer multiple severity levels by toggling category groups on/off - Reduce dependency on manual spreadsheet analysis for stress testing

Sources: Personio (2026-02-18), ON (2026-02-18) — Confirmed (2 sources)

"You have to think of the worst, worst, worst case." — Tom, Personio

5. Model the cash impact of timing shifts in payments and collections ✓¶

Desired Outcomes: - Minimize the manual effort to model 'what if payments are delayed by N days' scenarios - Increase the ability to quantify the cash impact of intercompany payment term changes - Reduce the time to model the impact of collection slowdowns on entity-level cash

Sources: ON (2026-03-04), Instacart (2025-07-01) — Confirmed (2 sources)

"Timing shift is actually quite, quite good and quite useful in lots of different cases." — Amanda, ON

6. Save and compare forecast versions to track how assumptions changed ✓¶

Desired Outcomes: - Minimize the difficulty of remembering past forecast assumptions ("What was my opinion in March?") - Increase ability to attribute variance to assumption changes vs actuals variance - Reduce unexplained variance by tracking all assumption changes systematically

Sources: Instacart (2025-07-01), ON (2026-01-22), Sonder (2024-10-03) — Confirmed (3 sources)

"Save a version of the forecast and call this the July 1st forecast... the system could tell you: you changed these three assumptions" — David Watt, Instacart

7. Adjust ML-based forecasts for market disruptors (COVID, tariffs, etc.) ✓¶

Desired Outcomes: - Minimize manual input by using reusable percentage-based adjustments - Increase transparency of adjustment reasons for future analysis - Reduce duplicate work when disruptors are already reflected in historical ML data

Sources: Levi's (2025-12-11), ON (2026-03-04) — Confirmed (2 sources)

"An explanation for the next poor soul who goes into the system and tries to understand what went on there." — Dette, Levi's

8. Determine and defend minimum cash thresholds per entity ⚡¶

Desired Outcomes: - Minimize the time to calculate data-driven cash buffer recommendations per entity - Reduce disputes with regional teams about required cash levels - Increase confidence in minimum cash guidance provided to CFO for M&A capacity decisions - Minimize the gap between perceived buffer needs (regional claims) and actual data-driven requirements

Source: ON (2026-03-04) — Dynamic line derived from forecasted outflows (e.g., 2 weeks of payables), not hardcoded

"You keep telling me that you need two hundred, but what we see is that you would be okay with 50." — Lucia, ON

9. Make quick operational decisions when plans change ⚡¶

Desired Outcomes: - Minimize the time to model a "what if this payment doesn't happen" scenario - Reduce the need to use Excel for ad-hoc calculations - Increase confidence that scenario impact is accurately calculated

Source: ON (2024-11-19) — "If the tool is not flexible or intuitive to make scenarios, you end up doing it in Excel"

10. Track fixed-point scenarios against rolling actuals over time ⚡¶

Desired Outcomes: - Minimize the loss of scenario baselines when forecasts re-baseline to actuals - Increase ability to see whether actuals are tracking best case, baseline, or worst case - Reduce confusion when comparing original scenario assumptions to updated forecasts

Source: ON (2026-02-18) — Jennifer raised, ON team validated

Key Pain Points¶

Full history: pain-points.md

No data-driven minimum cash calculation - Can't calculate and defend entity-level cash buffers; regions overestimate needs, CFO gets no confident guidance (Source: ON 2026-03-04)
Percentage-only adjustments insufficient - Real workflows need timing shifts, absolute values, and composable assumptions alongside percentages (Source: ON 2026-03-04, Personio 2026-02-18)
Manual scenario creation - Creating multiple forecast versions requires duplicating data and manually adjusting values (Source: Personio)
Rolling forecast re-baselines wipe out scenario comparison - Can't compare original scenarios against actuals over time (Source: ON 2026-02-18)
Cannot forecast for new entities with no history - New markets/stores have no ML baseline (Swedish store opening in April) (Source: ON 2026-02-18, 2026-03-04)
No tool for live regional negotiations - Can't run assumptions together in meetings to challenge regional buffer claims (Source: ON 2026-03-04)
TMS inflexibility drives Excel usage - If tool isn't easy for quick scenarios, people go back to Excel (Source: ON)
Overly conservative investment decisions - Without forward visibility, teams default to shorter-term, lower-yield investments (Source: Personio)

Key Opportunities¶

Assumption-based modeling - Allow users to define assumptions (e.g., -5% collections) that automatically adjust forecasts
Scenario comparison views - Side-by-side comparison of base vs. conservative vs. aggressive scenarios
Quick scenario generation - One-click creation of standard scenarios (best/worst/base case)
Investment scenario planning - Monthly planning view showing planned investment activity against forecasted cash positions
Dynamic minimum cash buffer line - Forecast-derived threshold overlaid on charts (e.g., 2 weeks of payables), moves with forecast and scenarios
Timing shift assumptions - Delay/accelerate inflows or outflows by N days, distinct from percentage adjustments
Budget sense-checking - Use direct cash forecast as independent validation of FP&A/controlling's indirect plan
Composable LEGO assumptions - Combine individual assumptions into named event scenarios (store opening, M&A, IC term changes)
Goal-seeking - "What needs to happen to reach X cash target?" for entity-level negotiations
Regional collaboration tool - Run through assumptions live in meetings with regional teams

Open Questions¶

[x] How do teams typically define their scenario assumptions? → Percentage-based adjustments on categories (Personio, ON validated) + one-off manual assumptions (Tom at Personio)
[x] Do teams need to scenario-plan at the category level or overall forecast level? → Both — category-level adjustments compose into scenario-level views (ON 2026-02-18)
[x] Bottom-up or top-down? → Bottom-up (entity-level) preferred for validation, then aggregate to group (ON 2026-03-04, Yulia)
[ ] What's the typical number of scenarios teams maintain?
[ ] How should fixed-point scenarios interact with rolling forecasts? (snapshot vs continuous overlay)
[ ] How to handle scenario modelling for entities with no historical data? (manual input, template from similar entity?)
[ ] How should minimum cash thresholds be defined? (formula-based? per-entity configurable? based on payables, revenue, or custom?)
[ ] What's the right cadence for scenario review — ad-hoc only, or regular (weekly/monthly)?

Last updated: 2026-03-10 | Sources: 12 transcripts (view all)