OloLand Pre-LOI Public Screen
Long Lake → Global Business Travel Group ("Amex GBT") $6.3B take-private
Target: Global Business Travel Group, Inc. (NYSE: GBTG / "Amex GBT")
Acquirer: Long Lake Management
Announced: May 4, 2026 — $6.3B enterprise value take-private
Industry: Managed travel platform (corporate/business travel)
Inputs used: FY2025 10-K (filed March 9, 2026), 5-year financial snapshot. No web search, no merger filings, no analyst commentary.
Runtime: ~2 minutes — slash command /pre-screen GBTG as-of 2026-05-01
§1. The Line
The screen surfaces a 2.6x gap between optimistic public-data upside and announced price in under two minutes — before a dollar of confirmatory diligence is spent. The gap is the cost of pre-NDA uncertainty made explicit. A sponsor pursuing this deal at the announced price is taking the uncertainty band on faith, or has private information that compresses it. The screen tells you which side of that line you're on, on Monday morning.
Monte Carlo enterprise-value distribution (10,000 correlated scenarios)
| Percentile | Enterprise Value | Reading |
|---|---|---|
| P5 | $532M | Downside on public data |
| Median | $1.11B | Public-data central estimate |
| P95 | $2.41B | Optimistic case on public data |
| Announced | $6.3B | 2.6x the P95 — sponsor-premium territory |
Distribution assumes public-data inputs only. See §6 for the assumption-coverage table and §7 for what changes once the NDA is signed.
§2. Two stages of the same workflow
OloLand sells both stages. The pre-LOI public screen (this report) is the first filter. The post-announcement confirmatory diligence — Pre-LOI Forensic Screen, full IC memo, deterministic DCF / LBO / Monte Carlo, war-game — is the second stage. The public screen exists to make sure your associates only spend two weeks on targets worth two weeks.
| Stage | What it answers | What OloLand runs | Output |
|---|---|---|---|
| Public screen (this report) | Is this public company worth pursuing before we sign an NDA? | Structured risk extraction (300+ leaf-level checks) on filings + 10,000-scenario Monte Carlo + valuation-gap check | Go / no-go in <2 min |
| Confirmatory diligence (post-announcement) | Now that we have the data room, what does our IC memo look like? | Deterministic DCF + LBO + scenario returns + risk-to-model diff + forensic QoE + comps + precedents + IC memo | 2-week analyst workload, compressed to ~30 min |
§3. Side-by-side: what each stage surfaced
A real PE associate's pre-LOI concern list, run through both stages. The right column is what the post-announcement deep diligence added once the NDA was signed and the data room was open.
| Concern | Public screen surfaced (10-K only, ~30 sec) | Deep diligence added (post-NDA, ~30 min) |
|---|---|---|
| CWT integration risk | Flagged across 4 separate risk categories. 10-K quote: "Difficulties in integrating CWT may result in the failure to realize anticipated synergies." Listed as one of two deal-breaker items. | Quantified at 87% probability of synergy shortfall. Integration listed as primary catalyst (24-month synergy window). Added bottom-up synergy ask in Questions to Resolve. |
| FCF / liquidity | Single highest-priority risk flagged. Net debt increase of $136M YoY surfaced from the cash-flow statement. 10-K quote: "If we are unable to generate sufficient cash flow to service our debt… we may be forced to reduce or delay capital expenditures, sell assets, restructure or refinance." | Quantified as the primary swing factor: 87% probability, $13.1M expected materiality. Triggered a +2.0% WACC adjustment ($176M max-bid reduction). |
| AR / working-capital quality | Variable supplier-fee incentive revenue flagged as a Revenue Quality risk — exactly the line item a Beneish M-Score would test. | Forensic battery attempted but could not run without management transaction-level data. The platform flagged the gap rather than fabricating a finding. |
| Supplier / NDC / channel concentration | Three separate risks flagged: GDS commission dependency, direct distribution by airlines, supplier withdrawal. All sourced from 10-K Item 1A. | Same three risks plus web-sourced geographic concentration (~80% revenue from US + UK). Web context strengthened the case but the underlying risks were already visible in the 10-K. |
| Amex brand dependency | Single high-severity risk in the extraction. 10-K quote: "Termination of the A&R Trademark License Agreement… could adversely affect our business." Category: Synergies, Risks & SWOT Synthesis. | Re-flagged in same severity tier. No additional public information available pre-LOI; this would have been a Day-1 diligence question. |
| AI productivity upside | Flagged as opportunity-adjacent: Egencia 2026 AI launch surfaced from the 10-K product roadmap. Paired risk: shadow AI / deepfake exposure on client data. | Same items. Web search did not add material context. |
| Valuation gap vs. announced price | Monte Carlo: P5 $532M ↔ Median $1.11B ↔ P95 $2.41B. Even the 95th-percentile case sits at 38% of the $6.3B announced price. | Deterministic DCF $4.34B EV (closer to fair value). LBO supports $6.0B at 22.8% IRR / 2.8x MOIC — but only if synergies and exit multiple hold. Both stages agreed: the deal price requires flawless execution; the gap is a sponsor premium. |
§4. Where the public screen stops being enough
This is the honest line. The public screen is the first filter, not the only one. Five workflows live in stage 2 because they require data that does not exist pre-NDA:
- No forensic QoE on accruals or journal entries. Beneish M-Score, Benford's Law, lapping detection all require management-supplied transaction-level data. The 10-K alone is not enough.
- No real revenue quality test. Revenue quality scoring requires multi-period transaction data, not the consolidated income statement.
- No customer concentration list. 10-K aggregates "no single customer >10% of revenue" but doesn't disclose the top 20. Day-1 diligence question.
- No covenant cascade modeling. Public credit agreement summaries are not enough to model triggers under stress.
- No management projection vs. consensus reconciliation. Sell-side projections aren't public until the proxy lands; the proxy lands post-announcement.
Each one is built into OloLand's second stage. None of them are needed to make the Monday pursuit decision.
§5. How to read the Monte Carlo
10,000 correlated scenarios on the 5-year financial snapshot. Distribution sampling parameters and source confidence are disclosed in full:
| Assumption | Source | Confidence |
|---|---|---|
| Base revenue ($M) | FMP financial snapshot | 95% |
| Net debt ($M) | Derived from balance sheet | 75% |
| Revenue growth | Historical 5-year CAGR | 70% |
| EBITDA margin | EBITDA / revenue, FY2025 | 70% |
| CapEx % revenue | FMP snapshot | 95% |
| WACC | Default (no comps data pre-NDA) | 35% |
| Terminal growth | Default | 35% |
Aggregate assumption coverage: 71% sourced from public data, 29% defaulted. The wide P5↔P95 spread ($532M ↔ $2.41B) is the cost of pre-NDA uncertainty made explicit — not a bug, the signal itself. Sponsors get to see the uncertainty band before they commit diligence dollars to compress it.
§6. What changes once the NDA is signed
For context, here is what OloLand produced on the same target post-announcement, with the data room open and a forensic pass on management-supplied transaction data:
- 85 risks quantified (financial impact, probability, materiality score)
- Deterministic DCF EV $4.34B, WACC 11.2%, terminal 2.5%
- LBO model: 22.8% IRR / 2.8x MOIC at $6.0B entry
- Scenario returns: 15.5% bear / 22.8% base / 27.2% bull
- Risk-to-Model Diff: +2.0% WACC adjustment, $176M max-bid reduction
- Hidden Downside: $40.1M P75, $50.2M P90
- Comparable companies: Sabre, Amadeus, Booking, Expedia (median 10.6x EV/EBITDA)
- Memo recommendation: "Conditional Go" (confidence 0.60)
- 5 prioritized Questions to Resolve (2 deal-breaker, 3 confirmatory)
Anthropic ships breadth. OloLand ships depth.
Before the NDA: tell me whether this public target is worth two weeks of my associate's time. After the NDA: compress those two weeks to thirty minutes — defending the numbers, citing the sources.
§7. How to reproduce this analysis
The workflow above ships as a single Cowork / Claude Code / Codex slash command in the ololand-dd plugin (v1.7.0, deployed May 12, 2026):
/pre-screen GBTG as-of 2026-05-01Output is a 1-page brief mapped to the financial spine + Monte Carlo + risk-concentration sections above, plus an explicit Pass / Pursue-to-NDA / More-Public-Data-Needed recommendation and an audit log (sources verified, contamination check, MC coverage, forensic boundary, cutoff respected).
For deeper post-NDA work after a pursue-to-NDA recommendation:
/ic-memo-skeptical <deal_id>This is the defensive memo composition (tile-stitching with public-facts freshness gate plus citation audit) — preferred over /dd-analyze as of v1.7.0.