debt_leverage_profile
Moderate leverage; net debt/EBITDA ~2.0-2.5x; 20% rate rise adds ~R$15-25M annual interest expense, compressing EBITDA margin ~2-3pp
Inferred
Agent_Inference
interest_rate_sensitivity
Floating-rate BRL debt exposure means 20% Selic rise materially increases cost of capital; refinancing risk moderate given regulated cash flows
Inferred
Agent_Inference
geopolitical_supply_exposure
High intensity; European gas dependency on Russia exposed structural energy security vulnerabilities.
Medium
GICS-commodity-overlay-v1
supply_chain_dependency
Bolivia-Brazil gas pipeline (GASBOL) and Petrobras TBG transmission network are top-two chokepoints controlling ~90% of feedstock supply
Inferred
Agent_Inference
international_expansion_readiness
No material international revenue markets; Compagas operates solely in Paraná state, Brazil — sovereign FX devaluation exposure is nil
Inferred
Agent_Inference
geographic_footprint
100% Brazil (Paraná state); zero international revenue; BRL/USD devaluation affects import costs for equipment but not revenue currency mix
Inferred
Agent_Inference
commodity_exposure_profile
High intensity; commodities: Natural Gas, Coal, Uranium, Crude Oil, Copper (grid), Lithium (storage); geopolitical: European gas dependency on Russia exposed structural energy security vulnerabilities.
Medium
GICS-commodity-overlay-v1
vendor_lock_dependency_score
Petrobras/TBG represent >80% of gas supply input cost; single non-substitutable supplier creating critical vendor lock dependency
Inferred
Agent_Inference
business_model_type_primary
Regulated utility/infrastructure; no meaningful cloud dependency — operational systems on-premise or via local providers; 30-day cloud termination impact is minimal
Inferred
Agent_Inference
business_model_type_secondary
Secondary billing and ERP systems may face disruption but legacy infrastructure-based operations would continue with limited cloud reliance
Inferred
Agent_Inference
switching_cost_profile
Minimal API coupling risk; utility operations rely on SCADA/OT systems, not third-party APIs; switching cost for customers is high due to pipeline infrastructure lock-in
Inferred
Agent_Inference
howey_test_risk_index
Low Howey Test risk; revenue from regulated gas distribution tariffs — not an investment contract; no token or passive-return structure
Inferred
Agent_Inference
regulatory_burden_tier
Very High
Medium
GICS-regulatory-overlay-v1
data_sovereignty_risk
Minimal GDPR exposure (no EU operations); LGPD (Brazil) applies to customer billing data; compliance costs low, estimated <R$5M annually
Inferred
Agent_Inference
antitrust_exposure_flag
High antitrust scrutiny as state-concession monopoly in Paraná; CADE monitors tariff-setting and exclusivity; concession terms constrain abuse risk
Inferred
Agent_Inference
regulatory_exposure_profile
Very High burden; regimes: FERC, NERC, EPA, NRC, State PUCs, DOE; Rate-case lag and clean-energy mandates compress returns on regulated asset base.
Medium
GICS-regulatory-overlay-v1
revenue_model_type
~85% recurring via regulated long-term concession contracts and volume tariffs; ~15% transactional (connections, services); highly predictable revenue base
Inferred
Agent_Inference
monetization_vector
Volume-based gas tariff per m³ distributed across industrial, commercial, residential, and vehicle (GNV) segments under regulated pricing framework
Inferred
Agent_Inference
pricing_architecture
Tariff set by AGEPAR regulator; stress scenario of 15% demand drop offset by pass-through clauses; margin protected but volume risk is real
Inferred
Agent_Inference
pricing_power_rating
Low independent pricing power; tariffs regulated by AGEPAR with periodic reviews; inflation pass-through partially allowed via adjustment indices
Inferred
Agent_Inference
target_gross_margin_bracket
Estimated gross margin 25-35%; regulated spread between gas purchase cost and distribution tariff compressed by Petrobras pricing and regulatory caps
Inferred
Agent_Inference
churn_vulnerability_index
Low free-rider risk; metered utility with contractual offtake; industrial customers have long-term contracts; residential churn limited by infrastructure lock-in
Inferred
Agent_Inference
headcount_cost_structure
Revenue growth is sublinear to headcount; pipeline network scales without proportional staff increases; doubling throughput requires <20% headcount growth
Inferred
Agent_Inference
marginal_cost_of_growth
Marginal cost of growth primarily capex (network expansion), not opex; incremental distribution is near-zero variable cost once pipeline infrastructure is in place
Inferred
Agent_Inference
franchise_compliance_risk
Not a franchise model; state concession granted by Paraná government; compliance drift risk is regulatory non-renewal risk, not franchise network drift
Inferred
Agent_Inference
customer_acquisition_metric
At 10x scale, CAC rises significantly due to geographic limits of Paraná concession; network extension cost per new customer increases as dense areas saturate
Inferred
Agent_Inference
network_effect_present
Weak network effects; pipeline utility has density economics not true network effects; value does not increase with more users in a platform sense
Inferred
Agent_Inference
asset_efficiency_ratio
AI displacement risk is low; physical gas distribution infrastructure cannot be AI-displaced; predictive maintenance AI may improve asset efficiency ~5-10%
Inferred
Agent_Inference
recession_resistance_tier
Tier 2 recession resilience; industrial gas demand falls in recession (~20-30% volume risk) but residential/CNG partially offsets; regulated tariffs provide floor
Inferred
Agent_Inference
customer_segment_primary
Industrial customers (automotive, food processing, ceramics) representing ~50-60% of volume; high concentration risk from top 10 industrial accounts
Inferred
Agent_Inference
customer_segment_secondary
Commercial and residential consumers plus CNG vehicle fueling stations; fragmented, lower concentration risk but lower margin contribution
Inferred
Agent_Inference
characteristic_occupations
["11-0000 Management Occupations", "13-0000 Business and Financial Operations Occupations", "15-0000 Computer and Mathematical Occupations", "17-0000 Architecture and Engineering Occupations", "23-0000 Legal Occupations", "41-0000 Sales and Related Occupations", "43-0000 Office and Administrative Support Occupations", "47-0000 Construction and Extraction Occupations", "49-0000 Installation, Maintenance, and Repair Occupations"]
High
SOC-2018/GICS-overlay
agent_automatable_labor_share
0.34 (HIL — ~34% of characteristic roles agent-automatable)
Medium
SOC-2018 + agentic-exposure-v1
capital_expenditure_profile
Capex oriented toward network extension into new municipalities and pipeline maintenance; limited reallocation to future-state digital infrastructure; legacy-heavy model
Inferred
Agent_Inference
sec_cik
null
Inferred
Agent_Inference
ticker
CGAS3/CGAS5 (B3); trades at discount reflecting Petrobras supply concentration risk, AGEPAR tariff compression, and illiquidity premium of ~15-20% vs. regulated peers
Inferred
Agent_Inference