debt_leverage_profile
Net debt ~NZD 1.8B; net debt/EBITDA ~1.8x; a 20% interest rate rise adds ~NZD 18-25M annual interest cost on floating-rate tranches
Inferred
Agent_Inference
interest_rate_sensitivity
Approximately 30-40% of debt is floating-rate; 20% rate increase compresses NPAT by roughly NZD 15-20M, ~5-8% earnings impact
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
1) Taiwan Strait (TSMC-dependent network hardware); 2) US-China tech export controls on Huawei/ZTE infrastructure components
Inferred
Agent_Inference
international_expansion_readiness
Revenue is ~95% NZD-denominated; minimal sovereign currency devaluation exposure; international markets (Pacific islands) are small and USD-linked
Inferred
Agent_Inference
geographic_footprint
Predominantly domestic NZ market; limited Pacific operations; negligible multi-currency exposure; AUD exposure via minor trans-Tasman services
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
High dependency on AWS for cloud services and Ericsson/Nokia for network infrastructure; Ericsson likely exceeds 30% of RAN capex input
Inferred
Agent_Inference
business_model_type_primary
AWS termination would critically disrupt Spark's cloud-resale and managed IT services division (Spark Digital); recovery timeline 6-12 months minimum
Inferred
Agent_Inference
business_model_type_secondary
Secondary impact: customer-facing digital platforms and billing systems hosted on cloud would face outage risk, damaging ~NZD 200M+ managed services revenue
Inferred
Agent_Inference
switching_cost_profile
Moderate-high API coupling to AWS; Spark Digital's managed services are architected around AWS tooling; multi-cloud strategy partially mitigates but not eliminates risk
Inferred
Agent_Inference
howey_test_risk_index
Primary revenue (telco subscriptions, data services) fails Howey Test; no expectation of profit from others' efforts; securities classification risk is negligible
Inferred
Agent_Inference
regulatory_burden_tier
Very High
Medium
GICS-regulatory-overlay-v1
data_sovereignty_risk
Subject to NZ Privacy Act 2020 (GDPR-equivalent); CCPA not directly applicable; key risk is cross-border data flows for Pacific/cloud services; moderate compliance cost
Inferred
Agent_Inference
antitrust_exposure_flag
Moderate; Commerce Commission scrutiny on wholesale broadband access pricing; duopoly dynamics with One NZ (Vodafone); regulated via Telecommunications Act 2001
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
Estimated ~75-80% recurring (mobile plans, broadband subscriptions, long-term IT contracts); ~20-25% transactional (hardware, project services)
Inferred
Agent_Inference
monetization_vector
Subscription-led (mobile ARPU, broadband monthly fees) plus B2B managed services contracts; hardware bundling as acquisition tool, not primary monetization
Inferred
Agent_Inference
pricing_architecture
Pricing anchored to regulated wholesale inputs; limited upward flexibility on broadband; mobile ARPU under pressure from competition; IT services margin more defensible
Inferred
Agent_Inference
pricing_power_rating
Low-to-moderate; regulated broadband limits pass-through; mobile market competitive; enterprise IT services offer moderate pricing power (3-4/10 overall)
Inferred
Agent_Inference
target_gross_margin_bracket
Group gross margin ~40-45%; IT services higher (~50%+); network services lower (~35%); hardware drag pulls blended margin down
Inferred
Agent_Inference
churn_vulnerability_index
Low free-rider risk; services are metered/authenticated; churn risk concentrated in mobile (~15-18% annual mobile churn estimated in competitive market)
Inferred
Agent_Inference
headcount_cost_structure
Sublinear growth model; automation and network investment enable revenue scaling without proportional headcount; IT services partially headcount-linear
Inferred
Agent_Inference
marginal_cost_of_growth
Marginal cost of adding mobile/broadband subscribers is low post-network build; incremental EBITDA margin on new subscribers estimated 50-60%
Inferred
Agent_Inference
franchise_compliance_risk
Not a franchise model; retail stores are company-owned or dealer-agent model; agent compliance drift is manageable and not a material risk
Inferred
Agent_Inference
customer_acquisition_metric
At 10x scale (not realistic domestically); CAC efficiency would plateau given NZ's 5M population cap; market saturation limits unit economics improvement
Inferred
Agent_Inference
network_effect_present
Weak direct network effects; value derives from network coverage quality, not user-to-user interaction; some indirect effects via app ecosystem partnerships
Inferred
Agent_Inference
asset_efficiency_ratio
AI can displace customer service (call centre ~500-800 FTEs at risk), network operations, and IT support roles; 15-20% headcount reduction feasible over 5 years
Inferred
Agent_Inference
recession_resistance_tier
Tier 2 (resilient); mobile and broadband are near-essential services; enterprise IT spending more cyclical; revenue decline in recession likely limited to 3-6%
Inferred
Agent_Inference
customer_segment_primary
Consumer (mobile + broadband): ~55% of revenue; low individual concentration risk but exposed to macro consumer spending pressure
Inferred
Agent_Inference
customer_segment_secondary
Enterprise/SME (Spark Digital IT services + connectivity): ~40% revenue; some customer concentration risk with large government contracts (IRD, MSD)
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 shifting from copper/legacy to 5G rollout and fibre; annual capex ~NZD 400-450M; reallocation toward future-state (5G SA, cloud-native) is underway
Inferred
Agent_Inference
sec_cik
null
Inferred
Agent_Inference
ticker
SPK.NX (NZX) / SPKKY (OTC); currently trading near fair value (~12-13x EV/EBITDA); geopolitical hardware supply risk not fully priced but partially offset by regulated revenue stability
Inferred
Agent_Inference