Lightning Network Economics — Fee Market, Channel Economics, LSP Competition

Section 1

LN Fee Structure Overview

Lightning Network routing fees have two components: a flat base fee and an amount-proportional fee. Every hop along a route charges both.

Fee Components

Parameter	Default	Unit
base_fee_msat	1000	msat (= 1 sat)
fee_rate_millionths	1	ppm (0.0001%)

fee = base_fee + amount × fee_rate / 1,000,000

fee_rate is in millionths (ppm). 1 ppm = 0.0001% = 1 sat per 1,000,000 sats sent.

Crossover Point

For small payments the flat base fee dominates. For large payments the proportional fee dominates. The crossover is when both contribute equally:

crossover = base_fee / fee_rate_ppm × 1,000,000

At defaults (1000 msat base, 1 ppm): crossover = 1,000,000,000 msat = 1,000 sats. Below 1k sats → base fee dominates. Above 1k sats → proportional dominates.

Key insight: For small payments (e.g. 100 sats), a 1-sat base fee is already 1% — extremely expensive. For large payments (e.g. 1M sats), 1 ppm adds only 1 sat but the base fee is 1 sat too, so total is 2 sats = 0.0002%. This asymmetry shapes routing strategy.

Interactive Fee Calculator

Base fee (msat) 500 msat

Fee rate (ppm) 100 ppm

Payment amount (sats) 1,000 sats

—

Total fee (msat)

—

Total fee (sats)

—

% of amount

—

Crossover (sats)

Fee breakdown

Fee vs payment amount (log scale)

Section 2

Channel Profitability Calculator

Full economics model for a routing node operator: revenue from routing fees vs. on-chain costs, opportunity cost, and rebalancing expenses.

Revenue Side

Payments routed/day 50

Avg payment (sats) 100,000

Fee rate (ppm) 500

Base fee (msat) 1000

Cost Side

Open tx fee (sat/vB) 20

Channel size (k sats) 1,000k

Opportunity cost (APY%) 3.0%

Channel lifetime (days) 365 days

—

Daily revenue (sats)

—

Annual revenue (sats)

—

Total channel cost (sats)

—

Break-even (days)

—

ROI over lifetime

Calculating…

Break-even timeline — cumulative profit vs cost

Section 3

Fee Market Dynamics

How routing fees reach equilibrium through supply (liquidity) and demand (payment flow). Path selection tradeoffs between cost and reliability.

Scenario: Alice sends 100k sats from A → B

50 sats

Path 1

A → X → B · 2 hops · Low liquidity, often fails near peak · Fee rate: 0.5 ppm

Low reliability

120 sats

Path 2

A → M → N → B · 3 hops · Well-funded, reliable routing · Fee rate: 1.2 ppm

Medium reliability

200 sats

Path 3

A → HUB → B · 2 hops · Hub node with large reserves · Fee rate: 2 ppm

High reliability

Tradeoff: Path 1 saves 150 sats but risks failed attempts (costing time and HTLC slots). Pathfinding algorithms weight probability of success — reliability premium is economically rational.

Supply-Demand Equilibrium

Supply curve: liquidity available at each fee level. Demand curve: volume of payments willing to pay that fee. Adjust market conditions below.

Network liquidity (shift supply) 0

Payment demand (shift demand) 0

—

Eq. fee (ppm)

—

Eq. volume

Market equilibrium: fee ≈ opportunity cost of liquidity. If staking yields 5% APY, nodes demand routing fees equivalent to that return. Nodes charging below opportunity cost operate at a loss; nodes charging above get routed around.

Section 4

LSP (Lightning Service Provider) Economics

LSPs are businesses providing liquidity and channel services. They earn through channel fees, routing revenue, and premium services.

Inbound Liquidity

LSP opens a channel to the user, granting them receiving capacity. User pays a one-time channel fee proportional to channel size.

fee = channel_fee_rate × size

Zero-Conf Channels

Instant channels opened before on-chain confirmation. Higher risk for LSP (double-spend window) → higher fee. User gets immediate liquidity.

JIT Channels (LSPS2)

LSP intercepts incoming payment, opens channel just-in-time, routes the payment through. Fee wrapped in the payment itself (wrapped invoice).

LSP Revenue Model — Interactive P&L

Channels opened/month 100

Avg channel size (k sats) 1,000k

Channel fee rate (ppm) 3,000

Routing fee/channel/mo (sats) 500

On-chain cost/open (sats) 3,000

Capital cost APY (%) 5.0%

Ops cost/month (k sats) 500k

Monthly P&L Statement

Channel opening revenue—

Routing revenue—

Total Revenue—

On-chain tx costs—

Capital opportunity cost—

Operations—

Total Costs—

Net Profit—

Margin—

Section 5

Liquidity Management & Rebalancing Economics

When channels drain, nodes must rebalance. Each strategy has different costs and break-even characteristics.

The Rebalancing Dilemma

Alice→Bob channel (drained)

Alice: 5k satsBob: 95k sats

Alice cannot route payments — all liquidity is on Bob's side. She needs to rebalance to recover routing revenue.

Imbalance (%) 90%

Daily payments missed 30

Fee earned per payment (sats) 50

Strategy Comparison

Circular Rebalance Simulator

Alice sends 30k sats in a circular route: Alice→Carol→Bob→Alice, restoring balance without touching the on-chain layer.

Alice↔Bob

5k (Alice)95k (Bob)

Bob↔Carol

50k (Bob)50k (Carol)

Carol↔Alice

95k (Carol)5k (Alice)

Press Start to begin circular rebalance animation.

Cost: ~50 sats in routing fees | Benefit: restored routing capacity

Section 6

Economic Attacks & Griefing

Fee incentives can be exploited. Understanding economic attack vectors is essential for robust node operation and protocol design.

HTLC Jamming (Economic Griefing)

Attacker holds HTLCs without settling. Victim's slots (max 483 HTLCs) and liquidity are consumed without the attacker paying fees (failed payments are free).

Scenario

Attacker / Victim Payoff

Slow jam (hold for hours)
Hold 483 HTLCs, victim earns nothing

Attacker: 0 cost / Victim: −routing revenue

Fast jam (spam then fail)
Flood with tiny payments, all fail

Attacker: 0 cost / Victim: −opportunity cost

With upfront fees (proposed)
Attacker pays per attempt

Attacker: −jam cost / Victim: +upfront fee

Root cause: Routing fees are only collected on successful payments. Failed attempts cost the attacker nothing. Upfront fees (even 1 msat) change the economics fundamentally — the attacker must pay proportional to the damage caused.

HTLC hold time (hours) 24h

Victim daily routing (sats) 10,000

—

Victim's loss (sats)

Attacker cost (no upfront)

—

Cost w/ 1 msat upfront

Routing Fee Sniping

Attacker preferentially routes through channels it plans to close, collecting fees just before the channel closes cooperatively. The final state is on-chain, so routing revenue is captured but the long-term relationship is abandoned.

Phase

Economic Effect

Pre-close burst
Attacker routes maximum volume through target channel

Attacker: +fees / Victim: +fees (unaware)

Cooperative close
Channel closes, attacker moves on

Net: channel destroyed, short-term gain

Defense: age-based fees
Young channels (few days old) charge higher fees; established nodes earn discount

Sniping cost ↑ / Long-term nodes rewarded

Fee Market Manipulation

Large hub nodes control many routes. They can extract monopoly rents if alternative paths are insufficient.

Hub market share	Pricing power	User impact
< 10%	Minimal	Competitive fees
10–30%	Moderate	Mild premium over competitive rate
30–60%	Significant	Toll road dynamics emerge
> 60%	Dominant	Near-monopoly, fee extraction

Counter-pressure: Competing routes incentivize node operators to undercut dominant hubs. Multi-path payments (MPP) split payments to avoid any single hub bottleneck. Channel factories reduce the cost of establishing diverse routes.

Section 7

Channel Economics — Research Insights

Key findings from Guasoni et al. (Mathematical Theory of PCN, 2024) and related work on optimal fee setting and channel sizing.

Optimal Fee Formula (Guasoni 2024)

fee* ≈ σ² × T / (2 × vol)

Variable	Meaning
σ	Payment flow volatility
T	Time horizon
vol	Channel volume (total flow)

Insight: Higher volatility in payment flow → higher optimal fees. This reflects the cost of balance uncertainty and the need to compensate for rebalancing risk.

Optimal Channel Size

C* = sqrt(2 × V_annual × capital_cost / rebalance_cost)

Too small: drains frequently → costly rebalancing erodes margin
Too large: capital locked inefficiently → low ROI per sat
Optimal: balances rebalancing cost against capital opportunity cost

Annual volume (M sats) 100M

Rebalance cost (sats) 200

—

Optimal size (k sats)

Balance Random Walk Simulation

Channel balance evolves as a random walk. Mean reversion occurs when payments are symmetric. Drift occurs when systematic imbalance exists (e.g., always buying goods, never selling).

Payment volatility (σ) 10

Drift (% per step) 0%

Mean-reversion strength 20%

Mean reversion: If payments flow symmetrically in both directions, the balance naturally returns toward 50/50. Strong mean-reversion (high setting) → stable channel needing less rebalancing. Zero mean-reversion + drift → systematic draining, requiring active management.

Section 8

Network-Level Economics

Macro perspective on LN's economic structure, fee distribution inequality, and why centralisation has systemic consequences.

~5,000

BTC locked (≈$500M+)

~50,000

Active channels (2025 est.)

~10,000

Active routing nodes

Fee Distribution (Lorenz Curve)

Routing fee income is highly unequal — hub nodes capture a disproportionate share.

Gini ≈ 0.72

0 = perfect equality · 1 = one node earns all fees

Estimated Fee Share by Node Tier

Node tier	% of nodes	% of fees
Top 10 nodes	0.1%	~40%
Top 100 nodes	1%	~75%
Top 1,000 nodes	10%	~95%
Long tail (9k+)	89%	~5%

Why Centralisation Matters

Fragility

Failure of a top-10 hub disconnects a large fraction of nodes, fragmenting the network. Single points of failure are inherent.

Fee Extraction

Dominant hubs gain pricing power. Users with few alternative routes pay above competitive rates — creating "toll road" dynamics.

Privacy

Hub nodes observe a large fraction of all payments. Even with onion routing, traffic analysis at hubs is more effective than at leaf nodes.

Counterforces: Channel factories, trampoline routing, and MPP all reduce dependence on large hubs. LSPs can serve the long tail without requiring global connectivity of every node. But the economic incentive to become a hub (and extract rents) remains strong.