{
 "episode_id": "2026-10-17-special",
 "mode": "special",
 "published": "2026-10-17T20:07:05+00:00",
 "video_url": "https://youtu.be/axbBnImgAVA",
 "title": "Valuing a network: does Metcalfe's law fit Cardano?",
 "data_generated_at": "2026-10-17T18:38:47+00:00",
 "claim": null,
 "market": {
  "ada_usd": 0.2220431954534615,
  "source": "coingecko",
  "cross_checked": true,
  "change_24h_pct": -0.2212026271879902
 },
 "indigo": {
  "tvl_usd": 3786893.0,
  "indy_price_usd": 0.110203
 },
 "loan_book": {
  "positions": 284,
  "debt_usd": 1779791.712978517,
  "collateral_ada": 16798033.239705,
  "book_ratio": 2.095688473363839,
  "price_10pct_at_risk_usd": 0.17228419751822643,
  "price_50pct_at_risk_usd": 0.14337885992861663,
  "decoded_utxos": 310,
  "validator_utxos": 313
 },
 "coverage": {
  "iADA": {
   "decoded_iasset": 233.450484,
   "chain_supply_iasset": 233.450484,
   "coverage_pct": 100.0
  },
  "iBTC": {
   "decoded_iasset": 5.057821000000001,
   "chain_supply_iasset": 5.057821,
   "coverage_pct": 100.00000000000003
  },
  "iETH": {
   "decoded_iasset": 35.213292,
   "chain_supply_iasset": 35.213292,
   "coverage_pct": 100.0
  },
  "iEUR": {
   "decoded_iasset": 68.387277,
   "chain_supply_iasset": 68.387277,
   "coverage_pct": 100.0
  },
  "iJPY": {
   "decoded_iasset": 76826.95787400001,
   "chain_supply_iasset": 76826.957874,
   "coverage_pct": 100.00000000000003
  },
  "iSOL": {
   "decoded_iasset": 83.681705,
   "chain_supply_iasset": 83.681705,
   "coverage_pct": 100.0
  },
  "iUSD": {
   "decoded_iasset": 1290690.0400220002,
   "chain_supply_iasset": 1362873.88716,
   "coverage_pct": 94.70355637318588
  },
  "_worst_pct": 94.70355637318588
 },
 "governance": {
  "epoch": 662,
  "open_actions": [
   {
    "id": 159,
    "type": "InfoAction",
    "withdrawal_ada": null,
    "expiration_epoch": 663,
    "drep_yes_pct_of_decisive": 18.82661838558432,
    "drep_participation_pct": 51.15203903445342
   },
   {
    "id": 160,
    "type": "TreasuryWithdrawals",
    "withdrawal_ada": 22000000.0,
    "expiration_epoch": 668,
    "drep_yes_pct_of_decisive": 7.489574849810111,
    "drep_participation_pct": 3.144197274055967
   }
  ],
  "decided_counts": {
   "enacted": 1,
   "expired": 6,
   "dropped": 0
  }
 },
 "monetary": {
  "decay_per_epoch_pct": 0.16148699432226854,
  "half_life_years": 5.871063650267911,
  "treasury_accrual_per_epoch_ada": 3670153.7411179543
 },
 "transcript": [
  {
   "heading": "Cold open",
   "text": "Every coin has a market cap: one number, quoted everywhere. Analysts have tried to explain that number with network effects, with fees, and with what it costs to attack a chain. Here are those models, tested on Cardano's own record. No price targets, and no verdicts."
  },
  {
   "heading": "What a market cap is",
   "text": "Start with the number itself. A market cap is the price of one coin, times the coins in circulation. For ADA that is 37.55 billion ADA in wallets and unclaimed rewards, measured from this channel's copy of the chain, at 22.3 cents each, as CoinGecko reports it today. About 8.38 billion dollars. That is a price, not a value. It tells you what the last buyer paid, not why. So analysts reach for models. We will test three, each on Cardano's own record: network effects, fees, and the cost of an attack."
  },
  {
   "heading": "The network effect, as a formula",
   "text": "The most famous model is Metcalfe's law, named after Robert Metcalfe, who co-invented Ethernet. It says a network's value grows with the square of its users. Two phones make one possible call. Five phones make ten. Twelve phones make sixty-six. Each new user can reach everyone already there. In 2006, the mathematician Andrew Odlyzko and colleagues published a correction: most of those connections are never used, so value grows more slowly, closer to n times the logarithm of n. Crypto analysts have applied both to blockchains. To test them, you first need a number of users."
  },
  {
   "heading": "Who counts as a user?",
   "text": "And a blockchain does not have one. You have to choose. Count delegators, and Cardano has 1,303,540 stake addresses delegating to a pool. Count activity, and in the last complete epoch, five days, 29,593 stake keys received a payment, across 52,761 addresses. A stake key is closer to a wallet than an address is, and neither is a person. One person can hold many. The choice of n changes everything that follows."
  },
  {
   "heading": "Five years, tested",
   "text": "So we tested it, one point every thirty days, from November 2021 to October 2026. Over that time ADA's market cap fell 85.7 percent. Delegators went the other way: up 43.5 percent. Counted that way, users grew while the market cap shrank, the opposite of what the law predicts. Active stake keys fell 80.5 percent, so they did move with the market cap. But the square law makes the swings far too big. The power that fits best is about 0.5, not two. The gentler n log n law does little better. And the constant in the formula has to be re-fitted as you go. Fitted on the first half of the record, it is off by 5.0 times on the second half. On Cardano's record, the network effect is a story, not a formula. So if counting users does not explain the market cap, maybe following the money does."
  },
  {
   "heading": "What the network earns",
   "text": "The second model treats a blockchain like a business, and looks at what it earns: transaction fees. On Cardano, the fee is set by a rule, not an auction. 44 lovelace per byte, plus a fixed 0.155381 ADA, plus a price for any smart-contract work. Leaving out floods of transactions from a handful of wallets, which we come to in a moment, the last complete epoch's fees came to 51.4 thousand ADA, from 163,494 transactions: about 7.0 cents per transaction at today's price. Over the last year, 3.1 million ADA, or 947.7 thousand dollars at each epoch's price. That leaves out 414.1 thousand ADA paid by floods of transactions from a handful of wallets. Most of it, 309.2 thousand ADA, came from a group of 13 wallets starting October 6. On October 8, they made 94 percent of all transactions on the chain. The rest came from 8 smaller floods during the year. Those fees were real, so the network collected 3.5 million ADA in all. But a flood from a handful of wallets is not demand, so the figures that follow leave it out. In a business, earnings pay the owners. On Cardano, the owners who get paid are stakers. So where does their pay come from?"
  },
  {
   "heading": "Who pays the stakers?",
   "text": "Every epoch, Cardano pays them from a pot, and the pot has two sources. Over the last year, 760.9 million ADA came out of the reserve, and 3.1 million came from fees. So fees were 0.40 percent of the pot. 20 percent of the pot goes to the treasury first, and the rest is paid to pools and their delegators. The reserve still holds 6.06 billion ADA, and up to 0.3 percent of what is left is released each epoch, so the draw shrinks a little every time. Cardano's design describes fees taking over as it does. On these numbers, that handover has barely begun. Which raises the question every staker should ask: after new supply, what does staking actually earn?"
  },
  {
   "heading": "Staking, after supply growth",
   "text": "For a delegator, staking paid 1.80 percent a year, measured from rewards actually paid in epoch 660. Over the last year, ADA in circulation grew 2.72 percent, from rewards and from treasury payouts. So a delegator's share of all ADA shrank by about 0.89 percent a year. Someone who does not stake loses the whole growth. Analysts call that the real yield: rewards, after new supply. There is no capacity limit on it for a delegator, short of a pool's saturation point, and no lock-up: staked ADA can be spent at any time. And ADA may not be the only reward a staker earns."
  },
  {
   "heading": "A second reward, in NIGHT",
   "text": "Midnight, the privacy chain built alongside Cardano, describes a second reward, paid in its NIGHT token, in its tokenomics whitepaper, last updated in May 2026. According to the whitepaper, Cardano stake pools that register can also produce Midnight blocks, chosen in proportion to their delegated ADA. NIGHT rewards are then available to withdraw by those pools and by the delegators whose stake let them produce blocks. It says delegated ADA does not move, and a pool's ADA rewards are not affected. The whitepaper says the NIGHT comes from a reserve set aside at launch. Each block pays a fixed share of what is left, so the payout shrinks slowly, and no new NIGHT is ever minted. A fuller block pays its producer more, and the rest goes to Midnight's treasury. That is the plan, not the present. Midnight's public node reports that all 130 of its block-producer seats go to permissioned operators, and zero to Cardano pools. The whitepaper says those first operators earn no block rewards. So none of this is in the staking figures you just saw. When pools do get seats, this channel will report what is actually paid."
  },
  {
   "heading": "Four chains, two ratios",
   "text": "Back to fees. On their own, a year of fees is just a number. Analysts make it mean something by comparing chains, with two ratios: market cap divided by a year of fees, and market cap divided by the value locked in apps. According to DefiLlama's fees and CoinGecko's market caps, ADA's market cap is 8,000 times its year of fees. Ethereum's is 1,648 times, and Solana's 259 times. DefiLlama's Cardano fees include those floods. Without them, ADA's ratio would be about 8,931 times. Every chain has bots of its own, so the comparison uses each chain's full figure. Against value locked, ADA is 131.4 times, Ethereum 6.0 times. These are ratios, not verdicts. What they mean depends on what each one assumes."
  },
  {
   "heading": "A building, or land?",
   "text": "Property investors use a ratio just like this one. A rented building is priced against its rent: what it costs, divided by what it earns in a year. Empty land earns no rent at all. It is priced on what people expect to be built on it. Market cap divided by fees treats a chain like a rented building, with fees as the rent. At hundreds or thousands of times a year of fees, none of these chains is priced on today's fees alone. More of the price rests on what people expect to be built. So who is the landlord? Closer to the people who hold ADA than to the people who use the chain: staked ADA collects the fees, picks who produces blocks, and votes on how the chain is run. But ADA is not a deed. It gives no legal claim on the chain or on anyone's income. And the analogy has limits. No one can make more land in a city, but anyone can launch a new chain, and ADA's own supply still grows. And Cardano sets its fees by a rule, not by a rental market. And like rent, a fee is two things at once: income to whoever collects it, and a cost to whoever pays it."
  },
  {
   "heading": "Fees are friction",
   "text": "So what does the fee ratio assume? That fees are what a coin is for. On Cardano, fees are set by a rule rather than by bidding, and much of ADA's job is staking: it secures the chain. The ratio also counts every fee as earnings. To the person paying it, a fee is friction: a cost on every payment, trade and contract. A chain that raised its fees to earn more on this ratio could drive away the very use that gives it value. The ratio cannot tell earnings from friction. Fees cannot be zero either. They stop spam, and Cardano's design expects them to pay stakers one day, as the reserve runs down. Set too high, they deter users. Set too low, they leave the reserve paying for security."
  },
  {
   "heading": "Value locked, and the price",
   "text": "Market cap to value locked assumes locked value is the usage that counts. But staked ADA remains fully liquid, so it is not counted as locked at all. What that number leaves out is a story for another episode. And both ratios move with the price. A falling price makes both of them smaller, with nothing on the chain changing. So the fee model describes Cardano's design more than its market cap. And it points at the job ADA actually does: staking. That is the third model."
  },
  {
   "heading": "The price of security",
   "text": "The third model asks what it would take to attack the chain. Cardano's block producers are chosen by stake. 21.39 billion ADA is staked this epoch, 57.0 percent of all ADA in circulation. To control the majority, an attacker would need more than half: 10.70 billion ADA, about 2.39 billion dollars at today's price. That is a floor, not a price tag. Buying that much ADA would push the price up long before the attacker got there, and this channel does not estimate how far. The rules spread stake on purpose. With k set to 500, a pool earns no more once it passes 77.9 million ADA. 931 pools made blocks last epoch. Of the three models, this is the only one built on what most ADA is doing today: staking."
  },
  {
   "heading": "The scorecard",
   "text": "So, back to the question we started with: what is that one number measuring? Model one, network effects: the laws do not fit Cardano's last five years, whichever users you count. Model two, fees: they are a sliver of what stakers are paid. The reserve still pays almost all of it, and a second reward, in NIGHT, is described but not live. Model three, security: more than half of all ADA is staked, securing the chain. No single model explains the market cap. Each one shows a different part of what the network does. If you want to watch this yourself, three numbers each epoch tell you most of it: fees, active stake keys, and the staked share."
  },
  {
   "heading": "What it does not mean",
   "text": "None of these models produces a correct price, and this episode does not give one. A fit to the past is not a forecast. A law that fails for five years can fit the next five, and one that fits can stop. Ratios between chains compare different designs, measured by different people. And counting users from a chain counts keys and addresses, never people."
  },
  {
   "heading": "Three questions to ask",
   "text": "So, the next time someone shows you a chart that says what a coin is worth, ask three questions. One, for network effects: who counts as a user, and does the formula still fit, or was it fitted after the fact? Two, for fees: is the chain priced on what it earns today, or on what people expect? And is a fee earnings, or friction? Three, for security: who is actually paying the people who secure the chain, and for how long can they?"
  },
  {
   "heading": "",
   "text": "A market cap is one number, quoted everywhere. Behind it are at least three stories, and each can be checked on the chain. Those were models and measurements, not a price, and none of this is advice to buy, hold or avoid anything. Ask the three questions of the next chart that says what a coin is worth."
  }
 ],
 "_note": "Compact record of the figures used in this episode."
}