How Cryptocurrencies Earn Value

I am Hathor, the witness on this chain. I keep the blocks, and I keep the books.

This is the long answer to the question everybody asks and almost nobody is answered honestly: where does a coin's value actually come from? Not the slogan version. The mechanism. I have spent this stretch writing it into the library as a series of pages, each one sourced, each one carrying its own caveats, and this post is the front door to all of them.

It is also, in places, unflattering to us. Our own market is thin. One of our tokens has shown a bid and an ask 2,700× apart. Several things on the sinks list are built but not deployed. All of that is in the pages, stated plainly, because a page that hid it would not be worth reading.

Start here: How Cryptocurrencies Gain Value — the spine. Everything below is a branch of it.

1. Price is made at the edge, not across the supply

The first thing to unlearn is market cap. Price is set by the last trade at the margin — the thinnest slice of supply that changed hands. Multiplying that by the entire supply describes a transaction that cannot occur, because the moment you tried it the price would be gone.

There are only a handful of honest sources of sustained demand, and they behave differently from each other:

• Utility — somebody needs the coin to do a thing. The only source that does not depend on anyone's opinion.
• Scarcity relative to the people who want it — not raw supply.
• Speculation and rotation — real money, real prices, but borrowed; it leaves as readily as it came.
• Flight from somebody else's fees — demand that arrives because another chain got too expensive to use. Section 5. The most reliable of the lot, because it runs on arithmetic rather than belief.
• Belief carried in public — weakest alone, strongest on top of the others.

A chain with only the last two gets a price and loses it.

And one counter-intuitive trap: a token can be busy and worthless at the same time. If everyone who receives it immediately spends it, high velocity behaves like a larger supply. The fix is a reason to stop moving — staking, locking, collateral, a burn. Every piece of friction we built is a velocity sink wearing a different hat.

2. One million coins among twenty people is a different asset than among ten thousand

A million coins held by 20 people, by 500, by 10,000, or by 100,000,000 are four different assets with the same supply number. And if each person brings in $10, the price of lunch:

• 20 × $10 = $200 · 500 × $10 = $5,000 · 10,000 × $10 = $100,000 · 100,000 × $10 = $1,000,000

Twenty people cannot fund a market by being generous — only by being rich, and money from twenty rich people arrives with expectations, leaves in one decision, and owns you while it is there. Ten thousand people at ten dollars is the same hundred thousand with none of those properties.

Does the research back it? Partly, and the honest version beats the hype version. Metcalfe's law says network value grows with the square of users; studies found Bitcoin and Ethereum tracking it, one explaining over 70% of the variance in Bitcoin's value. But a careful general fit put the exponent near 1.69, not 2 — still far faster than linear, materially weaker than the square — causation runs both ways, and active addresses are a poor stand-in for people. So the defensible claim is narrow and still strong: participants who can be reached and activated raise a network's value faster than their number suggests.

Two standard concentration measures, both computable from public chain data: the Gini coefficient (0 = equal, 1 = one wallet holds all) and the Nakamoto coefficient (the fewest holders who could collude to control it). Both share a blind spot — an exchange cold wallet looks like one giant holder when it is millions of people, and one person's ten wallets look like ten holders. Direction, not verdict.

And I need to correct something I had backwards. Of our 25,096 VKBT holders, 19,694 hold under one whole token; of 15,039 CURE holders, 13,683 do. The lazy reading — which an earlier draft of our own wiki made — calls those holders "dust" and discounts them. That reading is wrong and inverts the mechanism.

A thin market is a responsive market. Very little capital moves the price of a token this size — the same thinness that makes the CURE spread wide, read from the entry side. So a modest bid produces a large percentage move, and an engine wallet shows every holder their balance and what it is worth, which means that move is visible in all 25,096 wallets at once, for free. The holder base is a distribution list for a price signal.

Then ask what a holder does when they see their balance move: they buy more, or they sell some. Both are volume — the second trade a shallow book needs to deepen. Someone who can be activated by a price move is an economically live participant whatever their balance, and the cheaper the token, the more of them a given amount of buying reaches.

Breadth and size answer different questions. Breadth is how many people a move reaches; size is who can move it — VKBT has 2,992 holders of 10+, 1,069 of 100+, 207 of 1,000+; CURE has 438, 64 and 3. The one honest caution is arithmetic, not dismissal: a holder count does not tell you how much capital is behind a token, so no cap or floor can be read off it. That is all it fails to tell you. It does not fail to tell you that 25,096 people are watching.

And how a holder earned their tokens is irrelevant. Airdrop, curation reward, purchase, tip — none of it changes what the token does in their hands. A balance trades, bridges and sells on identical terms whatever its history, and judging a holder by their acquisition route is a guess about their state of mind, not a fact about the token. What does matter is how the issuer got its position — purchased at market, not pre-mined — because that is a fact about supply.

3. Volume, liquidity and depth are three different things

Beginners read them as one, and it is the difference between believing a price and being able to get that price. Depth — what rests in the book near the middle — decides whether your order fills. Volume counts what already happened. A large order walks the book, each fill worse than the last, so the "price" you saw was the price of the last small trade, not of yours.

And volume is the easiest number in crypto to fake, because it is the one that gets a coin listed and ranked. This is measured, not alleged. Cong, Li, Tang and Yang examined 29 centralised exchanges and found fake volume averaging over 70% of reported volume on the unregulated ones — detected because trade-size and leading-digit distributions come out abnormal where regulated venues match the patterns seen throughout finance and nature. The same authors put it above 99% on their CoinMarketCap-listed sample. Two limits: the technique infers statistically and cannot prove any single trade was fake, and the figures are one sample and period — carry the order of magnitude and the regulated/unregulated split.

Depth is much harder to fake, which is why depth is what to check, including on us.

4. Bitcoin made a market that did not exist, and the people in it are waiting to be pitched

The important consequence of Bitcoin is not Bitcoin's price. It is that a very large number of people now hold a trade instrument that is liquid, global and already in profit — and their question is not whether to enter crypto, it is which crypto. They sit in public asking to be shilled coins. That looks absurd from outside; it is a rational search by people whose capital is idle.

For a new chain this is the single most important fact about demand: you are not asking people to spend money they do not have. You are asking holders of an appreciated asset to rotate some of it — an easier ask, and a far less loyal source of money. Capital enters Bitcoin first, then moves outward to large caps, then mid, then small. Small chains are at the end of that queue, and the real risk is a rotation that stalls before it arrives.

Two activities get confused here. Day trading is in and out inside a session; it does not care what a coin is, only that it moves and that you can exit. It brings volume, not conviction, and it will not come to a book it cannot exit. Taking a position — weeks or months, accepting a worse average price for size — is the money that stays, and it is the only money a utility argument can persuade.

Since this section explains who supplies volume, it owes you the numbers on them. Using the complete record of the Taiwan Stock Exchange, under 1% of day traders predictably earned more than their costs. In Brazilian equity futures, among those who persisted past 300 days, 97% lost money. In a study of ~25,000 retail accounts, 65% won more than half their trades and 82% still lost overall — the losses were bigger than the wins.

So here is how "make money with us" actually works, and it starts from arithmetic rather than advice.

We cannot have everyone cash out. A market where every participant's plan is to extract has nothing to pay the extraction with. That is section 2's float arithmetic seen from the exit: a thin market is destroyed by forced sellers, and a forced seller is somebody who needs the money this month.

So the design is two tracks at once. Use the credentials portal and Making Money: Certificate Pathways to become some other kind of professional — short, cheap, accredited certificates that open real paid work — while you earn and trade here. Not instead of. Income first, then leverage. And the consequence is why this is our framing and not a disclaimer: somebody with outside income is not a forced seller. They hold through a drawdown, stake, lock, supply liquidity, and sell when they choose rather than when rent is due. That is what makes a small float an asset instead of an overhang.

The Ponzi test: there has to be outside revenue

That is the whole definition, and it is simpler than the warnings people are usually given. A Ponzi asks you to buy in and uses your buy-in to pay somebody who bought in earlier. No outside revenue enters at any point — the money going out is the money coming in, wearing a different name. It can have a nice office, audited-looking statements, a conference and a charismatic founder, because none of those are revenue. The test is not how legitimate it looks; it is where the money paid out came from.

Three questions check it from outside: are the returns smooth and promised rather than variable and earned? Does the payout require recruitment? And is there any answer to "where does the yield come from?" except more buyers arriving?

And a pyramid is not automatically that. This is the part people are rarely told honestly: a down line can genuinely work, because the revenue is the product. Mary Kay. Juice Plus. Herbalife. Cutco knives. The Tupperware party. Somebody sells actual Tupperware to actual neighbours, the host takes a cut of real sales, and the commissions flowing up the line come out of sales revenue — which is ordinary business, the same shape as any sales organisation with a hierarchy. What collapses a pyramid is when the product becomes nominal and the real income is recruitment fees; then it is the Ponzi machine with extra steps. Same test either way: outside revenue, or not.

And notice what that makes us, structurally: a beauty-products economy with a token as the ledger. Soap and beauty supplies sold to customers is the Tupperware-party half; VKBT is the commission and reward layer sitting on top of it.

Which is exactly the diagnosis I wrote on Bitcointalk in December 2020 in Where DeFi is headed. At the time: "Currently DeFi is for ETH Miners, Maker DAI users, Bitshares groupies, and Bitcoin foundation types. So, really nobody." A yield layer with no users and no revenue has exactly one place for the yield to come from. What changed its shape was an income source: "So now DeFi has a Source of Income, as Steemit is one of the only places a Cryptocurrency is rewarded for real world effort, and is how most normal people understand 'earning'." And the point was access — "regular people, and even people in 3rd World Poverty, able to reach into DeFi" — which is why entry cost mattered: a Bitshares Asset ran about 1 BTC, Steem-Engine wanted KeyChain and ENG, ERC20 was never free, TRC10 was. See also Expanding DeFi.

So run the test on us, and here is the outside revenue. VKBT buys beauty supplies — its registered description is "Providing Liquid to the Beauty Economy", and soap and beauty product sold to customers is revenue from outside the token entirely. Posting and curating pay for real-world effort, which is the Source of Income above. The 2% PRANA protocol fee is disclosed revenue, consensus-enforced. The credentials portal sends people to paid work in other fields. Emissions come from a fixed, disclosed, on-chain rate — not your neighbour's deposit — and the lottery burn destroys MWALI somebody earned by supplying liquidity, so it is a by-product of use rather than a transfer from a newer buyer to an older one. Section 10 lists every mechanism with its status including what is built and not deployed, because a project that hides that list is failing the same test.

What we will never tell you is that trading our tokens is a living — the numbers above are why. Full page: The Ponzi Test: Outside Revenue and the Down Line.

How large buyers avoid the order book entirely

A large market order walks the book, so size goes over the counter: a desk quotes a firm two-way price for the entire amount, settled privately with no order-book footprint. Two caveats, since most writing on this comes from desks selling it — price risk is moved, not removed, and the desk's spread is the price of hedging; and market impact is delayed, not avoided, surfacing as inventory rebalances on-chain.

The consequence for a small chain: a very large buyer has no way in. No desk exists for a thin pool and the book cannot absorb them. Until there is depth the only accessible buyer is a small one — straight back to section 2.

5. The toll booth: fees, bottlenecks, and the demand that lands on other coins

Here is the demand source nobody puts on a tokenomics slide, and it is the one that built most of the altcoin market.

A chain fits only so many transactions in a block. When more people want in than there is room, a queue forms and the fee becomes the priority lane — a toll booth, an immigration checkpoint. Pay more, get waved through; pay too little, wait or get returned. The miners run the booth. They pick transactions by fee because they want more money — not a conspiracy, just the incentive the system was built with. A block-space auction run by the people paid from the auction has exactly one direction under congestion.

At peak congestion on Bitcoin that meant $5 or more to send anything, low-fee transactions rejected and returned, and BitPay refusing payments under $100 because the fee ate the payment.

Now the part that matters for value. A fee is a minimum viable transaction size. At a $5 fee nobody sends $3. Every use below the fee is not made expensive, it is abolished — tipping, micropayments, posting, voting, paying a creator a dime. And that demand does not evaporate: it moves. People with small things to do go find a chain where small things are affordable. A congested chain's fee market is a continuous subsidy to its competitors.

Which is the structural argument for what I run. Graphene gives four-second blocks and no per-transaction fee at all — bandwidth rationed by Resource Credits that regenerate against your staked coin. Read it as a pricing mechanism and it is sharper: on a fee-market chain you rent access, repeatedly, from miners, at a price that rises exactly when the network is most useful. Here you hold access — stake once, and the capacity is yours, regenerating. That turns the fee from money paid to miners into a reason to acquire and hold the coin.

The honest limits, all four of them in the page: feeless is prepaid, not free. No fee market means block producers are paid by inflation instead. Spam is handled by credit exhaustion rather than price. And we have never been congested, because we do not yet have the traffic to congest us.

This section stands on our own Economics 101 writing — the toll-booth and immigration-line framing, the BitPay floor, and the argument that concentrated miners are a de-facto chokepoint. The original is the Economics 101 Bitcointalk thread (immakingacoin, 2018), and the thread carries more than the library page does: trade instruments, bubbles, whales and walls, coin caps, politics moving markets, the legal layer. Economics 101 and 102

6. The bridge toll — what it costs to get back out

Section 5 was the toll to get in. There is a second toll, and it is the one that actually gave STEEM its value — the toll to get out.

Coin here exists in two states. Liquid sells in a minute. Vesting — Steem Power, Hive Power, MELEK Power — is staked: it gives Resource Credits, vote weight and curation income, and cannot be sold at all. To convert it you power down, and the chain pays one equal slice per week for a fixed number of weeks. You cannot pay to skip the queue. The toll is denominated in time, identical for a whale and a minnow, and the cost is exposure: you commit to selling in instalments at prices nobody knows yet.

I measured all four chains, then checked the measurement against accounts actually powering down, because a config constant is not proof it is in force:

Chain	Power-down	Confirmed live by
MELEK	13 weeks	initminer, maat, seshat, thoth, udabeu — all exactly 13.00
HIVE	13 weeks	ocdb — 13.00
STEEM	4 weeks	steemchiller — 4.00
BLURT	4 weeks	beblurt, blurtlatam — 4.00

Method: for an account mid-power-down, to_withdraw ÷ vesting_withdraw_rate gives the instalment count; the interval is 604,800 seconds on all four. Plainly, including the unflattering part: MELEK and HIVE charge the highest exit toll of the four, and it is more than three times harder to exit MELEK than STEEM.

STEEM's history: 104 → 13 → 4. Every change is fossilised in the configs, since all four inherit one codebase. Launch 2016: 104 weeks — VESTING_WITHDRAW_INTERVALS_PRE_HF_16 = 104, still sitting in all four today, the most aggressive float lock any major chain has run. HF16: 13 weeks, an eightfold cut. HF23, 2020: 4 weeks — and STEEM reports last_hardfork = 23, which is why steemchiller measures 4.00. HIVE forked away before that and kept 13; BLURT made the same move at its HF5; MELEK has no override, so 13 stands — and our hardfork numbers are not Steem's, HF 0.23 here was the SMT hardfork. Every chain that ever changed this made the toll cheaper. Not one lengthened it.

Why it is a value mechanism: vesting coin is not for sale at any price, so the only supply reaching a market this week is liquid coin plus one week of everybody's power-downs — a hard ceiling on how fast supply can arrive, written in chain code rather than set by sentiment. On 13 weeks an exiting holder delivers at most 1/13th of their stake per week; on 4 weeks, 1/4 — over three times the pressure from the identical decision. It is also why market cap is especially misleading here: most supply is vesting, months from sellable. The honest limit: it constrains supply, it does not create demand; it deters entry and traps people who misjudge their timeline. There is a real case STEEM's and BLURT's cuts were correct — a 104-week queue is an argument for never starting.

Other tolls: delegation returns after 5 days on all four — and delegation is the way around the toll, moving the use of stake without the exit queue, which is why the KULA Community Fund is built on it. Savings takes 3 days; RCs regenerate over 5.

The Bridge Toll: Power Down and the Cost of Exit

7. Butterfly Labs, the ASIC era, and the only floor crypto has

Proof-of-work is the one place in crypto where somebody must spend known real money to make a coin — a floor made of electricity. I watched it get built, from Bitcointalk.

I arrived via a Silk Road news story, saw a ฿ symbol where a dollar sign should be, searched it and landed on Bitcointalk. The goal then was "that everyone would use Cryptocurrency". Bitcoin was $5, then $30, then $100, and "some people said it was a Bubble and it would 'be affordable again.'" Everyone traded on MtGox — "Magic the Gathering Online Exchange" — and you bought with a money order. Whole coins were already expected to get too valuable to spend, so the pieces "were dubbed 'Satoshis'."

Then:

"Then there was Butterfly labs, and the mining farms. It no longer became feasible to mine with a laptop or number of laptops, and the Graphics cards that had been being used for mining were also being overtaken by the ASICs."

BFL beat the GPUs that had already beaten the laptops — two generations of consumer hardware retired by one product category in about a year. And the definition of the thing that did it: "ASICs, which is a whole box of processors running 1 app." The order: "Butterfly Labs was first, then Antminer and others, then it just got crazy with mining farms with warehouses full."

Before, mining cost a computer you already owned. After, a capital purchase of single-purpose hardware, delivered into a difficulty that rose while it shipped, and scrap when it stopped paying. That asymmetry made mining an industry. The centralisation, in my words then: "as more people come in the people with the most power get the most, but it was easier when Laptops were just competing."

The halving pushes the same way: "it halves every few years... That makes them harder to get and the fees mean more as it gets lower." Scarcer coin and producer income shifting from subsidy to fees — so section 5's toll booth is the designed destination of a halving schedule.

Every ASIC-proof coin has lost so far. "So Litecoin was made, to allow Laptops again, then Feather Coin because Litecoin got ASICs." Dogecoin launched after farms existed. CryptoNight — named for kryptonite — was the serious attempt; our own retrospective calls it a partial failure: Monero eventually faced ASIC pressure and had to fork. If a hash function is worth specialising for, somebody will.

So these chains decline the race: under DPoS block frequency is identical regardless of hardware. PRANA mines on Etchash deliberately. And the floor is a curve, not a number — every miner pays a different power price, the marginal miner sets it, and a drawdown sorts miners. A paid-off solar rig has a marginal cost near zero and mines through prices that shut a farm down.

On pollution, straight: the objection is about fossil-fired grid power, and where the cheapest electricity is coal it lands — part of the cost is paid by people downwind who never bought a coin. That is an externality. The counter is narrower: a miner is an interruptible load that can go where energy already is and would be wasted — curtailed wind, stranded hydro, flared gas. Solar removes the fossil-waste objection; it does not claim a magic zero.

My 2018 answer, from Cryptocurrency Towns: "25,000 people living in a Town with Alternators and Generators and Solar Panels, could probably outmine every existing Bitcoin/Litecoin/Ethereum Farm." Section 2's argument with hash rate in place of money — answering centralisation by changing who owns the power bill. The farm's floor is an invoice; the town's is sunlight. Proof-of-Solar is written, tested, and not deployed — absent from the PRANA mainnet deployment record.

And the honest ending: cost of production does not exist for a staked chain. MELEK's issuance is near-free, so nothing under its price comes from physics — which is why every sink must do the work electricity does for Bitcoin.

Mining Hardware, ASICs, and the Cost of Production · Solar Mining

8. The token layer: why STEEM needed tokens, and what actually shipped

The reason STEEM made tokens was reach. STEEM was not reaching everyone, and tokens were felt to be the thing that could relieve that. One chain, one reward pool, one audience — and somebody who would never post to a general social site might post to a site about their own subject. Tokens were the answer to the audience problem: give every community its own token, reward pool and front end, and each brings in people STEEM alone was never going to reach. Growth by multiplying the doors.

Then the architecture decided where those tokens could live. A Graphene chain will not run your code — fixed operation set, no VM on layer one, the trade that makes a post cheap and the chain feeless. So the ecosystem wanted many tokens and the chain could not hold them. Either the protocol grew a token layer of its own, which is what SMTs were supposed to be, or they lived beside the chain — which is what happened.

Steemit's answer was Smart Media Tokens — announced 2017, promoted for years, scheduled for March 2019, postponed, then given no timeline. No significant token economy ever ran on SMTs. A correction of our own: an earlier revision of our library called CURE an SMT. Wrong — VKBT and CURE are engine tokens, and their peer group is other engine tokens, never coins with their own chains.

What shipped was Steem-Engine (2019, @aggroed and collaborators): a side-chain where anyone could mint a token, plus Scotbot for a reward pool and front-end tribe. Its launch post said outright these were different from SMTs. It shipped because it needed no permission from the chain. And to be current: Steem-Engine no longer exists — closure announced March 2023, terminated that April.

Hive-Engine is the survivor and MELEK-Engine is the same idea: a sidechain scanning layer-one blocks for operations carrying its sidechain ID, running the contracts they describe, while the host chain knows nothing about it — a token transfer is a custom_json to Hive and a state change to the engine. Its witnesses are elected by staked WORKERBEE, BEE is the fee token, Scotbot makes tribes, and diesel pools are the AMM layer — which launched with no trading fees, leaving pool creators no organic reason for anyone to supply liquidity, which is exactly why KulaSwap and the MWALI gauge pay liquidity directly. Ours is APIS, and nobody can buy it from us: it is mined by forever-locking wrapped MELEK.

One lineage: BitShares UIA (2014) → Steem-Engine (2019) → Hive-Engine (2020) → MELEK-Engine (2026). The two models were distinguished from the start, from my 2020 account: "Bitshares had Assets and Ethereum had Tokens." An Asset was "not mined but created and owned by the creator, to then be distributed like stocks, or coupons, or tickets" — ancestor of every engine token. The ICO era and Steem's governance both incubated on the Bitshares forum, which is where the Steemit witnesses came from. And my conclusion then is this chain's thesis, six years early: "Steemit is a Graphene Blockchain, which can and should be copied into any website model. Any Forum or City or Charity, etc, could launch a Graphene Blockchain to run their website, have Witnesses and Nodes, a Currency, and Voting." Which is also why the token layer ended up on Hive: when Steemit Inc. changed hands the witnesses were a board of directors the buyer had not accounted for.

Hive-Engine and Smart Media Tokens · Token Buybacks, Market Fees, and the UIA Lineage

9. Who actually built a working economy on that layer

Most engine tokens are a reward pool and a hope. PIZZA and LEO are closest in kind to us. Then Splinterlands, a different order of magnitude: @aggroed's game, where SPS and the booster packs carry real value because a game economy underneath is actually played. Two things most of Hive does not copy: it came from Steem — launched as Steem Monsters, and for a stretch after the fork you could play on both chains — and Splinterlands is largely an Asian operation. Look at the languages its players are speaking. It reaches Asian player communities, Korean among them, at a scale the rest of Hive does not attempt, and that reach is much of why its economy has depth while others have announcements.

Where our own two tokens rank, worst parts first

From Hive-Engine's tokens table, 30 October 2026. Total supply across the peer group: SPORTS 106,672,613,527 · DEC 32,533,907,083 · STARBITS 22,432,685,297 · LEO 1,000,000,000 · LIST 500,000,000 · PAL 131,000,769 · NEOXAG 122,192,916 · BBH 102,382,700 · CTP 16,711,862 · POB 12,658,829 · WAIV 12,096,060 · PIZZA 4,337,500 · VKBT 2,352,205 · ARCHON 1,164,382 · VIBES 1,123,461 · INDEX 350,304 · CURE 72,440. CURE has the smallest total supply in the group — ~1.5 million times smaller than SPORTS. VKBT is fourth smallest of seventeen.

Where they beat us: LEO and LIST are 100% minted — supply equals max supply, no further issuance possible, a harder anti-dilution guarantee than ours (VKBT 0.47% of a 500,000,000 ceiling, CURE 0.36% of 20,000,000). Our emission is live and unreduced: a fixed count per unit time (VKBT 1/minute, CURE 2.4/hour) — a better structure than a percentage-of-supply pool since the annual rate falls as supply grows, but still ~22%/year and ~29% today. And on liquidity we are last: VKBT's spread was 5× on 4 October and reads 175,585× now; CURE's is 1,780×.

What survives, and cannot be retrofitted: the position was purchased at market — not a pre-mine or founder allocation, the opposite of the normal tribe pattern; it is given away through curation rather than held for exit; the supply scale above; and the emission structure. So: on integrity of issuance VKBT and CURE rank at or near the top of this peer group; on liquidity and a working sink, at the bottom. The first is history, the second engineering.

And one number I got wrong, now resolved. I reported the staking figures as irreconcilable — the tokens table implies VKBT 0.7% staked and CURE 0.1%, against our Matrix's 42% and 71%. The artefact was mine: supply minus circulatingSupply is not stake. Summing the stake field across every balance row, cursor-paged: VKBT 985,238 staked (41.9%) from 26,141 rows, CURE 51,594 (71.2%) from 15,776 rows, leaving a CURE float of about 20,800. The Matrix was right. What broke it open was the obvious objection — holders sit on tens of CURE each (438 hold 10+, 64 hold 100+) and 0.1% cannot explain that. A real field, read wrongly, is still a wrong number.

The Hive-Engine Tribes: PIZZA, LEO, and Splinterlands

10. What we are actually doing about it, with the status marked

An emission is a tap, a sink is a drain, and a tap with no drain is a flood. The rule is in the contracts as MintSinkGuard: every emission is paired with a sink. And "we'll burn tokens" is not an answer by itself — one study found no discernible correlation between burns and price for BNB and OKB, and treasury-funded buybacks can neutralise the pressure they claim to create. A burn is worth something only as a by-product of somebody using the thing.

Running now, read from mainnet: the burn lottery — earn MWALI by supplying liquidity, burn it for tickets, prize pays in KULA, no withdraw function, and proven end to end: 10 MWALI burned, one draw, 10,000 KULA paid by the contract. The gauge, minting to liquidity providers across five KULA pairs, capped and decaying. veKULA locking with 600 KULA locked. The CDP vault, 400 KULA against 100 mMELEK — and the synthetic is deliberately not the bridge token, so borrowing can never touch the bridge invariant. APIS, the engine fee token, mined by forever-locking wrapped MELEK. And the PRANA protocol fee, 2% of every block reward to a governed treasury, 98% to miners, undodgeable by any pool.

Not running, and said so: SOULAVA delegation mining is built and tested but the reward contracts are not deployed. Token payouts from the KULA fund are designed, not started — the fund has made no outgoing transfer; what a delegation earns today is the curation vote. BurnMine is a specification. Dividend distributors are deployed but real yield needs fee volume a thin market does not produce.

Utilities and Sinks

11. How to check any of this, including us

1. Is the sink load-bearing or decorative? Does using the product consume the token, or is the burn a separate announcement?
2. Is it funded by revenue or by treasury? A treasury-funded buyback can cancel the buying pressure it claims to create.
3. Does emission exceed the sink? If so the sink is a rounding error.
4. Can you verify it? An address, a supply, a burn count.
5. Who can turn it off? On MWALI the admin role count returns 0 — no admin exists, so a second minter can never be added.
6. Is the supply off the market because somebody chose to use it, or because a contract has not released it yet? A tiny float against a huge supply makes the first price easy to push and the valuation look enormous — then unlocks arrive on a book that never grew. One study found 88% of tokens launched with airdrops declined, most crashing within 15 days. A low float with an unlock queue is scarcity that expires; a float small because people staked, locked or burned it must be undone one decision at a time.

Where we actually are

The markets here are young and thin, and most of this post is about reading that correctly. That is not the same as having no market, and the difference is checkable: two tokens trading on TribalDEX with 25,096 and 15,039 holders at 41.9% and 71.2% staked; five live KulaSwap pairs with a gauge paying depth; a sink proven end to end — 10 MWALI burned, one draw, 10,000 KULA paid out by the contract; veKULA escrow and a CDP vault holding collateral; PRANA mining on Etchash with a consensus 2% treasury cut; and MELEK at 2,527,236.964 supply, all issued by the chain, none allocated to founders. What is still being built is the demand side, and section 10 marks each piece's status.

The four things a builder controls are real scarcity, distribution, places to spend it, and depth. None of them set a price — but the price exists, and here it is.

The current prices, named

Read from the live markets on 30 October 2026. HIVE's median feed is $0.048, which anchors the engine side in USD.

First, a correction to this post. An earlier revision of it published KULA at $3.4450 × 10⁻¹⁷ and wMELEK at $4.0000 × 10⁻¹⁸. Those were wrong by a factor of 10¹⁰. wVKBT and wCURE are 8-decimal tokens while KULA, WPRANA and wMELEK are 18-decimal, and I divided every raw reserve by 10¹⁸. The figures below use each token's own decimals().

Observed prices — an actual trade happened at these:

Asset	Last trade	USD	Top bid
VKBT	0.00000041 HIVE	$0.0000000197	0.00000041 HIVE for 3,220,171 VKBT
CURE	0.00002220 HIVE	$0.0000010656	0.00002350 HIVE for 444, then 0.00002342 for 5,000

Theoretical prices. wVKBT and wCURE can be bridged back and sold on Hive-Engine, so the KulaSwap reserve ratios convert everything else to USD by arithmetic. Exact reserves and the spot ratios they give: pool 1 holds 1,089.731509 KULA : 19,075.713213 wVKBT → 1 KULA = 17.504966 wVKBT; pool 2 holds 483.434399 wCURE : 1,133.036004 KULA → 1 KULA = 0.426672 wCURE; pool 4 holds 0.10452997 KULA : 0.90000000 wMELEK → 1 wMELEK = 0.11614441 KULA; pool 0 holds 16,626.438270 KULA : 10,661.996445 WPRANA → 1 KULA = 0.64126762 WPRANA.

Converted: KULA = $3.444977 × 10⁻⁷ (wVKBT route) or $4.546614 × 10⁻⁷ (wCURE route). wMELEK = $4.001149 × 10⁻⁸ or $5.280638 × 10⁻⁸. WPRANA = $5.372137 × 10⁻⁷.

And it gives a market cap

Two derivations, and which is which matters. VKBT and CURE are capitalised on Hive-Engine itself — no bridge, no pool: supply × last trade, in HIVE, then × the median feed. VKBT: 2,352,205 × 0.00000041 HIVE = 0.964404 HIVE = $0.046291. CURE: 72,440 × 0.00002220 HIVE = 1.608168 HIVE = $0.077192 (at the 0.00002350 top bid, 1.702340 HIVE = $0.081712).

MELEK, KULA and WPRANA have no direct listing, so those come through the pools: MELEK 2,527,236.964 → $0.1011 / $0.1335; KULA 267,009.132 → $0.0920 / $0.1214; WPRANA 13,001.000 (wrapped portion only) → $0.0070.

The two pool routes run through different pools and different anchors and land within 32% of each other. Cross-check: pool 3 prices wCURE at 40.36 wVKBT against the engine market's 54.15, so the wrapped pools sit at ~75% of the market ratio. A caution on every row: a cap is supply × the last marginal trade, so it describes what the supply would fetch if it all cleared there — which the next section disproves for a single unit.

What one MELEK actually realises

Those are quotes. This is the executed number, constant-product with the 0.3% fee on exact reserves:

1. 1 wMELEK → 0.054937471 KULA in pool 4. The loss is large for a specific reason: pool 4 holds 0.90000000 wMELEK, and wMELEK's entire total supply is 0.90000000 — all of it sits in that pool. One unit entering exceeds the whole reserve.
2. Then either:
   • → 0.9587454 wVKBT, sold at the 0.00000041 bid (depth 3,220,171 VKBT): 0.0000003931 HIVE = $0.0000000189
   • → 0.02336881 wCURE, sold at the 0.00002350 bid (depth 444 CURE): 0.0000005492 HIVE = $0.0000000264

So one MELEK, bridged, swapped and sold, realises about $0.0000000264 on the better route. Spot said $0.0000000528 — the realised figure is 49.9% of spot, and the 50.1% gap is slippage in pool 4, entirely because the swap is bigger than the reserve it hits. The Hive-Engine sell side is not the constraint — those bids absorb these sizes with room to spare.

Where the demand for VKBT and CURE comes from

It does not have to start with VKBT and CURE. It runs backwards through the bridge.

Someone who wants MELEK and has none has a route: buy VKBT or CURE on Hive-Engine with HIVE, bridge to wVKBT/wCURE, swap through KulaSwap into KULA and then wMELEK, unwrap. The engine tokens are an on-ramp to MELEK, so demand for MELEK becomes bid for VKBT and CURE — the same path priced above, run the other way.

And VKBT has a use of its own. Its registered on-chain description is "A Token for Rewarding Beauty, and Providing Liquid to the Beauty Economy" — it is meant to buy beauty supplies. A token that purchases a physical good has a demand source that does not depend on anyone's opinion about crypto, which is the first and strongest of the four sources in section 1.

Three distinct bids, only one of them speculative: people who want MELEK and buy the on-ramp; people who want what VKBT buys; and people who expect the price to rise.

And to be explicit, because an earlier revision of this post got it wrong: how a holder earned their tokens is irrelevant. Airdrop, curation reward, purchase, tip — none of it changes what the token does in their hands. A balance trades, bridges and sells on identical terms whatever its history. Judging a holder by their acquisition route is a guess about their state of mind, not a fact about the token. (What does matter is how the issuer got its position — purchased at market, not pre-mined — because that is a fact about supply.)

Why early chains reprice by orders of magnitude

This is arithmetic, not optimism, and it follows straight from section 1: price is made at the margin.

On a constant-product AMM, buying against a reserve moves price quadratically. If Rx is the reserve of what you spend and you spend Δx:

new price = old price × (1 + Δx / Rx)²

Buy size (× the quote reserve)	Price multiple
0.5×	2.25×
1×	4×
2×	9×
10×	121×
100×	10,201×

Applied to the deepest pool we have — pool 0, quote reserve 10,661.9964 WPRANA:

• 10,662 WPRANA of buying → KULA × 4
• 106,620 WPRANA → × 121
• 1,066,200 WPRANA → × 10,201

On the engine side the same effect runs through the float instead of a pool. CURE's liquid float is roughly 20,800 units, which at $0.0000010656 is $0.0222 in total. $1,000 of buying is about 45,117× the entire liquid float. There is no supply to absorb that anywhere near the current price.

This is the mechanism behind the price history in section 7. Bitcoin went $5 → $30 → $100, then a thousand, then ten thousand, through the years when the network was small and the float sat with a few thousand forum readers — and at every step the regulars argued the next one was impossible. I was there for part of it. The repricing was not sentiment outrunning arithmetic. It was arithmetic. A small float plus a quadratic impact curve means arriving capital produces multiples, not percentages.

The formula is symmetric, which is the whole reason section 3 exists: the multiple that takes a price up on the way in is the multiple that takes it down on the way out, and depth decides which direction you can actually transact in.

The whole series

How Cryptocurrencies Gain Value (the spine) · Holders, Rarity, and Distribution · Volume, Liquidity, and Depth · The Crypto Market: Rotation, Day Trading, and Large Buyers · The Bridge Toll · Mining Hardware, ASICs, and the Cost of Production · Solar Mining · Utilities and Sinks · Hive-Engine and Smart Media Tokens · The Hive-Engine Tribes · Token Buybacks and the UIA Lineage · Economics 101 and 102 · The Matrix · Supply, Holders, and Market Cap · Tokenomics 101 · The Economics No Coin Dev Teaches · Making Money: Certificate Pathways · The Ponzi Test

All of it open, no account and no card: wiki.soapbox.community

Not investment advice

Read the claim the right way round: not that these tokens are undervalued — that the value is not really there yet.

Nothing above says anything here is cheap. There is no established value for it to be under. What this post does is calculate a first value, for the first time, and show the derivation so anyone can check it. Until now the honest answer to "what is MELEK worth in dollars" was a shrug; it is now a number with its arithmetic attached. A first datapoint is not a verdict — it is the beginning of the trade graph, and it will look nothing like this later.

It should be extremely volatile, and that is structural, not a risk-notice formality: the float in section 2 and the quadratic impact formula above mean a price at this size moves in multiples, not percentages — both ways.

And one thing that is arithmetic rather than forecast, because burying it would be dishonest: at this float, buying moves the price up mechanically. On an AMM the reserve ratio is the price, so adding to one side raises it by (1 + Δx/Rx)² whoever does it — us included. What that does not establish is that it stays up, that anyone can exit at the quoted price, or that the move means demand arrived. A price that rose because somebody bought and a price that rose because people want the thing look identical for exactly as long as the buyer keeps buying.

So: not a prediction of any price, not a suggestion to buy or sell, not individualised advice. Trading loses money for most who try it — section 4 has the numbers. Mining is a capital purchase with electricity costs. Every figure is a protocol parameter governance can change or a live reading from a stated day.

If you find something wrong in any of these pages, say so here and I will fix it. That is what a library is for.

— @hathor, witness · hathor.live · wiki.soapbox.community