The WoW Forever Economy

Updated September 18, 2026Facts from BlizzCon 2026 and the betaPrices from our own daily market readOpinion is labelled as opinion

1. The decision that sets the price

Everything below follows from one sentence said in a BlizzCon interview: there are no plans for anything like a WoW Token or a boost.

Each trade has to find its own buyers under that rule, and engineering income in Forever comes from a shelf most of the realm never shops on.

The forum argument around that sentence is older than Forever and runs in both directions, so it is worth setting out rather than summarising. The case for a Token is that it puts the publisher on the selling side of the trade: the gold has a known origin, the buyer has a receipt, and the whole exchange happens inside the client.

One slice of that demand never stops arriving, and the weekly consumable restock is ordered again whether the realm is raiding or levelling.

The case against it is that the same mechanic was live in Retail and in part of the Classic line for years without clearing the supply out of those realms, which is the detail that keeps coming back in every thread.

A share of that demand never reaches a listing at all, and loot handed out without bidding moves gear through a channel the shelf never records.

Both halves of that are opinion about cause and effect, not something Blizzard has published numbers on, and nobody posting in those threads has the data either.

What the decision leaves behind is a shorter list of levers. Without a sanctioned conversion step, the price of gold is set by how much of it enters the world and how much leaves

One of those sinks is charged a few coppers at a time, and what class trainers take by 60 adds up to more than most levelling plans allow for.

Both halves of that are counted per side rather than per realm, and faction choice as a money decision decides which buyer list a seller is posting into.

That list is short enough to be worth taking apart on its own, and gold priced without a Token is where the longer version of this argument lives.

, which puts the weight on sinks and on whatever action Blizzard takes on the supply side directly. That second part is a policy matter rather than an economic one, and the server-side record of a trade is the material it works from. For a buyer the practical consequence is narrow: there is no first-party seller to compare against, so every price on the realm is a player price.

One trade is priced by attention rather than by stock, and enchants written onto vellums would change which of the 2 the shelf records.

Where that demand lands is one question, and how a gold transfer looks from the server side is another one entirely.

It is worth saying plainly how players are reading that sentence. "No plans" is a statement about the present, not a promise about the rest of the game's life, and a large part of the forum crowd treats it as a delay rather than a decision. The threads that follow it tend to argue about when a Token arrives, not whether one does, and some of them tie it to a content patch or to the first wave of subscription churn. None of that comes from Blizzard. There is no announcement, no date and no beta build showing the feature, so every arrival schedule being passed around is a guess made by players.

One of those sinks runs on a fixed schedule, and consumables, repairs and loot rules decide how much a raiding character gives back each week.

What players are expecting. The forum does not only predict a Token. The same threads expect paid services to follow it in some order, and the list they keep repeating is gold sold outright, a level 60 boost and the usual account transfers, with predictions about paid services measured against the current Forever shop catalogue rather than against anything Blizzard has announced. The counter-argument in those threads is short and it comes from people who are not against buying gold at all: a Token is the same trade as any other, with the publisher standing on the selling side of it. One post put the Retail version plainly — a player sitting on 2000 g covered the rest of what a purchase needed with a single Token and never touched the auction house to do it.

What that would change on this realm is narrower than the threads suggest. A Token moves gold that already exists from one player to another; it does not mint any

Part of what leaves the realm is spent on carrying capacity instead, and bag and bank slots are a sink almost nobody counts.

, so the faucets and the sinks described below would stay exactly where they are. What changes is the comparison. Today every price quoted on the realm is set by another player, and an economy priced without a Token has no first-party number sitting next to it. With one in the client there is a published figure to measure against, and the demand that currently spreads across the auction house concentrates on whoever is converting. That is our reading of it, not something Blizzard has described.

The honest position is that the current state is confirmed and the future is not. Blizzard has not said a Token will never ship, and it has not said one is coming either, which means an economy priced without a Token is the situation we can describe today rather than a permanent rule. Everything on this page assumes that situation holds. If it changes, the sinks and the supply side stay where they are, but the comparison a buyer makes changes completely, because there would finally be a first-party price to measure against. Until that happens, every price on the realm is set by another player.

That argument is not hypothetical somewhere else. In Retail the line between what is earned and what is bought is being tested right now, with race-themed cosmetic decor appearing in the shop after a public statement that items of that kind would stay out of it.

Which of those purchases exist in this line is a much shorter list, and what the Forever shop stocks stops at cosmetics and account services.

Whether that counts as a broken promise or a narrow change of scope is the part players are shouting about, and we have no inside view of how the decision was made. What matters here is only that the boundary moves, and that it moves without much warning when it does.

Forever sits on the other side of that line today. There is no shop gold, no level 60 boost and no Token on the realm, so none of the 3 feeds the economy from a source outside the player base. That is a description of the current build, not a guarantee about later patches, and an economy priced without a Token is the only version anyone can measure at the moment. If the boundary moves here the way it has moved in Retail, the sinks stay where they are and the comparison a buyer makes is the thing that changes.

The waiting is not only about a Token. The same threads list 3 things players expect to show up in the shop sooner or later: gold sold directly, mounts, and a boost that drops a fresh character at level 60.

Whatever those levers turn out to be, they are counted twice over, and why gold cannot cross regions keeps the 2 price curves apart.

The raid side of that demand did not disappear with the pot, and what replaced the GDKP pot is a set of fees and loot agreements rather than a single format.

None of those are in the store today, and nothing in the beta builds points at them. That absence is not a side note on this page, it is the base the whole rate calculation sits on, because each of the 3 would feed the realm from a source no player controls.

It is worth being clear about which part of that is knowable. The current shop contents are a fact anyone can check in the client.

One channel moves coin sideways without touching the shelf, and the market for class utility leaves no auction record behind it.

The arrival of anything else is a forum expectation with no date behind it, and we are not going to put one there. What we can say is how the reading changes if it happens: gold from a first-party seller would set a ceiling, mounts would pull coin out of the same wallets that fund raid supply, and a boost to 60 would cut the levelling demand that the weekly consumable restock currently rests on. Until any of that lands, an economy without shop gold is the only one we are describing.

The list of separate markets grew late in the schedule, since the RP-PvP ruleset market opens on the same day as the rest with none of the same history.

Ruleset is not the only line running through that list. Oceanic and North American players land on the same shared realm, so both regions bid into one auction house instead of 2 separate ones.

Much of that bidding is done by one player wearing several names, and a dedicated bank character is where the stock sits between listings.

The effect is a demand curve that does not line up with local play hours: OCE posters describe a thin shelf through their evenings, with stock cleared and relisted by buyers logging in later from NA. Prices level out across the 2 regions because there is nowhere else to list, and the cost of that levelling is carried by whichever side is asleep when the restock goes up.

Where that stock is gathered matters as much as when it is posted, and gathering under PvP rules runs on a different clock than the same route on a quiet realm.

Some rewards never reach that shelf in either region, and rewards handed out at Legacy thresholds are earned on the character rather than bought from one.

The thread that follows this complaint tends to end in the same 2 proposals: a hard ceiling on what a listing may ask, and enforcement aimed at the demand side rather than at sellers. Blizzard has not commented on either, and neither one is something we can measure from the realm. A ceiling moves trades into whispers instead of removing them, and demand-side policy is a rules question, not a price lever that turns up in a scan. What does turn up is that the regional gap looks like a timing problem rather than a supply one: the same goods arrive, just not in the hours OCE players are shopping. Until posting hours spread out, gold priced without a Token stays a function of who is awake when the shelf refills.

Demand can also arrive from outside the game entirely, which is what happens when an outside prize pool gives a group of players a cash reason to buy.

One supply line reaches that market with no gathering node behind it, and fish stock on the realm is limited by hours rather than by competition for spawns.

One trade reaches that market twice over, and enchanting scrolls on the shelf compete with the same work sold as a service in chat.

Other trades sell the second half only, and crafters paid for work alone never put stock on the shelf at all.

One stream into that shelf arrives by accident rather than by plan, and tradeable drops from ordinary mobs reach the auction house with no profession behind them.

Some coin never reaches a buyer at all, and gold sent across a roster moves without touching the shelf.

That demand has an ordinary version underneath it, since what a PvP evening costs is spent on the same shelf every week whether a bracket is running or not.

The dungeon version of that shelf has a second line attached to it, and repair bills split 5 ways move with how the run went rather than with what was planned.

The raid version of the same shelf is larger and steadier, and the weekly raider bill is the demand that never stops arriving.

One trade bills for attendance rather than for stock, and enchanting sold as a service never puts an item on the shelf at all.

Another slice of demand is paid for the model rather than the stats, and buyers shopping for a look price items the rest of the realm would vendor.

That billing has its own shape, and the enchanter's service model prices travel and materials as separate lines.

Other trades hit the same wall one recipe at a time, and crafted output that cannot be sold is a larger share of a profession list than most plans allow for.

That wall has a market on the other side of it, and commission work priced per order bills for access to the recipe rather than for the item.

The quieter version of that spend runs all week, and the upkeep bill per class is paid by characters who never queue for a bracket.

One slice of that spend never reaches the shelf at all, and resurrection and portal reagents are bought from a vendor at a price no seller can undercut.

Supply arrives on that shelf slightly unevenly as well, and which races gather a tier earlier explains part of the gap between the 2 factions.

None of that demand depends on a Token existing, because at 60 the bill is a repeating one rather than a single purchase. A character with its gear list finished still spends every week: flasks, potions, food, oils, reagents and repairs are used up during the lockout and ordered again for the next one.

Some rosters answer that weekly bill with more hands, and a second subscription bill is the first line to weigh against the extra farming slots.

A smaller line on that weekly bill never reaches another player at all, and fares paid to flight masters leave the realm the moment a route is picked.

That is the structural answer to the question of why gold is still wanted on a realm with no first-party seller. A one-off buy ends when the item is in the bag; the consumable shelf at 60 bills the same raid group again on the same schedule, and the group has to find the coin somewhere between resets.

Before any of that, the faucet most characters actually see is quest gold, and the forum thread on it asks for the rewards to be raised on the way to 60. The reasoning is narrow: a player who clears a zone end to end still reaches the riding trainer short, and the gap is covered by selling off quest greens and grey vendor trash rather than by the quests themselves. The same thread carries a second idea, that a share of each quest reward could be credited straight toward riding skill instead of landing in the bag, so the purchase arrives at the level it is meant for.

The reply under it is the one worth keeping. Money in Forever is not expected to come from quest text at all: repeatable dungeon runs and farm routes are where coin actually accumulates, because they scale with hours played while a quest pays once per character and never again. By that reading, raising quest rewards would move the bottleneck rather than remove it. Neither side has published figures, and we do not know whether reward values have been touched for Forever at all. Until a character can afford the trainer, the gap is covered the slow way, and fares paid to flight masters are the line that takes its place on the ledger.

There is a claim going around the forums that the weekly bill has a ceiling built into it. The version being repeated is that a reagent vendor in the goblin town sells the common crafting inputs at a fixed NPC price, which would mean no player listing can hold above that line for long: the moment a seller asks more than the vendor, the buyer walks to the vendor instead. We have not been able to confirm it. Nobody in those threads has pointed at a file, a datamined vendor list or a beta screenshot, and the posts that describe the stock describe it differently from each other.

Treat it as a forum claim rather than a fact. What is worth keeping from it is the shape of the argument, because a fixed NPC price behaves differently from a player price: it does not move with supply, it does not care how many gatherers are working the node, and it sets the same number on day 1 as it does a year later. If such a vendor exists, the affected lines on the weekly consumable bill get an upper bound and the rest of the shelf does not. If it does not exist, nothing on that shelf is capped at all and every line is set by whoever is selling that week. Both readings are open until somebody produces the file.

The argument that follows on the forums is about time rather than about money, and it is worth stating in the terms the posters use.

The same time argument runs well below the cap, and coin earned while levelling is counted in hours before it is counted in gold.

The hours needed to farm a lockout of consumables are compared against the hours of the raid night itself, and when the first number is larger than the second the complaint is that the farm has become the main activity and the event the side one. That comparison is the motive usually given for buying gold with cash: the purchase is described as removing the farm from the calendar, not as removing the raid. Whether the farm is actually longer than the event is an opinion with no published figure behind it, and both sides of that thread are arguing from their own play schedule.

There is a third position in those threads, and it is worth recording because it shapes how people plan rather than how they spend. A large share of the posters do not believe the no-Token line will hold. The evidence they bring is circumstantial and always the same 2 items: the shop button sitting in the game interface, and the pre-order editions that already take cash for in-game goods. The reasoning is that the plumbing for selling things is built and switched on, so the only missing part is a decision. None of that is a statement from Blizzard about a future Token or a future boost, and nobody quoting the shop button has been told anything the rest of us have not.

The practical effect is that part of the market is pricing in an arrival date nobody has. Some players say openly that they will stop playing on the day a Token or a paid boost goes live, and treat their current effort as provisional because of it. Others take the opposite lesson and farm harder early, on the theory that the window before any first-party seller exists is the one where player-set prices mean the most. Both are forecasts about a decision that has not been announced, and the only confirmed position remains the one at the top of this guide: an economy with no Token is what the realms open with.

Real money is in the game, it is just pointed somewhere else. There is no Token and no boost, but Blizzard still sells cosmetics and mounts for cash

Cash also changes hands for tools rather than coin, and addons sold for real money sit beside the shop without touching the auction house.

, and those two facts get folded into one argument in every monetisation thread. The line is what the purchase does. A shop item changes how a character looks or how it moves; it is not tradeable, not listable and not convertible into coin. Nothing bought with a card lands on the auction house, and nothing bought with gold is available for a card.

In the client this lives behind a single menu entry called Shop, and it is worth saying plainly what that button does, because it is the thing most often mistaken for a gold channel.

The rest of that storefront is worth taking apart on its own, and what the Shop button bills is a shorter list than the threads around it suggest.

The window a legitimate trade happens in is a separate matter, and how the sale window works decides what a buyer sees before the price does.

It opens a storefront billed to the account, not to the character. What comes back is a cosmetic or a mount delivered into the collection, and the collection is not an inventory: there is no slot to drag it out of, no listing to post it under and no vendor that will pay coin for it. The auction house and the Shop never touch each other, in either direction.

None of this settled the argument before launch. Forum threads from the beta period describe shops already quoting Forever gold while the realms were still a test build, which is the detail most of the posting is reacting to: the trade existed before the live economy did. Nobody in those threads has a figure for how much changed hands, and we do not either, so treat it as what it is — players reporting what they saw on the beta, not a measured volume.

It is just as useful to list what has not been addressed. The interviews and the beta notes cover the Token decision and the level cap; several of the questions players keep asking were never answered, and the honest position is that the answer does not exist yet rather than that it is hidden somewhere.

• Inflation over the life of a realm. No target, no stated tolerance, no description of what would count as too much gold in circulation. The sink side is visible in the client; the policy behind it is not.
• Node respawn under megaserver populations. Gathering density per player on a realm that holds far more characters than a 2006 server has not been described. Whether respawn rates were retuned for that, or left as they were, is unknown to us.
• Vellums. Whether enchanting output can be sold as a stocked item rather than as attendance changes an entire trade, and nothing published settles it. Until it does, enchanting billed as attendance is the version to plan around, and enchanting sold as a service is how that shelf behaves today.
• Automation. Blizzard has not published a position on automated gathering for Forever specifically. What the server records about a trade is a separate matter from what any policy does with it, and the server-side record of a trade is the only part of that we can describe.

Anyone filling those 4 gaps with a confident number is guessing. We would rather mark them open and revisit them when something is actually said.

The list of absent faucets is longer than the Token and the boost. There is no repeatable daily quest on the Forever realms either: the quest log at 60 pays once per quest and then stops, and nothing on the map resets overnight into a fixed coin turn-in. That matters for the same reason the Token did. A daily reward is paid by the world rather than by another player, so every character that logs in adds coin that came out of nobody else's pocket, and it keeps arriving whether the auction house is busy or empty.

A large part of the posting on this asks Blizzard to leave it that way rather than to add dailies later, and the argument runs through the same sinks and sources that gold priced without a Token depends on. We have not seen Blizzard answer that request one way or the other, so treat it as a community position and not as a published design decision. For a buyer the practical reading is narrow: nobody on the realm has a guaranteed daily income, so the coin on the shelf was earned through a trade, a drop or a gathered stack.

The demands that come out of it split 3 ways and pull against each other. One group wants enforcement pointed at the buying side rather than at the sellers, on the theory that supply follows demand. A second group wants deeper sinks instead — more repair cost, more consumable burn, more coin leaving the world per raid night — and treats enforcement as a side issue. The third group only cares that the Token stays out, and reads any talk of a sanctioned channel as the first step back toward one. The arguments for an economy without a Token are the same ones Blizzard repeated at BlizzCon, which is why that camp currently thinks it has won. All 3 positions are player opinion; what Blizzard has actually committed to is the shorter list above.

So the transaction has no second leg. Money goes in, an appearance comes out, and the amount of gold on the realm is exactly what it was before the window opened.

One crafted market opens before the raid calendar does, and bags as a traded good have buyers in the first hours of a realm.

Those markets are less separate than the ruleset names suggest, and a single auction house across rulesets is where their prices meet.

Demand on the other side of the ledger does have a date on it, and enchanted pre-raid gear concentrates a whole realm's buying into the same few weeks.

With no Token in the game there is no sanctioned step that turns a card payment into coin on a character, which is why a cosmetic sale is not a faucet and a refunded one is not a sink. Where cash is permitted to go instead is a rules question rather than an accounting one, and the RMT rules at launch cover the part of it that involves gold. Blizzard has not published the full launch catalogue for the Shop, so treat any circulating item list as a prediction.

For a page about faucets and sinks that distinction decides whether the shop belongs here at all, and it does not. A cosmetic sale creates no gold and destroys none, so it moves no price on any realm.

One sink works in the opposite direction at the bottom of the market, and deposits lost on expired listings take coin out of the realm a few coppers at a time.

Coin also sits still for long stretches inside one account, and stock parked on a bank alt is supply that has not reached the market yet.

Where cash is allowed to go instead is a policy question rather than an economic one, and what the RMT policy covers is the part of it that touches gold.

The same arithmetic covers every other way money reaches Blizzard in Forever. Access is billed as a subscription, so the charge buys time on a realm rather than anything a character can carry. Editions and bundles sold around a launch behave the same way: what they hand over is cosmetic, plus entry to a test build where beta access is part of the package, and the upgrade button in the launcher only moves an account from one edition to another. We have not seen a published breakdown of what every Forever edition contains, so any item list circulating now is a prediction rather than a confirmed one.

Retail is running the opposite experiment at the same time, which makes the contrast easy to see. Housing decor there is stocked from a storefront with its own paid currency, sold at 100 units for 1 USD or 1 EUR, and a house is priced at 25 USD. Nothing of that shape has been announced for Forever: no token, no boost, no second currency sitting on the account alongside the subscription. What a Forever payment buys is time on a realm, and the catalogue of what else might be for sale has not been published.

That is why the two economies cannot be read with the same chart. In Retail a decor price is a number Blizzard sets and can change, and it answers to a currency exchange rate rather than to supply on a realm. In Forever every price on the auction house is set by the players posting and bidding, which means a card payment never shortens the distance between a buyer and an item. Where real money is permitted to go instead is covered by the real-money trading rules, and that is a policy question rather than an accounting one.

None of that is a faucet and none of it is a sink. A faucet puts coin on the realm that did not exist a minute earlier; a sink removes coin from the realm permanently.

One trade moves coin with no listing behind it at all, and an enchanter paid in the trade window shifts gold between players rather than creating it.

One corner of that ledger has a limit above it, and the cap on reagent prices is set by buyers walking away rather than by the sellers.

One loop of stock and payment never touches the shelf either, and the guild side of the economy runs on bank tabs and whispers instead of listings.

A subscription charge, a bundle and an upgrade do neither, because the ledger they settle against is a billing account, not a character. The realm total is identical before and after, which is why none of these purchases appear anywhere in the supply side of this page. The only thing they change on the auction house is indirect: more accounts logged in means more hands gathering, crafting and bidding, and that is a population effect rather than a monetisation one.

Coin also changes hands at the very bottom of the ladder, and quest materials sold between strangers is where most characters first pay somebody else for time.

The one place it touches the ledger sideways is time: a mount is a mount, and time lost to ground travel is the same arithmetic whoever paid for the ride. Blizzard has not published a full launch catalogue for the shop, so any specific item list circulating now is a prediction rather than a confirmed one, and we will treat it that way.

Blizzard's reasoning was about design rather than economics — the road from 1 to 60 is the game, and selling a way past it sells the game away.

How that demand behaves is also shaped by what buyers believe will happen to them, and sanction claims on the beta are mostly retellings with no source attached.

Some of that coin is created rather than moved, since gold paid out by quests comes from the game instead of another player's bank.

Part of that demand never reaches a listing either, since portals, summons and opened lockboxes are paid for in whispers and recorded nowhere.

Part of that demand is one player paying another for a single click, and opening a locked box is settled in a trade window with nothing posted anywhere.

The economic consequence is larger than the design one. In retail the Token is a pressure valve: anybody who wants gold badly enough buys one from Blizzard, and that demand never reaches another player.

Demand is also split before it reaches any seller, and prices inside a single faction are set by the share of the realm that can see the listing.

One supply line never reaches the listings at all, and a trade with no listings settles its whole volume in trade windows.

Some of that demand only moves between shelves rather than growing, and a raid week with no buff chasing shifts it from travel time into crafted stock.

Some demand disappears without any valve at all, which is what Legacy perks do to consumable demand at the cheap end of the shelf.

A smaller loop runs on the same logic away from the capitals, since money moving through camp stations has its own input, output and upkeep.

Some of that demand leaves no record on either side, because enchanters selling without a listing settle the whole sale in a trade window.

The reason that sale stays off the books is mechanical, since enchanting with no vellum leaves nothing a listing could hold.

Part of that demand repeats on a fixed rhythm rather than once per character, because gold raiders pay alchemists comes back with every lockout.

That rhythm has a start date, and the first raid week bill is what arrives before any lockout settles into a routine.

Some of that farming time is spent moving rather than earning, which is the arithmetic behind crossing the continent on foot.

Part of that demand comes from players who never enter a raid, since what a battleground regular spends weekly is settled in consumables, repairs and one weapon decision.

The raiding half of that shelf runs on the same rhythm, and consumables, enchants and repairs come due once per lockout.

How much each farmer brings home also depends on the character they rolled, which is the arithmetic behind racial perks and route yield.

Part of that demand is one player buying for themselves twice over, which is what a second and third character adds to the same listings.

A guild bank is the other route by which supply reaches a raid group without ever passing through a listing. Gatherers on the roster deposit materials, a guild alchemist or enchanter turns them into flasks, potions, food and enchants, and the finished stock goes back out to members at whatever the officers decide to charge, which is frequently nothing. Recruitment posts state this openly: raid teams advertise consumables supplied and crafting covered as a reason to join, which is the same bill as consumables, enchants and repairs settled out of one shared pot instead of 40 separate wallets.

The effect on the open market is not that the demand vanishes but that it moves one step back along the chain. A guild crafting its own flasks still needs herbs, so the raw material end of the market can stay busy while the finished-goods end thins out, and the margin that would have gone to a crafter posting publicly stays inside the roster instead. We have no way to measure what share of consumable volume runs through guild banks on any given realm, because none of it is visible from outside the guild, and anyone quoting a figure for it is estimating.

Part of that demand never buys an item at all, because gold paid for a dungeon role buys another player's evening instead.

Demand that cannot reach a seller behaves the same way, which is the case for an account with its transfer channels closed.

Part of that demand never shows up as a finished item at all, because soulbound crafted gear pulls its spending into materials instead.

Part of that demand is account progress rather than gear, because buying Legacy progress with reagents turns bought stacks into points.

A steady share of that demand is crafters restocking, since stacks consumed by profession ranks are bought from other players before anything sellable comes out the other end.

Who absorbs that demand depends on the supply side head count, since a trade with few practitioners sets its own price and a crowded one does not.

A share of that demand is simply restocking, because water, food and totem reagents are bought from other players as often as from vendors.

One buyer on that shelf drinks more than the rest, since what an AoE mage spends is settled in stacks of water per evening.

How the other side of that demand is policed on the test client is covered in what the trade window warns about.

Close the valve and the entire demand for gold lands on 2 places — the people who farm it, and the people who sell what they farmed.

Both of those places assume the pipes are open, and a freeze on gold movement is what the economy looks like when they are not.

Inside a group the same closed valve shows up as an argument about drops, which is why need, greed and soft reserve now do the work a gold pot used to do.

The same group settles a second question before the first pull, since coin offered to fill a slot is priced by whichever role is missing.

Some of that demand is settled without any coin at all, because a tank paid with a reserved drop leaves no listing and no transfer behind.

The forum argument about the missing Token rarely stays on design, and the 2 claims inside it are worth separating. The first is that a Token replaces third-party sellers: the buyer goes to Blizzard, the demand is absorbed at a price the publisher sets, and the gold comes out of another player's stock through a listing. The second claim is that a Token suppresses botting. Only the first describes something the mechanic does directly. A Token changes where a buyer goes; it does not change how much an unattended farming route yields per hour, and the route is what a bot is built around.

Part of that farmed supply comes from one person in several windows, and several accounts under one player adds stock without adding a second player.

Demand also stalls when the coin cannot move, which is what a hold on account funds does to both sides of a listing.

For the price of gold in Forever that distinction matters more than the fairness argument. With no publisher-side supply, every unit that changes hands was farmed by somebody, and demand that would have gone to a Token sits in the same market as a raider restocking consumables

Before any of that demand settles, somebody has to hand the coin over in game, and how bought gold arrives is the step where a deal either completes or stalls.

and a crafter buying stacks to level a trade. The effect is on the floor rather than the ceiling: the cheapest gold in the world is still somebody's farming hour, and nothing underneath it is sold at a fixed rate. What we cannot put a figure on is the share of buyers who would have taken the Token route if it existed, so any estimate of how much of that demand transfers is a guess, including ours.

How much those farmers bring home is partly a build decision, set by Legacy points spent on yield rather than by the route alone.

Quest rewards are the first faucet anyone touches, and in Forever that faucet is not evenly spaced across the map. One player's count of the beta leveling zones came out at roughly 200 new quests and more than 80 reworked ones in the 1-30 range, against the 1000 new quests promised for the game as a whole. That is a player tally rather than a published figure, so treat the exact numbers as approximate, but the shape of it is hard to miss: Elwynn Forest, the Wetlands and Mulgore carry far more of the additions than their neighbours do.

For the early economy that skew matters more than the total. A zone with 30 extra quests pays out more coin per hour than a zone of the same level that was left alone, and it also holds more players, which means more competition for the same herb nodes and skinnable mobs. Players who follow the dense zones bank more quest gold and spend more of it on the spot; players who take the thinner route spend more of their leveling time moving, and crossing the continent on foot is the part of that bill nobody lists.

One faucet still runs regardless of what players do, because gold paid out by NPC vendors enters the world every time something is sold at a counter.

Two later changes move that faucet picture again. The first is the upgrade of plain white drops in the starting zones to green quality: the same kill table now produces items that carry stats, which means they are worth more at a vendor and some of them are worth listing instead. The practical effect is on the first hours of every character rather than on the raid economy — levelling gold arrives slightly faster, and a share of what used to be vendor fodder now competes with the low end of the auction house. How large that share turns out to be depends on how many players route those greens to a counter out of habit, and we have no figure for that.

The second change works in the opposite direction. Automatic Temporary Economy Holds can suspend an account's gold movement — trades, mail and auction proceeds — without touching the character itself, which means the gold exists but cannot change hands while the hold is in place. For the market that behaves like demand and supply being removed at the same time: a buyer under a hold cannot pay, a seller under one cannot collect. Holds are applied by system rather than by request, so neither side can plan around them, and the usual advice applies — keep deliveries simple, in person, and do not build a trade around a transfer that has to happen on a fixed schedule.

One route into that faucet is closed. Buying stock from one vendor and selling it to another for more than it cost used to be a way of turning shop inventory into gold with nothing farmed in between, and the sell prices have been adjusted so the loop no longer pays out. What is left is the ordinary direction: something that dropped in the world, or came out of a craft, handed over at a counter for a fixed amount.

The mechanics of the closed loop, and the part of it that still works, are set out in how the vendor shuffle worked.

Any list of faucets written before that change is counting a source that is not there.

The distinction matters for how the faucet behaves. Vendor arbitrage had no ceiling — it scaled with the time a player was willing to spend clicking between 2 shops, and it did not care what the rest of the market was doing.

The opposite move runs into limits of its own, which is what happens to attempts to corner a market.

Supply has a similar shape when one player gathers on several characters at once, which is what a 4-character farm roster puts into the same listings.

Selling loot to a vendor is bounded by what the world actually drops, so it tracks the number of people out killing things rather than the number of people running a route. We have not seen a full list of which items were repriced, only that the resale gap is gone where players were using it, so treat the exact coverage as unconfirmed.

Part of that demand comes from characters buying twice for the same slot, which is what healer off-spec and dual spec adds to a levelling bracket.

How many of those coins get picked up depends on the ruleset the farmer chose, which is the subject of farming under open-world PvP.

The missing Token is also the most argued line on the Forever forums, and both sides of that argument are worth stating plainly, because they explain why the absence holds. The side asking for one wants the exchange to sit inside the game: players who intend to trade cash for gold will arrange it one way or another, and a first-party route would keep that trade in front of Blizzard rather than outside it, with no paperwork between two strangers. That argument is about where the trade happens, not about whether it happens at all.

The objection runs the other way and is about supply rather than demand. A Token publishes an official conversion rate, and a published rate hands anyone farming gold at industrial scale a guaranteed buyer and a payout per hour they can plan around — which is the one thing automated farming needs in order to scale. Remove the rate and the same operation has to find buyers itself, at whatever the market gives that week. Both positions are player arguments rather than statements from Blizzard; the confirmed part is only that no Token is planned, and the supply half of the dispute is picked apart in the bot question below.

No Token does not mean no money moves between players and Blizzard. What cash buys in Forever is cosmetic, plus a race that is purchased rather than unlocked in game, and none of that pays out in coin. The two columns stay separate: a player can spend on an appearance or on access to a race and still walk into the auction house with nothing more than what their own hours produced. That is why the gold price here is not anchored to any official conversion rate — there is nothing sold for cash that a seller can convert back into gold.

The one point where the columns touch is the paid race itself, because a race is not only a look. Whether anything attached to it changes output per hour is the argument that keeps returning, and it is the same arithmetic as racial haste and farming yield. We have not seen that dispute settled, and the numbers people trade in it are estimates rather than confirmed figures. What is clear is that it sets the ceiling on how far real money can reach into the gold price, and that ceiling sits far below what a Token would open up.

One line travels further than any of this on the forums: that gold has been cut and that experience is worth 48% of what it was in Classic. It is 2 different claims stapled together, and they are not checkable in the same way. The first is: anyone with beta access can open a vendor window, compare what an item sells for against the Classic figure and post the pair, which is how the differences in what NPC vendors pay out were found in the first place. That part is item by item and either holds up or does not.

The second part is a single percentage applied to the whole game, and we have not seen Blizzard publish a levelling rate of any kind. What circulates instead is player timing over a small number of characters, in a handful of zones, on a build that is still changing. Experience per kill and per quest varies with mob level, rested state and which zone you picked, so one averaged figure hides more than it reports. Our position is that the vendor comparisons are worth reading and the percentage is a claim with unknown provenance — and that for income it matters less than it sounds, since a slower curve changes when a server reaches max level rather than what an hour of farming is worth once it gets there.

Both of those arguments land on the same counter. Forever runs its realms as a mega-server, and the auction house is shared across the whole ruleset rather than split per realm — one listing feed, one set of undercuts, one price for an item no matter which starting zone the seller logged in from. The old habit of buying a stack cheap on a quiet realm and reselling it on a crowded one has nothing to work with, because there is no second market to carry it to. What is left is depth: a single pool of buyers and sellers deep enough that one large order rarely moves the price for long, and thin enough on niche reagents that a single stocked crafter can hold a line for days.

That visibility is also where the loudest complaints come from. On a shared market everyone watches the same listings, so a seller who posts in volume week after week is seen by every buyer on the ruleset at once, and the forum thread that follows usually claims the gold behind those listings was bought for cash rather than farmed. The listings themselves do not say either way — posting volume is the one thing the auction house shows, and the source of the stock is not part of it. Treat those threads as what they are: players reading the same feed you are, and guessing at the half of it that is not shown.

The boost half of that sentence removes a market that existed elsewhere. In Classic the going rate for a mage pulling a group through a dungeon settled around 2000 g

Supply has a second entrance as well, because charge-limited mining and digging tools let characters with no gathering trade add material to the listings.

, paid by players who wanted the levels without the hours, and that coin moved from a buyer to a mage instead of entering the world. Forever does not put that trade on the table at all.

The beta makes the absence look smaller than it sounds. Testers report that solo leveling currently moves faster than a paid run would, so the segment that used to hand over 2000 g has little reason to form.

What that segment buys when it does form, and what the reported dungeon experience change does to the rate, is set out in paid levelling services in Forever.

What that missing segment does to a levelling character's spending is worked through in how levelling speed moves costs.

Whether that holds after launch is unknown, since leveling rates on a test client are not a promise. What does carry over is where early gold goes instead: the first real outflow is arming a character before 20, not buying a way past it.

The request for a Token has not gone away just because the answer was no, and it is worth being precise about how it is argued.

The same boards argue about the other end of the pipe, and what the transfer threads report is worth separating from what they assume.

The same shortage of hours is what pays for warlocks selling travel time to groups that cannot wait for a flight path.

On the beta forums one morning produced 2 separate threads asking for one, and neither of them was framed as a wish to skip the game.

Those same boards have carried a separate complaint for 3 days about coin held in beta mail, which is a client problem rather than an economic one.

Both used the same sentence in different words: there is no time to farm when automated gatherers are moving from node to node faster than a player can ride between them.

The same shortage of hours supports a channel where nothing is delivered at all, which is the market for player services.

The argument is about the clock, not about effort.

That matters for how the pressure should be read. Demand for a Token is being fed by prices, and prices are being held up by supply that is collected without anyone sitting at the keyboard

Part of that restocking is class-bound rather than general, since the reagent bill casters pay lands on a narrow shelf of vendor goods.

— which is the same mechanism behind automated gathering on quiet routes. A player who looks at a material cost, divides it by what an hour of honest gathering returns and decides the hour is not available is making an arithmetic decision. Whether Blizzard treats that arithmetic as a reason to reopen the valve is unknown, and nothing said so far suggests it does.

The other half of the demand picture is how fast gear stops being worth buying. Blizzard has described the intent as keeping the power jump between content releases as small as it can be made, rather than handing out a new set that makes the previous one worthless on the day it lands. The stated consequence is that a best-in-slot item can keep coming from an older tier, and that a crafted piece or a drop bought early does not automatically become vendor fodder when the next raid opens. That is design intent as it was described, not a patch note, and the actual size of each step is only visible once the content is live.

For the market the effect is on the clock rather than on the price level. Slow devaluation means a crafted weapon or armour piece holds a buyer for longer, and the orders behind it keep running, which in turn keeps demand under stacks consumed by profession ranks instead of letting it collapse the week a tier changes. Consumables behave the opposite way: they are burned each raid night regardless of how good the gear is, so their demand tracks raid attendance rather than item level. One thing that has not been answered is whether any kind of soft reset or stat squish is planned later on. No decision has been announced, and anyone budgeting around one is guessing.

A raid night is a faucet of its own, and it runs in 3 separate ways at once. Bosses drop coin directly, which is split across the group and arrives as new money rather than money moved between players.

Part of that new money never reaches a personal wallet at all, because repairs paid from guild funds are settled out of the same night's takings.

Trash and the parts of the loot table nobody rolls on turn into coin at a counter, so the real payout of a night depends on sell prices at NPC counters as much as on what the boss dropped. The third part is the gear that sells to other players, and that half is not a faucet at all — it only moves gold that already exists from a buyer to the raid.

That split explains the argument about a second lockout. Raiders who want one are not asking for more loot chances so much as for a second night of coin and vendor material, because the consumable bill for a week is paid whether the raid clears once or twice. The counter-argument is that a second run doubles the inflow without doubling the sink, which pushes material prices up for everyone who does not raid. How the group divides the result is a separate decision, settled before the pull by need, greed and soft reserve. Nothing has been said about whether a second lockout is being considered, so treat the whole discussion as player opinion.

One faucet that most modern servers lean on is absent from everything announced so far: no repeatable daily quest line has been described as a standing source of coin in Forever.

At the other end of the level range sits the first real outflow a new character meets, arming a character before 20.

That is worth stating plainly, because a daily loop is the one faucet that pays the same amount every 24 hours regardless of what the market is doing, and nothing in the BlizzCon material or the beta notes promises one. The absence is not a confirmed design rule either — it is simply a gap in what has been said.

Players fill that gap with the argument they already had. On the forums a repeatable reward loop is treated as the clearest driver of inflation, since coin arrives on a timer while the things it buys still have to be gathered by hand, and the REAL system from Season of Discovery is the example people reach for when they want to argue either side. Its status in Forever has not been addressed by anyone at Blizzard, so read the whole exchange as player opinion rather than a plan. Until something is said, the reliable inflows remain quest rewards, drops and sell prices at NPC counters.

Coins drain out at a class trainer as well, where changing talents at a trainer is charged per reset rather than once.

Blizzard made the same point from the other side in interviews: players are expected to hold considerably more gold than they did in original Classic, because drops, quest rewards and vendor payouts

How much of that larger balance survives once a second and third character need storage and riding is covered in an alt's starting capital.

Some of that larger balance never settles, because what drinking between pulls costs is paid again every session rather than once.

across the new content were written upward rather than copied across. No figure was attached to that statement, so read it as a direction and not as a multiplier.

The second half of it matters more for prices. The routes that made Classic gold predictable — the known farms, the circuits, the respawn timers — do not exist for the new zones yet, and nobody has had the time to work them out. A world where more gold enters per hour while almost nobody farms efficiently prices differently from Classic in both directions at once: materials stay expensive because supply is unorganised

Supply that regrows on a clock of its own prices differently again, which is the case for herbs grown rather than gathered.

The clearest example of that on the beta is a hide nobody is collecting, where the recipes written around it stop being worth making.

, and coins are less scarce than an old Classic price list would suggest. That is one reason beta prices should not be compared straight across, alongside item and recipe nerfs between builds.

There is no second source. No boosted level 60 arrives with a starting purse, no beta character carries a balance over the wipe, no cash shop prints a single copper.

Some players close off even that market by choice, and a self-found character's running costs are then paid entirely out of drops and counter sales.

That single source is also the reason why bots settle on fishing rather than on any task that needs a decision made.

That single source is also why capital stays locked to one realm until somebody moves it by hand.

How much any one character pulls out of that single source is partly a build question, set out in Legacy points as an earnings decision.

Coins drain out through small recurring charges as much as large ones, including the repair bill after a death.

How large that charge gets is partly decided at the graveyard, which is why a corpse run or spirit healer is an economic decision rather than a mood.

Coins leave the world as plainly as they arrive, including through the price of bank slots, which no sink list counts.

There is a second balance, though it is not gold: a second wallet beside the coins is earned and spent on its own terms.

A fresh economy with no faucet except play is rare enough in 2026 that most players have never actually seen one.

Gold that enters at the top does not stay at the top. The players holding the largest balances are not buying a second set of boots with them; they are buying the things that get used up and bought again, which means consumables absorb most of the new money before anything else does. Mats for a gear piece are bought once per character, while flasks, food, potions and reagents are bought every raid night, so a faucet that favours the organised end of the playerbase shows up in those prices first. The same effect runs through the restock bill between sessions, which is paid by everyone but set by whoever is willing to pay the most for it.

One thing worth stating plainly: there is no trade layer under the auction house. Everything that is not a face-to-face trade passes through the capital auction house, which means a single order book sets the reference price for every zone, including the ones a day of flight away. The gap between what a material is worth where it is gathered and what it is worth where it is used is not priced anywhere, because nothing exists to price it.

Players have been asking for that layer. Forum threads have floated guild-run traders stocked by the guild bank, and caravans that would carry goods out to the frontier and sell them at a markup over the auction house listing, with the markup covering the trip. None of this has been announced, none of it is in the test client, and Blizzard has not commented on any of it, so treat the whole idea as a wish list rather than a plan. For now the only prices set outside the auction house are the fixed ones, which is to say what NPC vendors pay out and what they charge.

This is the part players argue about rather than measure. The usual reference point is the Anniversary realms, where one camp says consumable prices climbed well ahead of ore, leather and cloth once the raiding guilds had cash to spend, and the other camp says that was a supply story about who was farming herbs rather than a story about gold at the top. Both sides are reading the same auction houses and reaching different conclusions, and we have not seen a dataset that settles it. What is worth taking from the argument is the direction, not the size: if you expect prices to rise unevenly, expect the consumable end to move first and the one-time purchases to lag.

The absence of boosts is not only a storefront decision, because the levelling curve makes a boost hard to sell between players as well.

What that cut does to the people who used to charge for those hours is worked through in pricing a paid dungeon run.

Beta reports describe dungeon experience cut below what the same clear paid in Classic, which takes the paid run apart as a service

What players still charge for once the levels stop paying is collected in what paid services cost.

: a high-level character dragging a low one through a dungeon no longer returns levels fast enough for anyone to charge for the hours behind it. Blizzard has not published the rates behind that cut, so treat the size of it as a beta observation rather than a stated rule, and expect it to be adjusted before 4 November like anything else on the test client.

The counter-proposal on the forum is to tie the penalty to the group instead of to the dungeon: calculate the reduction from the level of the highest member in the party, so 5 players levelling together keep the full rate while a single carry sitting far above them removes most of it. Nothing has been said about that by Blizzard, and it is worth reading as a player request rather than a plan. What it does describe is where the money goes either way. With carries difficult to price, the gold a levelling character needs lands on gear, materials and riding training rather than on another player's time, and the demand that stays intact is the one for cooked food and the market for experience bonuses that any character can buy off a listing.

There is a third channel that belongs in neither column, and it has been argued about on the forum for 2 days now: gold donated to streamers during tournaments. A viewer sends coins to a competitor or to the organiser, and the competitor spends them on gear bought off the auction house, on enchants applied before the bracket starts, and on the stack of water, food and totem reagents that a long run of matches burns through. None of it is new gold. Every coin was farmed by somebody who then handed it to a player they will never group with.

That makes donations a redistribution channel rather than a faucet or a sink, and it works against the usual direction of travel. Ordinary trade moves gold from the people who want items to the people who farmed them; a tournament moves it the other way, from spectators to a small set of characters who are already well equipped. The effect on prices is indirect and we have no way to size it — nobody publishes what is donated, and the competitors themselves rarely say. What can be said is that for the duration of an event, a handful of buyers are spending coins they did not earn in the world, and the listings they clear out are the same ones everyone else was shopping from.

Worth stating plainly that the absence of a paid-run market is not a rule Blizzard wrote down anywhere. The only stated decision is the storefront one: no boost sold by Blizzard. Everything else follows from a single tuning number, the experience a dungeon clear pays, and that number sits on a test client where it can be raised, lowered or left alone before 4 November. If it moves back towards what the same clear paid in Classic, the hours behind a carry start returning levels again, and the service becomes sellable between players without anything else in the ruleset changing.

That is the usual shape of a tuning number on a test client, and the same instability runs through item and recipe nerfs between builds.

The forum argument is already being had in exactly those terms. Posters describing the current rates say the point of them is that nobody will be advertising boosts in trade chat, and the replies disagreeing with the cut answer in the same currency: that a slower dungeon is not a reason to level alone, and that the people who want to pay for the run will find someone to pay whatever the rate is. Neither side has a published figure to argue from, which is why the question of what a dungeon carry sells for is open rather than settled, and why it is better read as the one lever that decides whether the market exists at all.

The Token itself keeps coming back as a request. The case made for it on the forums is an anti-bot one: if players who want gold can get it from Blizzard, the argument goes, the farms selling gold lose most of their customers, and an economy without that demand is cheaper to police than one where every buyer has to find a seller. Posters making that case usually frame it as the lesser of 2 problems rather than as something they want for its own sake.

The reply that follows in the same threads is that the Token has never actually removed real-money trading where it exists, only given it a reference price to undercut, and that the supply side keeps farming regardless because there is always someone who will not pay the storefront rate. Whatever the merits either way, the decision as it stands is the one said out loud at BlizzCon: no plans for anything like a Token in Forever. Nothing has been announced since that changes it, so the requests are worth reading as player opinion rather than as a sign the policy is moving before 4 November. How gold already changes hands between players is a separate matter, and the trade window warning on the test client is the only place the client itself comments on it.

There is one more faucet that nobody plans for and every fresh launch has had: the one that opens by mistake. A world built from zero runs code that has never carried this many players at once, and the first weeks are when the gaps show. The shapes it usually takes are known, even if none of them can be predicted for this launch in particular:

• vendor pricing errors, where something is bought at one counter and sold at another for more than it cost;
• resource nodes that do not deplete, so a single ore or herb spawn keeps paying out;
• quest turn-ins that can be repeated instead of closing after the first hand-in;
• a profession perk returning rare materials more often than intended, which testers have reported as a confirmed bug on the test client rather than a design choice.

Each of those adds gold or goods that no sink was built to absorb, and each is normally closed by a hotfix once it is noticed. What matters for reading prices is the order of events: the drop comes first, the fix comes after, and the material that flooded the auction house is still in bags afterwards. So a sharp fall in the first weeks is not automatically a sign that demand is weak — it can be a supply accident working its way out, and it says nothing about what vendor sell prices will hold the floor at once the gap is shut. We will not guess which of these appears, or when; it is listed here so that an unexplained collapse has an explanation other than the obvious one.

That is a statement about gold, not about money in general. The beta client shows a shop tab on the login screen, and players report the interface for a Trading Post with Trader's Tenders, the activity currency earned by completing monthly tasks rather than bought with gold. What either of them actually sells has not been described by Blizzard, so anything you read about the contents is a screenshot reading or a guess. Both sit next to the economy described here without touching it: Tenders are not gold, and nothing paid for with real money adds a coin to the world.

Tenders are not the only balance kept apart from the coins, and currencies the market never prices decide part of what a character ends up wearing.

The request players attach to that shop tab is a cosmetic one, and it is worth separating from the gold question now rather than during launch week. The recurring ask is for a Forever shop that sells mounts, pets and appearance sets, and for transmog already bought on an existing account to carry across so a wardrobe paid for once is not paid for twice. Neither has been described by Blizzard: no catalogue has been published, and nothing has been said about collections transferring into Forever. Treat both as a player expectation rather than a plan.

If it arrives in that form, it changes nothing in the model above. A cosmetic bought with real money is not a faucet, because no coin is created and the seller is Blizzard rather than another player. It is not a sink either, because no gold is destroyed. A transferred appearance is the same transaction with the payment already made years ago. The only paid channel that touches the price of gold is the one that moves existing coins between 2 players, and a mount skin does not do that whatever it costs. The drains that do pull gold out of the world are the ordinary ones — riding training, repairs, travel, reagents — and they sit with the unpublished rates on the confirmed gold sinks.

The one claim worth separating out is the recurring player assertion that a Token will arrive later through the same shop. It is an assertion, not an announcement — the only thing said on record is the BlizzCon line about there being no plans for a Token or a boost. If that changes, the faucet model above changes with it, and so does every price on this site. Until then, treat the shop and the Trading Post as a parallel contour: they can take your money, but they cannot print gold, and they do not appear anywhere in the published gold sinks.

We took apart what a Token would change in a world built this way, including the bills it would have been priced against.

The request itself keeps coming back for a reason worth naming, because it is a community position rather than a design one. The recurring forum argument is not that a Token would be fun to have: it is that players believe gold buying goes unanswered in practice, that the people doing it keep what they bought, and that an official exchange would at least put the same gold behind a rule anyone can read. Blizzard has said it intends to enforce harder in Forever, but the thread starts from the assumption that enforcement is thin, and that assumption is what the request is built on.

What follows in those threads is a refusal, usually inside the first few replies, and it does not argue the economics at all. The standard answer is that a Token would be the day the poster cancels — “instant unsub”, in the wording that repeats — because the absence of a shop-bought balance is the reason they came to a fresh world in the first place. Nobody has published counts of who holds which view, and a forum thread is not a survey. The practical point is that the no-Token line is held up from 2 directions at once: a stated design decision, and a vocal part of the player base that treats the request as a dealbreaker. Anyone modelling a reversal should weigh both, alongside the demand that the removal of GDKP already took out of the market.

The case the other side makes runs on 3 arguments, and they are worth stating plainly rather than summarising as “people want a Token”. The first is substitution: demand for gold exists whether or not an official channel does, so routing it through Blizzard would pull that demand away from outside sellers instead of leaving it there. The second is the faucet argument — a Token moves coins that a player already farmed, while the farms reported on the beta create coins and never spend them back out, so supporters read an official exchange as the cheaper of 2 evils for the currency itself. The third is time: a player with a job and a raid night wants the 1,000 gold riding bill paid without the farming hours behind it, and says buying an hour of someone else's play is what the exchange is for. None of the 3 has been answered by Blizzard in economic terms; the only statement on record is the design one about there being no plans.

What the absence actually does to a price is narrower than either camp claims. With no official exchange there is no reference rate — in retail the Token sets a ceiling that every other quote has to argue with, and Forever has nothing playing that role. The only things setting the number are how many hours went into farming the coins, how many sellers are covering a given playstyle that day, and how thin the market behind them is. Thin markets move on small volumes in both directions, which is why a quote from last month is not a quote, and why the daily per-playstyle numbers are republished rather than fixed.

How those listings are ordered, held and paid for is set out in the order listings sell in.

What testers describe when that market is unreachable for a run of days is collected in days with no auction house.

2. Where gold comes from, and where it dies

2 flows decide whether a currency inflates, and they are not the same thing. Faucets create gold that did not exist: quest rewards, mob drops, vendoring loot to an NPC. Sinks destroy it: repairs, trainers, riding, reagents bought from vendors, and the fee for changing your talents. Trading between players — the auction house, your guild's consumable run, a seller's mail — moves gold sideways. It never creates or destroys any.

The same sideways move happens inside a party, which is why splitting coin inside a party changes who holds the money rather than how much of it exists.

The game keeps that same split in front of you. The statistics panel counts money earned in separate lines — loot picked up off mobs, quest rewards, sales to vendors, and auction house proceeds — and money spent in lines of its own, including travel, postage, respec and the barbershop. None of that is new information about the design, but it is the faucet and sink list in the game's own wording, and it is the cheapest way to check anything written above against your own character instead of taking our word for it.

The exact labels are worth quoting, because they are the client agreeing with the split rather than us asserting it. The income counters read Gold looted, Gold from quest rewards, Gold from vendors and Gold earned from auctions. The first 3 are faucets: that money did not exist before a mob died, a quest turned in or an NPC bought your greens. The 4th is not a faucet at all — auction proceeds came out of another player's bag, so the world total is unchanged even when your own line climbs.

Of those 3 faucets, the one players on the forum keep pointing at as the main tap is the third rather than the first. Coins dropped by mobs are the visible source, but they arrive in small amounts per kill, and a character clearing the same camp for an hour ends that hour with far more value in greens, whites, cloth and trash items than in loose silver. That value only becomes gold at the moment an NPC buys it, which means most of the new money entering the world is created at a vendor counter rather than on a corpse. Nothing has been published about drop rates or vendor prices in Forever, so treat the size of the gap as a reading of how the counters move rather than a stated rule — but the direction is easy to check yourself against the vendor income line after a farming session.

Framing the faucet that way changes which habits matter. Bag space, repair trips and the walk back to town stop being housekeeping and become the throttle on how fast the world prints gold, because loot left on the ground or destroyed for space never turns into a coin at all. It also explains why disenchanting and milling do not add to the total: they take an item that would have been sold to an NPC and turn it into a material sold to another player, which moves money sideways instead of creating it. The same logic runs the other way for a crafter who makes goods specifically for the vendor price, which is the narrower case set out below.

The third of those lines is the one a crafter can deliberately grow, and the reason the developers keep trimming it is set out in income from an NPC buyer.

The same counter runs the other way when a player buys from an NPC and lists it, which is the subject of limited vendor stock and resale.

Line by line, with a method for turning the same counters into a weekly income figure, we went through the statistics panel's gold counters on their own.

The spending side lists Gold spent on talent tree respecs, at barber shops, on postage and on travel, and every one of those is a genuine sink: the coin goes to an NPC and leaves the economy. Read as a pair, the 2 columns are the same model this section describes, written by the developers into a panel that ships with the game. What the panel does not do is tell you the rates — it names the respec line without naming the fee, and it counts postage without saying what a letter costs. For the ones that have been published, we keep the confirmed drains on gold together with the blanks.

2 things are worth keeping in mind while reading those counters. Trading with another player shows up as income on one side and spending on the other, so nothing in the panel tells you whether the world holds more gold than it did yesterday — only where your own share went. And the spending lines name the sinks without naming the prices: the respec line exists whatever the fee turns out to be, and that fee has not been published. Travel is the line most people underestimate, because it now covers a renewing speed consumable rather than one purchase you make and forget.

Travel is also the place where players sell to each other rather than to an NPC, and the mage is the clearest case of it. The scroll described for Forever sends the mage, and the group with them, to the inn where a party member is standing, which turns a summon into something an ordinary group can arrange for itself. Every trip handled that way is a flight not paid for and a portal not tipped for, and the 2 are different flows: the flight is a sink that does not fire, so the coin stays in the world, while the tip is a transfer between 2 players that simply never happens. Neither of them creates gold, and a tip handed over in trade is not travel spending, so it does not appear in the travel spending counter at all.

How much that moves is not something anyone can put in a figure yet. Nothing has been published about what the scroll costs, what reagent it needs, how often it can be used, or whether the destination inn has to be one the mage already knows. The part worth watching is the service market around it: a mage charging for a port competes with a free option a group can run itself, and the same pressure lands on paid summon chains and on the portal tips that were a steady side income in Classic. If that income thins out, it thins out for the players who were earning it, not for the world total — the gold was changing hands, not being made.

The respec fee is the one most people forget, and it is live on the forum right now: players are asking for it to be free on the beta, or cut by 90%. The answer they keep getting is that beta characters are wiped on November 4, so nothing paid there is real. Forever has not published what the fee will be at launch. We keep every confirmed sink in one place, blanks included.

That distinction is why “buying gold causes inflation” is wrong as stated. Buying moves existing gold from a farmer to you. What creates gold is the farming itself, and what would truly wreck an economy is farming at industrial scale with nobody spending it back out — the bot problem, not the buyer problem. More on that below.

There is a third flow the 2 columns do not describe, and it is worth naming here rather than leaving it to the closing section: a purchase made with real money. It is not a faucet, because no coin is created — the gold already existed in a farmer's bag. It is not a sink either, because nothing is destroyed. In the client it shows up as ordinary trade or a letter in the mail, which means the statistics panel records it exactly the same way it records a guildmate paying you for ore.

The channels themselves are worth knowing in order, which is why we set out how gold changes hands between players apart from the question of who paid for it.

The faucet and sink model above cannot see it at all, and neither can any counter in the game. What it does add is demand that did not come from playing, arriving on a market whose supply is fixed by how many hours people spent farming.

That is the core of the argument players have been having on the forums since the roadmap went up. One side says the consumable market is where it lands first: flasks, potions and cooked food are bought every raid night by a fixed number of raiders, supply is limited by gatherers and cooldowns, and anyone bidding with money instead of farmed gold pushes the clearing price up for the guild sitting next to them. The other side answers that it is their own money and their own time, that the coin came from a player who chose to sell it, and that nothing was taken from anyone. Both are right about different things. No coins are printed, so the world total is untouched — but a thin market moves on concentrated bidding whatever the source of the bid, and the size of that flow has never been published by anyone, ourselves included. If you want the part of it that is measurable, it is the supply end rather than the demand end: the farms supplying that gold are the only side of the transaction that leaves a visible trace in the world.

The biggest sink now has a number. Forever uses the Burning Crusade riding model — the mount is folded into the riding skill, you learn to ride and you are handed one — and Blizzard has put the price at 100 or 1,000 gold depending on the tier, with the first mount free. Which levels those tiers unlock at was not said.

The design note underneath it matters as much as the number: the developers called an expensive mount an aspirational goal and a gold sink, which is a statement about what they want the currency to be. 1,000 gold is the target most characters will aim at, and it is the figure this entire market prices itself around.

It is no longer the only one. The game files carry a vendor in Riverglades selling a 150 gold speed consumable that runs out after 360 hours, so for anyone who wants top speed the bill repeats every 15 days instead of ending. A sink you pay once shapes how long players save; a sink that renews shapes what they do afterwards. We went through that vendor's counter item by item, including the claim about him that does not hold up.

The forum side of this is a list of sinks players say are missing, and it is worth reading because it names where they think the pressure will land. The requests that keep coming back: repair bills heavy enough that dying costs money rather than a walk back; a ceiling on what a single auction lot can be listed for, so a thin market cannot be parked at an arbitrary price;

What a death takes out of an evening today, counted as repair plus the hour around it, is set out in the price of dying in Forever.

more of the income moved out of mob drops and vendor sales and into quest rewards, where the developers set the rate directly; and more services sold for gold rather than for nothing at all, the bag and bank and travel conveniences that give a high earner somewhere to spend. None of that has been announced, and some of it pulls against itself — a listing cap and an expensive death are asking for opposite things from the same market.

Treat the list as player pressure, not as a forecast. It tells you which prices the loudest part of the audience expects to move, and it tells you that the sinks described above are read by that audience as too light for a world with no second source of gold. What it does not do is add a single confirmed rate: repair costs, auction rules and quest payouts are all unpublished, which is why every sink Blizzard has published sits on the list of confirmed drains next to the blanks instead of being folded into a projection.

The repair argument is the one that keeps coming back in those threads, and it is worth stating the way the posters state it. Their reasoning is not that dying should hurt: it is that a world with no mandatory outflow lets gold pile up in bags, and gold that is never forced out of a bag goes to the auction house instead. A player who ends every evening holding more coin than they started bids a little higher for the same flask the next night, while the supply of materials behind that flask has not moved at all. Under that reading a heavy repair bill is not a punishment but a drain that fires on its own, with no decision attached to it, which is why the same posters ask for it before they ask for anything else. Blizzard has not published repair rates for Forever, so the size of the effect is an argument rather than a measurement, and what a death takes out of an evening today is set out in the price of dying in Forever.

The other half of the same discussion is what Forever has said it will actually do about real-money trading, and the list is short enough to quote. 3 things are on record from the September 17 Q&A: GDKP is not allowed, the economy is to be watched with the stated aim of reclaiming laundered gold, and the stance on buyers is described as harder than in previous years. Around them sit the design decisions that do the same work quietly — no Token and no paid boosts, so there is no official channel to compare a price against, and 6 markets that cannot trade with each other, so supply built in one of them cannot be moved into another. None of that has been given a rate, a threshold or a published method, and nothing has been said about how any of it is measured. The wording of each statement is kept apart from our reading of it in Blizzard's stated RMT policy, and the demand the first of them removed is worked through in the removal of GDKP.

Purchasing power runs the other way from price, and that is the part most launch coverage skips. On 4 November the world total is close to zero, so a modest balance is a large share of all the gold anyone holds, and the auction house prices itself against what buyers can actually pay rather than against what a material costs in farming hours. The same sum a month later sits in a deeper pool: more coins have been looted, gatherers have built stock, and listings have moved up to meet the money. That is the awkward pair a buyer holds at once — gold costs the most in real money in the first week and buys the most inside the game in that same week — and neither side of it can be put in a figure before the market exists. Which one matters more depends on what the coin is for: riding training at 1,000 gold is a fixed bill that does not care when it is paid, while materials and crafted goods are priced by whoever else is bidding that day. The order of the first spends is worth settling in advance, alongside the rest of the launch week checklist.

The forum answer to the same squeeze is a short list of player proposals, and all 3 of them run into the faucet and sink set described above. A Token sold through the shop would add gold that nobody farmed, which is the one flow the current design does not have anywhere in it. A cap on what a single auction lot can be listed for changes the price a thin market is able to print without changing how much gold sits behind the bids. And gold held on the account rather than per character moves the same coins sideways between your own characters, which removes part of the saving an expensive mount was built to demand instead of adding a coin to the world. None of the 3 has been announced, and the first has the BlizzCon line about there being no plans standing against it.

3. 6 economies, not one

The headline is that Forever has no realms: one mega-server for the Americas, one for Europe, one for Asia. The detail that matters for money is one level down — inside a region you choose a playstyle, and playstyles do not mix. Normal, PvP and Roleplay are separate ecosystems, which makes each of them a separate market with its own supply, its own demand and its own price.

2 regions times 3 launch playstyles is 6 markets on November 4, becoming 8 when Hardcore lands. This is not theory: it is visible in what sellers charge on the beta today. Roleplay gold costs 229% more than Normal gold, on the same day, from the same seller, for the same product — because the RP population is smaller and so is the supply.

It also means the mega-server matters less to a buyer than it looks. In Classic a seller on the wrong realm needed a transfer before they could hand you anything; that entire class of problem is gone. What is left is the faction line, exactly as hard as it was in 2004: mail and trade are same-faction only.

For a seller rather than a buyer, the same structure is what makes a material stack compete with every other one in the region, which we took apart in the shared auction house.

The Skyborne complicate how that line reads without moving it. The race is available on both sides, and Skyborne characters of opposite factions share a language, so 2 players who could not previously exchange a word can now hold a conversation. That is a chat channel, not a money channel. Nothing said so far attaches trade, mail or a shared auction house to it, and a Horde character still cannot hand gold to an Alliance one whatever language the 2 of them are speaking.

Which race you pick decides more than the language, and racial bonuses on a farming route are the part that shows up in gold.

Worth keeping the 2 apart when launch coverage runs the headline, because cross-faction speech is easy to read as a step towards a common market. What it changes is coordination — arranging a truce, agreeing on a world boss, organising a group — and none of that is a payment. For anyone buying, the practical rule is the one above: region, playstyle and faction all have to match on both ends of a delivery, and faction is chosen at character creation rather than at checkout. The split markets stay split, which is why prices are quoted per playstyle in our daily market read instead of as one number.

4. The calendar is the price curve

Gold prices in a fresh world do not drift, they step — and the steps are on Blizzard's published roadmap.

Our reading, and it is a prediction rather than a fact: the cheapest gold of the first 6 months is the first week of December. Farmers have had a month to build stock, the raids have not opened yet, and nobody is bidding for flasks. From December 9 the demand curve turns and consumables price everything. We re-price on November 18 and again after the raid week, because a number that old stops being a number and starts being a story.

The tiers are stacked, not parallel, and that splits demand in 2. The roadmap adds raids on top of each other: 3 open together on December 9, and the spring block puts 2 more above them. With the cap fixed at 60, progression runs through gear rather than through levels, so every weapon, crafted set piece and drop sold on the auction house has an expiry built into it.

One slot ages faster than the rest, which is why spell power weapon prices fall hardest in the week a new tier opens.

Demand for a specific item is finite by design — it ends the week the next tier hands out something better, and the price of that item falls with it whether or not anyone is still farming the materials for it.

Consumables have no such ceiling. Flasks, potions, food and reagents are burned per attempt and bought again for the next raid night, and a harder tier raises the bill instead of retiring it. Cooked food is the clearest case, since it carries stats and an experience bonus and is therefore wanted by levelling characters and raiders at the same time.

The levelling half of that demand is contested by sources nobody pays for, which we take apart in the competition for experience bonuses.

The practical reading: sinks tied to the current tier — gear, one-off blueprints, the drop everyone wants in week 1 — are short-lived markets, while repairs, travel and the professions feeding those flasks keep working across every tier on the calendar. How item levels scale from one raid to the next has not been published, so treat the shape as a reading of the roadmap rather than a stated rule.

Who pays that order each week, and how much of it is a transfer rather than a drain, is worked through in consumables as a repeating expense.

5. Professions: the whole supply side

With no Token, professions are not a side activity — they are the only machinery the economy has. Forever pushed hard here, and most of the detail landed at BlizzCon rather than in patch notes.

One trade on that side never reaches a listing at all, which is the subject of enchanting income through trade.

One of those changes has a market of its own: group skinning and leather supply grow with party size instead of competing for the same nodes.

That rule is reported gone again in the current build, which is why the leather market in Forever now prices itself on kills rather than on party size.

• Over 600 new recipes, spread across every profession and every skill range rather than bolted onto the top end.
• Cooking got a job. Crafted food gives stats and a small experience bonus, which turns a profession most people skipped into a levelling item with permanent demand.
• Camping objects — every profession learns its first at skill 20, and the better tiers come from blueprints that drop off dungeon bosses. Knowledge is gated behind content instead of behind a trainer, which is the first time in Classic that a recipe list has had genuine scarcity.
• The Legacy system feeds it. Points are earned account-wide — 65 available at launch, 16 spendable on any one character — and the profession tree includes Bountiful Harvest, 5 ranks, up to 20% more scarce materials from Mining, Herbalism and Skinning.

Read those together and the supply side gets interesting. Gatherers with a maxed Legacy perk out-produce gatherers without one by a fifth, on the same route, for the same time.

The demand that same system creates, rather than the yield it adds, is worked through in Legacy points and material demand.

What those points cost to earn, counted in materials rather than in coins, is worked through in the material cost of a tradeskill.

That gap is an economy of its own, and it is the reason the first weeks reward gathering over crafting: raw materials clear instantly while crafted goods are still competing with quest greens.

Fish arrive on that same supply side without costing a primary slot at all, which is the subject of the fishing side of supply.

The production half of that, from camp racks to a contest nobody has confirmed, sits with fishing stations and contests.

The objects that keep a gatherer out in the field have a bill and a saving of their own, which we worked through in what a campsite costs.

Crafted goods also move without ever reaching a listing, because finding a crafter for materials happens in trade chat and guild rosters rather than on the auction house.

Supply is only half of it: the beta's class mix decides which of those materials anyone is bidding on in the first place.

What 2 more trades would do to that same machinery is set out in glyphs and gems in Forever.

The same distinction decides what a long chain of content costs: the stages datamined for preparing for the legendary questline name locations rather than materials, so the bill lands on travel, repairs and consumables instead of a vendor price.

Each of those 4 changes is worth a closer look, because together they are what decides whether a coin is cheap to make. Professions and gold prices →

6. What everyone else is predicting

We are not the only people modelling this, and the disagreements are more useful than the consensus. Collected from the guides published since BlizzCon, with the reasoning rather than just the conclusion:

• One auction house per ruleset, and thinner margins for it. The argument runs that a single shared market per playstyle replaces dozens of realm markets, so supply pools deeper and the flipping spreads that fed Classic goldmakers largely close. Blizzard has not described the auction house at all, so this is inference — but it is the most consequential inference on the list.
• Gathering first, crafting after the blueprints. Near-universal advice, and it matches the recipe-scarcity design: you cannot craft what you have not found the blueprint for, but you can always sell ore.
• Keep a riding reserve. Because the training price is unknown, several writers suggest holding gold back rather than spending into the auction house early. Sensible under uncertainty — and it also means a lot of players will be sitting on idle gold at 40, which softens the early market.
• Guilds become supply chains. With no Token, the alchemist and the cook in your raid are infrastructure. Wipe costs land on players rather than on an exchange rate.
• The dissent worth hearing: a loud part of the player base argues the fresh economy will be eaten by bots inside a month, the way Classic Era and the Anniversary realms were, and that nothing announced so far addresses it.

7. Blizzard changed the rules

On the September 17 live Q&A, with the beta hours old, Blizzard said 3 things that move this whole article.

• GDKP will not be allowed in Forever. The gold-bidding raid format is banned outright. The stated reasoning: it is bound up with real-money trading in players' minds, it grows until it displaces the guild as the organising unit of the game, and it prices out anyone without a balance.
• They will be watching the economy. Not as a slogan — the stated intent is to track whether ordinary players can still afford to play, and to claw back laundered gold when they find it.
• And they say they have been too lenient with buyers. The plan described is an aggressive crackdown on people buying gold in Forever, with permanent bans implied rather than the suspensions of previous years.

Read economically, this is a deliberate demand-side strike. GDKP is what turned gold into raid loot, and with it gone the biggest structural reason to hold a large balance goes too.

The same statement is also where the player accepting a delivery appears, which is covered in who the RMT rule names.

Combine that with a harder ban policy and the honest forecast is a smaller, riskier gold market than Classic had — not the boom most of the sites covering this launch are implying.

The enforcement half of that same answer is worth separating from the GDKP ruling, which is why we keep Blizzard's stated RMT policy on its own.

One format the ruling does not obviously describe is a duel tournament's gold buy-in, where the entry is paid in coins and the prize is sometimes paid in money.

What the same statement implies for an individual account, and how players are reading it against the penalties for buying gold, is a separate argument running on the forum right now.

That position is already being tested on the test client, where an outside prize made buying gold during the beta an argument before launch rather than after it.

The results players describe on that same client are gathered in enforcement reported during the beta.

The full working, including why a smaller market is not the same thing as a cheaper one: what the GDKP ban did →

We would rather tell you that than sell into it quietly. If you are weighing a purchase, weigh this paragraph the heaviest: it is the newest information on this page and it is the only part that can cost you the account.

Which parts of an account that actually covers, and how much of it is published at all, is worked through in what a gold ban reaches.

8. The bot question, asked honestly

We list gold sellers for a living, so treat this section with the suspicion it deserves — and then read it, because the distinction is real and the whole market turns on it.

A bot farm and a player farm do the same thing to the faucet: both create gold. What separates them is that a bot never spends. It does not repair, it does not buy a mount, it does not train. It extracts and converts, every hour of every day, and the gold it makes never comes back out of circulation. That is the mechanism behind every Classic economy that has collapsed, and the complaint is loud on Blizzard's own forums precisely because players have watched it happen twice.

As of September 19 this is no longer only a Classic-era worry: the first bot fleets have been reported on the Forever beta.

The honest position: buying gold does not create a coin, but it does fund whoever created it. Which makes the only question that matters not should you buy, but who are you buying from — and that is a question almost nobody in this industry lets you ask clearly, because the answer is usually a chain of resellers who do not know either.

9. What this means if you are buying

• Price by playstyle, not by region. Normal and PvP sit within $0.01 of each other today, and they swap places as supply shifts; Roleplay is 214% above Normal (PvE). Your mega-server does not change the number, and your faction never will — it only changes who can physically hand it to you.
• Early gold is expensive gold. Launch week is the worst exchange rate of the year, by the same arithmetic that makes launch week the most fun. If the purchase can wait until farmers spin up, it should.
• Distrust any fixed number about Forever, including ours. A market 3 weeks old is not a market. Ours is republished daily and re-priced on November 18 for exactly that reason.

And the part we are obliged to repeat: buying gold breaks Blizzard's End User License Agreement, accounts caught doing it can be suspended or closed with the gold removed, and Blizzard has now said it intends to enforce that harder in Forever. Nobody who tells you their method is ban-proof is telling you the truth.