What Actually Happens to Prices When an Expansion Launches

Every expansion gets the same article a month before it ships. Stockpile these, level that, here is your checklist. Almost none of those articles come back afterwards to check whether the advice worked, because checking requires price history nobody bothered to keep.

We kept it. Forty seven consecutive monthly price observations for every item on the region-wide commodity market, September 2022 through July 2026, covering three retail launches: November 2022, August 2024 and February 2026. This page is the curve those launches traced, with the method written out so you can run it again yourself.

How this was measured

One row per item per month, carrying that month's low, average and high market price and the average quantity listed. Monthly rather than daily, because the daily series only reaches back a few weeks and the interesting question spans years.

A launch cohort is defined mechanically, with no judgement calls. Every commodity whose first month of trading anywhere in the series is the launch month counts as a new material of that expansion. That is 941 items in November 2022, 1,096 in August 2024 and 678 in February 2026. The data dates each launch by itself: those items arrive carrying two or three days of trading in the 2022 and 2026 launch months, and about ten in August 2024, which is what early access looks like.

Then a liquidity filter, at least a thousand units listed in the first full month, which throws out the single-listing curiosities and leaves 580, 523 and 365 items. Each surviving item is indexed to its own first full month, and the reported figure is the median across the cohort, not the mean. One luxury reagent moving a hundredfold would otherwise carry the whole curve on its back. The quartiles behind these medians are wide and individual outcomes vary enormously, which is the standing caveat on every number below.

The curve

Median price, with the first full month after each launch set to 100.

Months after launch	Nov 2022 launch	Aug 2024 launch	Feb 2026 launch
1	100	100	100
2	50	67	58
3	32	47	36
4	21	34	28
5	25	33	24
6	26	36	not yet
9	21	30	not yet
12	26	39	not yet

Three launches, three and a half years apart, same shape. Half the value is gone by the second full month. Two thirds to three quarters is gone by the third. The floor arrives in month four or five at somewhere between a fifth and a third of the opening price, and then the line stops falling.

Supply explains it, and supply is measurable too. Against the launch month, median listed quantity across the new cohort reached roughly eight times in the first full month of the 2022 launch, about twice for 2024, and roughly six times for 2026. Nobody is farming faster in month two. The stockpiles built during the levelling rush simply arrive on the market at once, and they arrive before the crafting demand that eventually absorbs them.

Launch week is not a price

The launch month is not a month. It is two or three days of trading, or about ten when there is early access, and during those days the market is nothing but scarcity.

Measured that way the numbers look apocalyptic. The median new material traded at 17, 33 and 32 percent of its launch-month price by the first full month, across the three launches in order. Reading that as a crash is a mistake. Most of it is the launch-week figure being a poor description of a market, which is why the curve above is indexed to the first full month instead.

This matters if you are the one selling into launch week. That window is real, it is short, and it is the only part of the whole cycle where a stockpile built before launch is worth more than the time it took to build.

What the old materials do

Prep guides assume last expansion's materials spike before a launch. Across 2,115 items already trading six months before the 2024 launch and 2,346 before the 2026 one, the median tells a duller story: a slow firming a few months out, then a fade into launch day, then a real dip.

Indexed to six months before launch, the median old commodity sat at 89 the month before the 2024 launch, 65 in the launch month, 74 a month later, and back to 99 by month three. The 2026 launch traced the same path at 95, 84, 70 and 98. The month after a launch is when the previous expansion's materials are cheapest, not the month before.

Restrict the same query to herbs and ore, where the folklore comes from, and a genuine pre-launch move does appear. Median index for 221 gathered materials around the 2026 launch:

Months before launch	6	4	3	2	1	launch	+1	+3	+5
Median index	100	117	124	155	121	112	90	99	91

That is a real spike, and it peaks two months before launch day rather than on it. The equivalent 200-item basket around the 2024 launch peaked at 108 four months out, which is barely a spike at all. So the pattern is directionally true, wildly inconsistent in size, and over before the expansion ships. Anyone still holding gathered stock on launch day has already missed it.

The split: what comes back

Take each new material's trough, the lowest month between months three and six, and compare it with month twelve.

Outcome at month 12 against the floor	Nov 2022 cohort	Aug 2024 cohort
At least doubled	207	59
Up 50 to 100 percent	62	116
Up 10 to 50 percent	77	148
Flat	99	87
Still below the floor	135	113
Median multiple	1.35	1.25

Both cohorts split roughly the same way. Three in five end the year meaningfully above their floor, about one in six are flat, and a little over one in five are still below it. The median material is worth a quarter to a third more a year later than at its worst, which is a long way from the launch price and a long way from worthless.

The recovery is also lumpy rather than steady. In the 2024 cohort the median sat at 33 in month five, jumped to 43 in month seven, then sagged back to 30 by month nine. Bumps of that shape land on content patches rather than on the calendar, which fits the obvious mechanism: new gear to craft is what pulls a material off its floor, and a material with no crafting demand behind it has nothing to be pulled by.

Bismuth Common

7 gold

The recovery is a lease

The clearest illustration is what the 2024 launch materials did when the next launch arrived. Bismuth and Mycobloom are the flagship ore and herb of that expansion, both deep, both heavily crafted, both textbook recoveries during their own cycle.

Then February 2026 happened. Both fell hard through March, the first full month of the new expansion. By July, Bismuth had climbed back above where it had been that January and Mycobloom was sitting at a little over half. Same launch, same market, one material recovered fully and the other did not, and there is no way to tell which is which from the price alone at the moment of the crash.

Older cohorts do not get even that much. Hochenblume and Serevite Ore, the equivalents from the 2022 launch, sit today between a fifth and a tenth of their own first full month, three and a half years on. They did not recover and they did not keep falling either. They stopped mattering. A recovery is a lease on relevance, not a purchase, and it expires when the next expansion ships.

Using this without a crystal ball

Five things the curve supports, none of which require guessing what comes next.

The selling window for pre-farmed stock is launch week, not launch month. Two or three days.
Measure your holdings against the first full month, not the launch-week price. It is the first honest number the market produces.
Months four and five after a launch are where new materials stop falling. That is the buy window for anything with real crafting demand behind it, and the checks in how to read auction house price data are how you tell real demand from a thin market.
Last expansion's materials are cheapest one month after the new launch, not before it. If you need old materials for levelling a profession or filling recipes, that is the month.
Recoveries follow content patches. If a material has no recipe demand attached, nothing will pull it back up.

Where things stand today across three generations of gathered material:

Material	Launch window	Live price
Argentleaf	Feb 2026	13 gold68 silver
Mycobloom	Aug 2024	2 gold40 silver
Hochenblume	Nov 2022	53 silver

The full list, with history on every one, is on the materials index, and the per-expansion item lists sit on the Dragonflight and The War Within pages.

FAQ

Does the same curve apply to crafted goods and gear?

The cohorts above are commodities, which are priced region-wide. Finished gear is realm-scoped and behaves differently, with far thinner markets and much noisier prices. The shape of the fall is similar, the timing much less reliable.

Can I use this to time the next launch?

You can use it to recognise where you are once a launch has happened, which is the useful half. Three launches is three, not a law, and the sizes vary a great deal between them. Anything that reads as a price prediction is not coming from this data.

Which materials never recovered?

Ones with no crafting demand behind them. Null Lotus is a clean example from the 2024 cohort. It fell to a fraction of its opening price, stayed there through the whole expansion, and the first major patch did not lift it because nothing new consumed it. Check the recipe list on an item page before assuming a cheap material is a bargain.

How far back does this data go?

The monthly series starts in September 2022 and runs to July 2026, with the tracked region-wide commodity list growing from 6,345 items to 10,010 across those forty seven months. Every figure above is a query against that series, not an estimate.