Here's Gevlon from his best selling glyphs post
But there is a better reason why you should never buy a glyph for resale (unless it's below ink price of course). To have the mentioned profit, the glyph must sell. The reason for no sale can be undercut or simply low demand. Most people already have their glyphs and the summer decreases player activity. Imagine that you have monopoly (no other campers), but only 3 glyphs are bought by users a day of a type. I post 2, 2 times a day. So you must buy 4. You paid 80G for it, and get 150G income instead of 200. Also, when I'm collecting mail and crafting, I notice that my glyphs are sold. I will assume that players are buying it and increase my craft/post quota. If I sell 2 x 3 a day, you have to spend 120G to buy me out, but you still just have 150G income, 30G profit/day while I have 16*6 = 96G. I guess you can figure out the next step: believing that users buying my stuff, I increase my crafting even more, selling 2 x 4, that would cost you 160G to buy out. Of course on a high population server the numbers are higher, but the problem is the same. As deep undercutter I'm happy to work for the price I set. So if you buy me out, I'm happy and encouraged to work more, crafting and selling more, finally reaching the point where you can't buy out the legions of glyphs I'm pouring into the AH (cluelessly believing they are sold to users).
The mistake made here is that the putative monopolist is as automatic as using a streamlined QA3 or auctioneer strategy.
But it isn't. For exactly the reasons Gevlon gives, there doesn't seem to be a simple algorithm that works for the monopolist strategy. That's the big advantage of the deep undercut strategy. You can program a deep uncercut strategy into QA3 and put little or no further thinking into your strategy, while making gold.
As a monopolist, you do not have to to camp the ah to win, but you do have to put a bit more thought into when and how you buy and list.
The example I quote here shows the *danger* of buying glyphs for more than they cost to craft, but it does not demonstrate that you should never do it, only that buying every glyph that comes in order to prop the price up indefinitely is a long run bad strategy. If you want to be a monopolist, you must understand the market better than your competitors (or their algorithms) do.
Let's assume the consistent market sellers of some Glyph consist solely of X (potential monopolist) and Y (deep undercut crafter following your automatic strategy), X can use the following strategy and win (Y doesn't really lose, btw, X just makes gold from trading alone if he judges the market enough better than Y, Y still does quite well, just not as well as s/he could have):
Y lists 2 of a glyph twice a day at 20g fallback, X believes that the market supports selling 4 glyphs per day at 50g ea, so s/he figures out Y's posting schedule, and arranges to buy out those glyphs and repost them for 50g as soon as Y is offline.
All glyphs sell for 200g ea. day, for 120g/day profit.
After a few days, Y needs to craft, notices quick sales, so decides to post 3 of that glyph. X buys out all 6, sells 4 for 200g, 80g profit for a few days.
Now Y decides to craft and sell 8/day. X buys them out, and sells 4 for a 40g profit. Y ups again and goes for 5/half-day. Now X would make zero direct profit, buy 10, sell 4 leaves 0g. But there is still *some* profit, in that X now has 6 glyphs. So x does this also to build up a nice stock.
Now Y is listing 6 glyphs twice a day and that actually *loses* gold in the short term, plus X already has a whole pile of unsold glyphs stored up. So here's where the strategy changes up:
X stops buying the glyphs now, and doesn't try to control the price. Let's see what happens.
So for a while, X goes to a standard or deep undercut strategy with a nice low threshold. If Y updates listing and crafting behaviors every 5 days, X now has a nice stock of 60 glyphs that he has effectively paid nothing for (other than the time spent with zero resell profit to accumulate 30 glyphs while Y was posting 5/day), so he no longer needs to buy Y's to sell them. So he goes ahead and undercuts Y, selling them all for somewhere between 6g99s and 19g99s, given Y's 7-20g price range. Suddenly Y is selling fewer than 4/day, possibly none, while X is getting rid of the excess stock. If X has done this right, X is already in the black, so s/he can sell those glyphs at any price: sell them all at 6g, under Y's threshold, and that's another 360g. Meanwhile Y hasn't been selling much, so Y goes back to maintenance posting only thinking "hm, weird, market was really hot, now it's crashed", and once X's glyph stock gets low and Y has gone back to posting/crafting smaller amounts, X can start the strategy all over again.
So here's the deal -- this strategy makes money, but it requires more effort. Y is rolling along automatically, making however much g/day without having to think about the market. X is doing pure AH pvp, not the stupid kind where you give up your own profit in order to make the other guy lose, but a smart kind, where you look at markets carefully and decide that you can make some gold that the other guy is leaving on the table.
If there are multiple goblins keeping an eye your market, it doesn't work as well (because as soon as you raise the price for a few days, they come in, so you can't run the full course). Too many other monopolists, especially if one or more are campers, also spoil the fun rather quickly.
Further, in markets where there really is massive price elasticity (i.e. if a pair of glyphs are reasonable to use as a consumables for raiders at low prices, but will have relatively little takeup at high prices), there may be more money in crafting 10-15/day for 8-12g than in trying to push the price to 50g.
Now, one can easily argue that the undercutter could alter strategy as well, and ultimately keep the monopolist from making any gold. And this is true. But it would require understanding the market as well as the monopolist, and making the same sort of time commitment to that market as the monopolist is prepared to. Basically, you find the optimum monopoly price and listing strategy and change your settings to react appropriately to what the monopolist does, and it becomes very difficult for the monopolist to make any money short of camping you 24/7. But it may be more work than it's worth compared to simply playing the standard low fallback game and letting the monopolist make their gold.
In general the big disadvantage to the monopolist strategy is that it takes more work if you want to do it well. And honestly isn't generally worth it in the glyph market where your best case is getting profits of 20-30g per glyph for a few sales a day and that is quite rare. Where it makes sense to use, is in middle to big ticket markets (where profits of at least 40-50g per item are possible), and markets that move a lot of stock with relatively few big providers. It's also very helpful if you have a way to control the mat markets along with the finished markets.
I did *very* well with enchant mats in burning crusade, propping up prices on shards and essences and riding the wave back down, buying up cheap supply to reset the price -- ended up getting half way to cap without really having cap as a goal.
Lately, I'm more likely to do industry than try to control markets, because with more goblins around today, the inefficiencies are rarer and smaller.
-- The Gnome of Zurich