Talk:French auction

Proposed deletion
Don't delete yet. There's not enough information either here or in the Walrasian auction article. I'm still not convinced they're the same thing. In any case, A search on "French auction" on both the web generally and academic papers specifically turns up a good number of results, mostly relating to IPO's. Whether it's importantly distinct from Walrasian auction, it clearly is a term that's being used. References need radical improvement of course. At worst, redirect the article to Walrasian auction until a proper distinction can be made. Cretog8 (talk) 13:26, 25 June 2008 (UTC)
 * My main concern is that the article provides no substantial or useful info on the term you're getting results for as-is, and indeed, appears to be misleading if not entirely incorrect. As noted in the prod, the term 'French auction' is used (as you say, in IPOs) but the relationship of the term used there to the article here is indirect, at best. At worst, the term used in IPOs and the term described here (as a synonym of the 'fix') are entirely different things. As noted in the prod, in my view there's nothing here worth saving, and a better article can be re-created at any time. nb. Removing the prod means you've now taken on the job of improving it. :) Debate   木  13:42, 25 June 2008 (UTC)
 * Prod text, to save people looking up the history: "There is some evidence that the term "French auction" is occasionally used in some circumstances, however nothing in this article refers to these circumstances. The only substantive content in the article that isn't already in Walrasian auction is the unsourced claim that the London bullion market uses this system, however a search of the London bullion market website and a Google search for "London bullion market" and "French auction" returns no relevant hits. It seems likely that this is a neologism, however there really isn't sufficient meaningful content to say one way or another. Given the lack of verifiability and substantive content here there is no justification in keeping it, either as-is or as a merge/redirect." Debate   木  13:58, 25 June 2008 (UTC)
 * That'll teach me to remove a Prod. Anyway, it's very much a stub, but it's now a correct stub with two good refs. What I haven't found yet is a specific source that makes it clear "French auction" is used as shorthand for "an IPO auction mechanism like the OPM they use in France". It's clear holistically from a lot of sources, but there's no single gem. O well.Cretog8 (talk) 05:38, 26 June 2008 (UTC)
 * LOL. Good job. ;-) Debate   木  06:36, 26 June 2008 (UTC)
 * BTW, based on your research, do you still feel that a merger remains an option here? Debate   木  06:38, 26 June 2008 (UTC)
 * I'm staying silent on that until I learn more about different uses of "Walrasian auction". Just not sure. Cretog8 (talk) 19:14, 27 June 2008 (UTC)
 * Here is a good reference to a well-cited article for how Walrasian auctions are utilized at least in my field: The WALRAS Algorithm: A Convergent Distributed Implementation of General Equilibrium Outcomes by John Cheng and Mike Wellman in Computational Economics 12 (1): 1-24 (1998). The paper presents one such auction/algorithm based on a tatonnement process.  Perhaps if someone who is familiar with French auctions could take a look at this, it might help provide some clarification. Halcyonhazard (talk) 04:14, 30 June 2008 (UTC)
 * OK, after a really quick skim of that, I'm back to my original impression of Walrasian auction. It's a thought experiment, or a metaphor for what could happen in the economy. The stuff the paper goes into for its actual implementation is all getting away from the Walrasian auction (for good reason), and it appears, trying to show that much looser conditions can achieve similar ends. So, no, this shouldn't be merged with Walrasian auction, its something totally different. Cretog8 (talk) 04:30, 30 June 2008 (UTC)
 * Got rid of the merge tags. (Thanks for interesting reading Halcyonhazard; I hope I can find time to do better than skim.) Cretog8 (talk) 04:34, 30 June 2008 (UTC)

Bookbuilds vs French Auction
Although the term French auction isn't consistently used, there is a very limited amount of academic literature suggesting that bookbuilds are not auctions while French auctions are, I'm not convinced that for our purposes the distinction is so clear. The main qualification in the literature appears to be that bookbuilds are essentially managed by a third party (an "auctioneer"?) who receives bids from various institutions for an allocation at a particular price (or a range of amounts at different prices - "demand curves"?). Bids lodged at prices lower than the final price generally receive no shares while those who bid above the final offer price pay the offer price. (Sounds like a Walrasian auction to me - is the literature simply using the term auction as a synonym for the English auction?) Also, the bids happen within a set range, and although this may not strictly be 100% unadulterated Walrasian, perhaps it could be characterized as a pragmatic way to keep it manageable. The main issue in distinguishing bookbuilding from an auction appears to be that there is an element of subjectivity remaining with the bookbuilder at the end of the day (eg capacity to pay). Anyhow, this is stepping well over the line into the realm of original research, but I do wonder whether we are perhaps missing important qualifications here.

There does remain an issue in this article with the relative lack of multiple reliable sources, and possibly there is an element at the moment of stretching the literature a little further than perhaps it is advisable to stretch it. I also wonder whether, in practice, this is a share-trading/finance topic that might be better thrown to editors interested in that rather than discussing it largely within the context of economic theory and game theory. I might be bold and put it to an appropriate project group at some stage... Debate  木  14:26, 28 June 2008 (UTC)


 * I'm with the literature on bookbuilding not-auctions versus these being auctions. These auctions are a bit odd however, in that the true rules are hammered out after the bids are in! And I also know little about bookbuilding, so I could be wrong there.
 * I'm comfortable with this as a topic, but I'm not positive it's an article. There is more stuff out there, including discussion in various sources of using the French auction system elsewhere. However, I know I wouldn't spend a great deal of time expanding this article.
 * My feeling is that it can remain here as a bitty article, which I'm OK with. Alternatively, it could possibly be merged with Initial public offering? I think if it should get merged anywhere, its into the use-category (IPO's) rather than the form-category (auction systems). Cretog8 (talk) 17:10, 28 June 2008 (UTC)


 * That all sounds sensible to me. Given the finance/IPO implications I can't see the value of merging this into Walrasian auction at this stage. Debate   木  00:13, 29 June 2008 (UTC)


 * From what I've seen here and from the brief bookbuilding article, it seems that bookbuilding is not an auction but rather a part of an auction because the allocations are not made, not because the auctioneer has leeway in making its decision. Regarding Debate's comment about the use of the term auction in literature: typically auction means any mechanism in which the participants submit some set/tuple/etc. of prices in a particular format and the auctioneer/mechanism determines an allocation.  This can include iterations such as in the English auction as well as SPCA and iBundle.  It may also be the case that the payoffs and strategies of the one-shot French auction are equivalent or approximately equivalent to some other auction in the way that the Vickrey auction is to the English auction, but I do not know enough about the French auction to say (and if it is, the proof may or may not be original research at this point). Halcyonhazard (talk) 04:14, 30 June 2008 (UTC)