« Regulation, a dialogue with Warren Buffett | Main | The European collective response »
Wikipedia on reverse auctions
Gartner's keys to success as a supplier in reverse auctions are: (a) Thorough preparation – it's essential to know your costs, your suppliers, and your market to the greatest extent possible – tiny details can make the difference between winning and losing, and between being profitable or not; (b) Reverse auctions should be largely kept to the supply of commodity products rather than proprietary ones; and (c) Having a strong, competent bidder leading your effort at the time of the auction, with clear guidelines on when to bid and when to fold is essential.
Anticipating Hank Paulson, Gartner adds:
"I know most people don't look at reverse auctions positively, but we see them as a process that makes you better,"
Here is the link. The article will soon be much longer. Here is a website devoted to summarizing the research against reverse auctions; it appears unrelated to critiques of Paulson and the Paulson plan. It seems to be fighting a personal war and so I doubt its objectivity:
The bottom line is BUYERS should not use reverse auctions because the amount of savings that can actually be achieved is greatly overstated. In addition, reverse auctions create numerous other problems for buyers. SELLERS should not participate in reverse auctions because there is nothing in it for them; especially incumbent suppliers. In almost every case, neither buyers nor sellers benefit from this purchasing tool because it is an unhealthy continuation of zero sum power-based bargaining that degrades the competitiveness of both parties. Reverse auctions are undeniably a bad purchasing practice and a wrong approach to spend management.
I hope to soon consider other research on reverse auctions.
Posted by Tyler Cowen on October 5, 2008 at 06:42 AM in Economics | Permalink
Comments
My only experience with reverse bidding is anecdotal. Last year at a conference I host, a presenter from a large pharma company showed a few slides on how his company uses reverse online auctions for R&D outsourcing. The first slide (X-axis = bid, Y-axis = time) showed the wide-ranging bids of the various contract research organizations on the X-axis. The next slide showed the bids 5 hours later; rebids were conducted every 3 minutes. All the bids had collapsed to the bottom of the X-axis. I was aghast. The presenter told us, "Of course, we're not just looking for the lowest cost. We're trying to ADD VALUE..."
I hit up some of my associates in the industry. They're on the contract manufacturing side of the industry (which sounds like a commodity, but isn't), and asked them if they've been subjected to the same bidding process. Most replied, "Yeah. And that's why we don't pursue business with the major pharma companies."
Like I said: anecdotal, but the underlying assumption was that whoever won the bidding ended up having to build in more charges to make up for the incredibly low "winning" bid. Or cut corners, do the job poorly, and watch as the sponsor has to go back to one of the higher bidders to do the same work in a shorter timeframe.
Posted by: Virtual Memories at Oct 5, 2008 7:55:02 AM
"we see them as a process that makes you better,"
So, government really is a journey of self-discovery!
How is the government going to know the tranches are identical? Is to ask such a question to answer it?
Shall there be clawback provisions if sellers are better at valuing these assets than they are letting on?
Posted by: Andrew at Oct 5, 2008 7:55:38 AM
take a look at this:
http://marketdesigner.blogspot.com/2008/10/auction-design-for-bailout-by-treasury.html
Posted by: samson at Oct 5, 2008 9:22:25 AM
The thing that has been pointed out on CNBC constantly by the talking heads - and in this they're correct - is that Treasury doesn't want to overpay too much for these assets, but at the same time getting them at firesale prices won't help solve the problem. They need to strike a balance.
This is why I think the warrants might be a preferable idea - require the banks to put more skin in the game in the event we're overpaying by too much (whatever too much is).
Posted by: meter at Oct 5, 2008 10:36:31 AM
A reverse auction is conceptually no different from a normal auction, except that the primary party wants to buy, rather than sell, the item. As such, a reverse auction does not "drive prices" anywhere in particular.
Auctions make little sense for specified, fungible products (e,g, ounces of gold.)
Like a normal auction, a reverse auction makes sense when the item is specific, yet not a commodity (e.g. a specified widget.) The key is your ability to specify (e.g. you wouldn't do an RA to buy a 1979 Trans-Am car: you can't specify enough to avoid getting junk.)
A reverse auction for complex financial derivatives is a pretty insane idea. If you under-specify (e.g. an AA rated $10MM CDO tranche,) you get junk, If you specify tightly, small sellers are unable to bid, while large sellers repackage existing derivatives into even more complex securities that conform to the specification (so small sellers sell to big sellers, and the govt winds up owning a lot of pointless derivatives.)
Posted by: gorobei at Oct 5, 2008 11:43:24 AM
I have run many reverse auctions for a large healthcare organization on the buying side, and Miliani's characterization of them as "zero sum" bargaining describes a pitfall to avoid, not an inherent quality to all auctions.
The primary fallacy is that the low bid in an auction will automatically win the business. Auctions are simply a mechanism to speed up the price negotiation process, and to provide suppliers with instant feedback on where they sit in the overall market for the goods or services being bid (price discovery). Well before you get to an auction, several rounds of RFIs or RFPs will have taken place, with all of the non-price attributes collected and scored, including quality, service, critical contract terms, and of course assurance of supply. Miliani's examples of poor outcomes all have to do with a poor sourcing process overall, where these factors were ignored in favor of focusing on price--obviously those outcomes will be poor regardless of the mechanism used to negotiate.
Also, auctions do not necessarily need to be kept to pure commodities. I've run auctions on everything from proprietary hardware to consulting services (setting base rate cards). The key is having well-defined specifications, and again, to reassure suppliers that their non-price attributes will be taken into account. That way they can participate and not necessarily bid the lowest, but still win the business.
Posted by: BartJones at Oct 5, 2008 12:16:18 PM
As a sometime participant in reverse auctions primarily as seller of services, my experiences with them are wholly negative, even if I was able to win some business. For a complicated product like a software development, generally both buyers and sellers can manipulate things so the auction and procurement is no longer about getting the job done, its about suckering in their counterpart so they get value from the counterpart at deep discount, without having to do the work.
For example, I've seen buyers put the most superficial of specs for software out there, hire someone at deep discount to design and build the job, then fire them after the design is done and documented only to return to the same auction site with documented design in hand. Protests that procurements aren't specified enough are met with shrugged from the site operator and continuing bids.
On the seller side, I've known buyers who hired someone at discount, then found that the product of the seller simply did not work. I learned this because they were back at the auction site looking for someone to fix up the mess the earlier people had left.
Without auditing and enforcement of anti-fraud (which probably impose overhead costs making the reverse auction far less attractive), I don't see it being a good way of getting anything, apart from the most standardized of goods.
Posted by: ekzept at Oct 5, 2008 3:25:29 PM
It sounds like an auction has a tendency to sanitize the one thing I'd be interested in in a business deal, an ongoing relationship.
It sounds kind of like walking up to someone at a bar and saying "What's the least I could get away with to get you in the sack?" You might say it takes the mystery and romance out of the equation.
Again, how do we make sure we don't get screwed without a call back?
Posted by: Andrew at Oct 6, 2008 6:05:12 AM
Reverse auctions are already being widely used by big corporations (such as GM or IBM) and by government entities through FedBid. As a business owner, I'd rather say that one sale with short margins is always better than no sales, so I welcome every new buyer's request that come in for the products we sell.
I agree that there should be some sort of regulations and limitations to avoid abuse but I think initiatives such as www.oltiby.com will help make the reverse auctions more popular to the general public.
Posted by: reverse auction at Oct 6, 2008 9:27:39 AM
Wasn't Google's IPO a reverse auction? Seemed to work ok for them.
Posted by: Cobb at Oct 6, 2008 11:48:07 AM
