The Winner’s Curse and Reverse Auctions

The winner’s curse is an interesting (certainly it sounds interesting) feature usually discussed in relation to forward auctions – i.e. selling auctions. Basically it says that the winner of an auction is the one who has over-estimated the value of the item being auctioned.

This is how it works. It’s a very theoretical example but the point is not to be accurate, it is to illustrate a point:

5 companies want to buy an oilfield. They all have roughly the same technology at their disposal for estimating the amount of oil in the oilfield. They can all generate roughly the same amount of income from the oil. Theoretically some companies will under-estimate the value of the oilfield, and some will over-estimate.

Let’s say the oilfield is worth 100 units. Let’s say, that the oil companies have estimated the oilfield’s worth at 80, 90, 100, 110 and 120 units respectively.

Let’s now say that these 5 companies have been invited to submit sealed bids for the oilfield. And let’s say that they all bid 10 units less than the price they think it’s worth to them (so that they can make some profit), then you would get bids of 70, 80, 90, 100 and 110.

I know that this is a very simplified example – and I’m not going to go into digressions about net present value and the rates of return et al.

The point is that, when you are selling something that is worth roughly the same to everyone who is buying it, and you sell via a sealed bid, then the person who wins it is the one who over-estimated its value. In this example the company who estimated the value at 120 bids 110 and finds that the oilfield is worth 100.

That is the winner’s curse. Similar logic goes for normal “english” auctions as well.

But what about reverse auctions?

In a reverse auction, bidders (suppliers) are bidding down their margin. You would hope that bidders understand their cost structures, so would be unlikely to bid below their cost due to having under-estimated their cost. Of course, there are other reasons why bidders might bid below cost if they have particular strategic reasons for trying to win a particular contract. They would do this with their eyes wide open, though, so you can’t really talk of a curse.

If this is true (and I haven’t seen any serious studies on the subject) then it would mean that reverse auctions are even more efficient at allocating resources optimally than forward auctions.


Was my father in law’s 65th birthday celebration over the weekend. I was hoping to go for a whole weekend without getting dragged into anything computer-related, and nearly did, except for the fact that he had a new digital camera for his birthday and needed to install the software on his PC for it.

The main thing that threw him was that the CD you have to install is called the Solutions disk. He had filed the CD away because he assumed it was for trouble shooting: He had assumed that the Solutions disk was for Solving problems.

Whereas any techie will tell you that solution is just a piece of meaningless jargon which can mean almost anything, but only very rarely does it actually solve anything. When a techie means solution, in the sense that everyone else understands it, they will more likely use the word resolution.

Don’t blame me, I didn’t invent the culture. I am convinced that the plethora of jargon is invented for the benefit of sales people – so they have something with a distinct name that they can sell.

However, from the buyer’s or user’s point of view the jargon is usually irrelevant, even misleading. SOA is an example doing the rounds right now (and I still don’t really understand it myself, or even particularly care). Web 2.0 is another.

Even in my own industry – the electronic reverse auction industry – the term auction has become a confusing jargon term. Sometimes it can mean a piece of software capable of running auctions, sometimes it can mean an event complete with suppliers, some times it can mean the whole process of negotiating a contract – from designing the specifications through to finally awarding the contract to the chosen supplier.

I suppose some of this is inevitable, and thankfully it’s not as bad as using obviously misleading language or meaningless neologisms. But it goes to show that even phrases that people can assume are commonly understood can have wildly different interpretations.

In other words: If the jargon doesn’t make sense, don’t think you’re dumb – it’s the jargon that’s dumb, not you. And also, even if the jargon does make sense – beware – because even common words can have very different meanings depending on who you are hearing them from!

Buyers have more fun

Was fortunate enough to be able to go to a dinner hosted by CBR last night, entitled “Spanning the IT/Business Divide”. It was being paid for by an SOA company, so there was a lot of talk about whether SOA does or does not help span the IT/Business Divide (conclusion – the name doesn’t help, because an SOA does not provide a “Service” in the way business people would understand).

More interesting was that the most provocative questions and issues, which provoked a full and frank exchange of views, came from a gentleman who used to be in IT but is now a buyer of software rather than a technologist. Sample question: “why should I invest an unknown $$$ value into an SOA  architecture when I could instead, at a low, known cost hire some bodies to type information from one system to another”.

If your immediate response to that question is to get riled and believe that the questionner is stuck somewhere in the dark ages then you have fallen into the trap of perpetuating the IT/Business divide. Because I bet you, “The Business” is asking the same question. It is a serious question. IT departments need to be able to step this far back from their day to day operations to be able to challenge themselves and their futures with questions like this.

And then I was reminded of this post on Deal Architect I read a while back, in which Vinnie Marchandi intimates that the most important skill needed in IT departments is vendor management. IT directors need to address this challenge and justify why they (assuming they are technologists) are even there in the first place. Only then does the IT/Business divide stand a chance of being spanned.

p.s. Thanks to Jason Stamperer from CBR for running the event. And apologies for not having paid attention to the new layout of CBR magazine. Truth be told I have been a big fan of the magazine for a while now but now I think about it I haven’t seen a copy in a while.