Re: Fixed Prices [Re: Shoppers in your Palm and Rainy Day Checks]

Dave Long (dl@silcom.com)
Tue, 22 Jun 1999 22:28:54 -0700


Kragen opens an unrelated can of worms:
> If the auction doesn't last long, then this price will often be lower,
> ... If the auction lasts a significant amount of time, then this price will
> be higher, ...

Oops. I was much too vague. Vickrey Auctions go in a single round: all
parties turn in their bids, and then the winning parties and price are
determined, so there isn't really any time factor involved.

I do find the time factor an interesting question in normal market clearing.
We have all been exposed to a simple model: a demand curve, a supply curve,
and a clearing price which changes smoothly as the curves change. Now imagine
a less liquid market, where buyers and sellers arrive such that they average
the prior equilibrium case, but not at any given instant. What do the price
excursions look like? Does it matter if we tag the sellers and buyers with
their supply/demand values? What if we allow the sellers and buyers to carry
inventory between periods, or make investments based on the price signals?

RST, following Orwell, catches me defending the indefensible in passive voice:
> Ummm... held by whom to be superior how? I can certainly see disadvantages
> from the seller's point of view...

It's been years since I ran into this idea; I'll have to dig up the references
to properly justify seller benefits. I believe the argument is that in a
single-round first-price auction, buyers bid below their valuations, but will
bid at them in a second-price auction. Try starting with either Vickrey or
Hal Varian if you wish to beat me to it...

-Dave