6 ms·
This is an interesting question, what is worse for buyers, blind or open bids? Blinds bids prevent bidding wars. They use them in Scotland. Not sure if buyer o
by bennysomething 5y ago
This is an interesting question, what is worse for buyers, blind or open bids?
Blinds bids prevent bidding wars. They use them in Scotland. Not sure if buyer or seller is better off using them.
- thethimble 5y agoAlso another interesting dynamic is that bids can’t simply be compared by headline price. There’s so many other factors to consider in a bid: contingencies, time to close, reliance on a lender, likelihood of close, etc. I don’t see how a blind system can take those factors into account to algorithmically select a single winning bid.
- robocat 5y agoIt isn’t done algorithmically. Often the best bid is obvious - for example an unconditional offer at the highest bid. Sometimes the best bid is not so obvious, and the vendor chooses the bid that suits their situation the best, or they renegotiate price or conditions with the highest bidding buyer(s). A vendor that doesn’t really need to sell and that believes prices are still rising, might accept a high price with a long settling date, or other conditions that might cause the sale to fall through. A vendor than desperately needs the money might accept a lower price with immediate settlement, or if they know their house has problems they will accept a lower bid if it reduces their future liabilities (e.g. zero inspection, as-is clauses, etcetera).
- jefftk 5y ago> Not sure if buyer or seller is better off using them. There's a theorem an auction design that says that, under some vaguely reasonable assumptions about bidder information, it doesn't matter what form you use for the auction. Open auction, sealed auction, even second price auction should all give the same winner and same final price: https://en.wikipedia.org/wiki/Vickrey_auction#:~:text=Revenue%20equivalence%20of%20the%20Vickrey%20auction%20and%20sealed%20first%20price%20auction%5Bedit%5D https://en.wikipedia.org/wiki/Vickrey_auction#:~:text=Revenu...
- Beldin 5y agoThat's not entirely correct. The theorem says that in Vickrey auctions (sealed bid 2nd price), each bidder's optimal strategy is to bid their actual valuation of the auctioned object. It doesn't work with open bidding, because then you could bid artificially much as long as the 2nd highest bid was at your real valuation or below.
- bennysomething 5y agoIs there anything in game theory related to this? (Very naive question, so probably not) But yes my intuition when selling my house was "surely I'd get more if this was open bids"
- jefftk 5y agoSorry, that's not the theorem I'm referencing. See the section "Revenue equivalence of the Vickrey auction and sealed first price auction". (Or https://en.wikipedia.org/wiki/Revenue_equivalence https://en.wikipedia.org/wiki/Revenue_equivalence)
- peeters 5y agoBlind bids prevent bidding wars, but from an idealistic standpoint they don't find the true price of a home. Particularly when you have buying agents routinely advising their clients to bid 50k over what turns out to be the next highest bid. Though psychology comes into play in a bidding war, at least it's happening because the asset for sale is highly desired. It's efficient.
- chii 5y agoBlind bidding, but the winning bid only pays the 2nd highest bid. This will lead to an actual price as the highest bid. The only problem is that the seller won't receive the highest price, even if they know what it is afterwards.