Sunday, June 26, 2011

Lessons from a Busted Auction

In 2003 the transit authority of the state of Massachusetts (the MTA) announced that it would sell, by sealed bid auction, a ninety-acre plot of land in Allston.  The land had some significant "development issues" (being occupied at the time by CSX rail yards) but more importantly it bordered Harvard University on one end and Boston University on the other.  Both schools had deep pockets, a long term need to expand, and a serious shortage of available real estate opportunities.

An article in the Boston Globe cited an internal MTA paper's estimate that Harvard would pay approximately $150 million for the land, a figure an insider to the school's decision team told me was "disturbingly accurate."  Harvard had paid a similar price for a plot of land three years earlier that was much less encumbered but also only half the size.

Harvard did buy the land, but at half the expected price -- just $75 million.  What happened?

Before looking at what went wrong, let's revisit one of the two key differences between a negotiation and an auction: price pressure on buyers.  In a negotiation, that pressure comes from the seller, e.g. through an opening offer and statements about what prices are or aren't acceptable.  For example, if the MTA were negotiating with Harvard directly it might have opened with a request for $200 million and perhaps had a reservation price of $120 million.  Where the price ended up would depend on how the across-the-table discussions went.

In an auction the source of price pressure is from the same side of the table, i.e. rival bidders.  The MTA didn't make an opening offer or hire a team of negotiators to push Harvard towards a higher price.  Instead, it looked to Boston University to accomplish that goal on its behalf by being a rival bidder for the property.

The most obvious problem with this negotiation is that there were only two credible bidders.  Moreover, the two bidders were far from equal.  BU's endowment was less than one billion dollars; the Harvard University endowment was over twenty billion.  BU was also somewhat less constrained in terms of growth; more small plots of land were available to it along its southern border.  For these reasons it was clear that Harvard's ability and willingness to pay were both significantly higher than BU's.

The Harvard decision team created a sub-group assigned to put themselves into Boston University's position and try to determine what their rival might bid.  That team came back with a best guess of $60 million, but with the caveat that the range of possible bids was huge -- from no bid at all to over $100 million.

Some people on the Harvard team still favored a large bid.  One person told Harvard's president something along the lines of, "No one will remember who overpaid for this land but everyone will remember who secured the university's future growth for a century."

There was one more piece to the analysis.  The Harvard team recognized that their chance to get the land didn't end with the auction.  If they bid low and BU won the property they could then negotiate with BU to purchase some or all of the land.  Thus, a low bid didn't mean "save money" or "lose the property" it most likely meant "save money" or "pay a full price to BU".

Harvard submitted a bid of $75 million and BU declined to bid.  Quite likely the BU team concluded that they were unlikely to prevail and thus chose not to come in second.  I suspect that Harvard wouldn't have bid much less than $75 million even if they knew BU wasn't going to bid as any further reduction in price would increase the political risk of the sale being blocked.  (At it happens there were a number of statements of outrage at the low price and the Massachusetts rail authority threatened to seize some of the land through eminent domain.)

An effective auction can enable a seller to realize a better price than might be possible from negotiation, but a busted auction can be a disaster.  Since the power of an auction comes from same-side-of-the-table pressure, a would-be seller should needs to ask whether the available bidders will provide that pressure.  In this case, with just two bidders and a clear difference in need and resources between them, sufficient pressure didn't exist.  Boston University didn't force Harvard University to bid anything like the value it placed on the property.  A one-on-one negotiation between the MTA and Harvard would likely have resulted in a substantially higher price.

Another lesson to draw from this case is that Harvard didn't limit their thinking to the auction itself.  By thinking through their options in the event that they lost the initial auction they realized that the risk of a low bid was lower than it might seem.  Understanding your BATNA is just as important in an auction as in a negotiation.


  1. I invented an auction system designed to deal with this exact case (only two bidders, one of whom values winning much more than the other and so the latter has no incentive to bother pushing the price up). I have not been able to figure out how to interest anyone in it, even though I think there is potentially enormous value to be unlocked in certain situations like this one.

  2. P.S. You don't mention the other side, BU probably left $50 million on the table by not just buying the property themselves and then reselling it to Harvard. Or some third party could have done it.