A Case Where Auction Mechanics Make Sense

Oh my goodness. There are homes and there are HOMES. 2511 Baker Street is the latter – a roughly 3,700 square foot gem in Pacific Heights that is somehow more spectacular in person than in photos. The layout is wonderful with kitchen and entertainment space on the ground floor, bedrooms on the 2nd floor, and one of San Francisco’s most stunning home offices on the 3rd floor. It has multiple balconies and patios with that rare combination of views and privacy. If you lived here you would just stay home and order everything you needed from Amazon and Instacart. It is listed at $5,750,000 and I don’t see how that number isn’t surpassed in auction, but let’s dig into why.

One of my central theories is that most homes don’t benefit as much as you might think from auction mechanics. By auction mechanics I’m referring to the process of running some number of well-marketed opens homes, collecting offers, and then getting buyers to bid up to the highest possible sale price until the last person standing wins the deal. A counterargument to my theory is that many homes in the bay area are “going for 20% over list price” as indication that exactly this dynamic is playing out, but I think this has more to do with a specific listing strategy (list very low) and we should treat the sale to original list price ratio as a fairly easy to manipulate number.

The reason I think this is because most homes are not, strictly speaking, unique. Yes that home you have your eye on might be the only house for sale in the last 20 months on Commonwealth Avenue but if you broaden just a little bit to the San Francisco market for $3m homes, you are going to quickly find other places you could put your money which will prevent you from getting too detached from fundamentals when you bid. You can see this line of thinking in my analysis of 1178 Greenwich Street – imagine that home, minus the roof deck, somewhere else and what it would be worth and then work out how much you would have to value the deck and neighborhood to pay $1.7m, $1.9m, $2.1m etc for it.

2511 Baker Street is unique. My guess is that there are ~15 homes that go on the market each year that you could say are comparable. They are all in Presidio Heights, Pacific Heights, Sea Cliff, and maybe Russian Hill. That’s it. I remember seeing one on the market a few months ago – 100 Locust – just a couple years too early for me to get in on the bidding 🙁 🙁 🙁 .

So, if you happened to stop by 2511 Baker Street on Sunday and are in the market for a $5-10m home in these neighborhoods, what are you thinking to yourself? Do you want to wait until the next home of this caliber comes on the market or would you “overpay” a little bit to get this one? And let me add one thing – I think the market for modern/contemporary homes does not track the same way – there’s almost a glut of contemporary homes all over the city but homes like this? It feels more and more like 1/1000.

This is where running some version of an auction really makes sense. My first instinct is to model the auction premium as the range of observed premiums you would obtain if you auctioned the home a large number of times. The wider the range, the more we are basically saying that we don’t know what buyer/bidder dynamics are going to play out and consequently it is impossible to price the asset with high confidence.

My wild guess is that if we look at all the buyers who have the means to purchase at this price level and decide they like the property, several of them would be willing to go up to $2m over the fundamental value of the property.

To put some numbers to this, I’m assigning a neighborhood score of 1.3, a usable square footage score of 1.0, a view premium of $2,000,000 and an auction premium range of $500k-$2m. From our prior model:

Estimated sale price = ($ / square foot) * neighborhood score * square footage * usable square footage score + view premium + auction premium

Estimated sale price = ($1000 / square foot) * 1.3 * 3700 * 1.0 + $2,000,000 + ($500,000 <-> $2,000,000)

Estimated sale price = $6,810,000 + ($500,000 <-> $2,000,000)

Estimated sale price = $7,310,000 <-> $8,810,000

In other words – I think in another neighborhood (and with no view) this is a $3.7m home. In the same neighborhood (still with no view) it’s a $4.8m home. And with the view it jumps up to $6.8m, with an additional auction premium somewhere between $500k to $2m depending on how many buyers want to stretch (and how far) for a truly unique property.

My guess is that someone will try to take it off the market in the first round of bids with a $7.5m offer.

Let’s see what happens!