Zestimate Prize

Hat tip to SM for sending in this detailed article about Zillow’s contest to improve the zestimate. No surprise that the winner was a group of analyzers who blended the algorithms to move the needle a couple of ticks.  If they would have asked me, I’d say score each agent, and factor in +/- 5% based on who is selling the house – your agent makes that much of a difference!

In Seattle, the typical Zestimate is off by 4.7 percent, which amounts to $35,000 on the median home. Data scientists from around the world competed to improve the algorithm and expect to get the median error rate down to about 4 percent.

On Wednesday, the Seattle-based company awarded a $1 million prize to the winners of a public contest to improve its algorithm. The winning team, three guys from Raleigh, Toronto and Morocco who teamed up despite never having met in person, came up with a way to beat Zillow’s own data scientists to a better estimate.

The contest started a year and a half ago with 3,800 teams from 91 countries and was narrowed down to 100 finalists last year. The teams were given seven years’ worth of data on a sample of millions of homes across the country, and were tested to see how closely their estimated values for each home matched up with the actual sale prices of homes that sold in the ensuing months.

Jordan Meyer, the American on the winning team, reduced his workload at his day job as CTO of an analytics company and poured about six hours a day into the contest, communicating with his teammates, Moroccan computer science professor Chahhou Mohamed and Canadian artificial intelligence startup founder Nima Shahbazi, on the messaging application Slack.

Meyer started by finding every data source he could — the exact longitude and latitude of houses could be used to determine the proximity to streets and therefore determine noise near the house. Slight differences in distance from a body of water could influence a home price by thousands of dollars. In the end each home had hundreds of different data points.

But the strategy that set them apart was trying wildly different algorithms and merging the ones that worked together to get the best blended average.

“It was extremely hard,” Meyer said in an interview. He called the process “relentless experimentation” and echoed Shahbazi, who said in a statement: “For every idea that worked, there were a hundred that didn’t work. But we kept going.”

Zillow has slowly improved its Zestimate from a median error rate of 14 percent when it started in 2006 to 5.7 percent when the contest began in mid-2017. It’s now down to 4.5 percent nationally (it’s higher in some cities and lower in others), and once the winners’ tweaks to the algorithm are incorporated, the company expects the error rate to dip to about 4 percent.

Link to Article

Slower Start for 2019

Click to enlarge

Just what we expected – more homes for sale, and more priced wrong:

Only six months following the most competitive home buying season of all time, January data shows the U.S. housing market is off to a slower start in 2019. Although home prices are increasing, 15 percent of U.S. listings had price cuts in January, and declines in days on market have significantly decelerated since last year.

The U.S. housing market is off to a slower start this year in many markets, compared to the rapid acceleration we saw last January. Although the market is slowing, it’s important to remember that we’re coming off of four straight years of inventory declines that pushed the market to a record low availability of homes for sale. The real metric to keep an eye on is entry-level homes, which are the key to getting today’s market back in balance. These homes are still in short-supply.



Keep an eye on the entry-level homes? Let’s look at all of them!

Active listings of detached-homes between La Jolla and Carlsbad:

Date, Year
Above $2.0M
Jan 29, 2018
Jan 28, 2019
% change

Our percentage of ‘entry-level’ inventory has grown the most, year-over-year, though with the cheapest home listed at $665,000, they aren’t exactly affordable to the masses.

Though we will have more homes sitting around not selling, buyers will forget them quickly and easily – as long as at least a few homes are selling that will give direction to all involved.

Mortgage Rates Should Stabilize

Rates with no points

It sounds like mortgage rates will be rangebound for the foreseeable future, which is good for buyers – but not necessarily for sellers.  Without the threat of rates going higher, buyers will be even more patient:

From MND:

It’s true that markets were already expecting a dovish Fed announcement.  This created an asymmetric risk that the Fed would only be as friendly as they needed to be and that rates would have been positioned too low for such a thing.  As it happened, however, the Fed was noticeably friendlier than most anyone guessed.

They dropped their verbiage pertaining to additional rate hikes.  This effectively begins an era where rates will remain at current levels until economic data or other considerations motivate a change.  Big news indeed!

They also dropped the reference to economic risks being balanced.  The only way to be any friendlier to bonds would have been to say that economic risks had tilted to the downside (which can already be inferred from this change).

They also said they were prepared to adjust the balance sheet normalization policy if needed (i.e. they could start buying MBS again).  Unsurprisingly, MBS liked this news and improved at a faster clip than 10yr Treasuries.  Part of the 10yr’s problem was the big advantage the Fed’s announcement provided for shorter maturity Treasuries.  In other words, traders were buying Treasuries, but most of the love went to the 2-5yr sector.

In the bigger picture, this is just another decently-sized green day for 10yr Treasuries, but reading between the lines, there’s a bit more to like.  For MBS, it was easily the best day since January 4th.  Not only that, but there’s a chance we look back at this as the day the Fed confirmed the end of the rising rate environment of the past 2 years.  Granted, the past 2.5 months already did quite a bit in that regard, but it’s going to take a bit more time and stability to get comfortable with that idea.

Link to Article

Nice improvement today, and lowest in last 12 months – Jan 31st:

More Tidbits

More tidbits, mostly from Leonard’s daily Compass email:

Here are some interesting Los Angeles-area market stats:

  • There were 580 closed sales of $5 million+ in 2018, versus 602 in 2017.
  • There were 157 closed sales of $10 million+ in 2018, versus 185 in 2017, (down 15%).
  • 43 of these were $20 million+ in 2018, versus 52 in 2017, (down 17%).
  • 19 of the sales were $30 million+ in 2018, versus 23 in 2017, (down 17%).  Both years had 9 sales of $40 million+.
  • Of the 43 sales of $20 million+ this year, 34 of the buyers were American, (79%).  The other 9 buyers were from Monaco, China, Australia, Saudi Arabia, Switzerland, Japan,
  • The majority of the 43 sales were in Beverly Hills with 12, followed by Malibu with 9,  There were 5 in Bel Air, 5 in the Sunset Strip, 3 each in BHPO, Palisades, Brentwood and Holmby Hills.
  • 21 of the 43 sales of $20 million+ sales were not officially listed when sold.



Coronado Bridge Construction

For those who get a little queasy while driving over the Coronado Bridge, looking at these photos from its construction might help:


One of the comments – Beautiful pictures –thank you. For those of us who lived “State-side” and had to commute to jobs on North Island via monotonous ferry or long Silver Strand drive, we were thrilled to watch this beautiful bridge take shape. A little note for those wondering why the long, sweeping arch: the bridge had to be high enough for Navy ships to pass under it. A straight line from San Diego to Coronado would not allow for enough height; the roadway would simply be too steep at both ends. The long, sweeping arch shape allowed for enough roadway distance to get to the required height with a reasonable grade. Nice. It added greatly to the aesthetics. While many residents of Coronado were not happy, most San Diegans were thrilled and still are to have this beautiful bridge.

San Diego Case-Shiller Index, Nov

The non-seasonally-adjusted San Diego Case-Shiller Index dropped again in November, which seems like 100 years ago now that we’re rolling into February. A couple of notes:

  1. If you like your data processed and refined, the seasonally-adjusted index has been flat the last four months. I like mine raw and uncooked.
  2. They revise these numbers slightly every month.

We are roughly back to where we were in March.  You could make the case that the better-quality homes sell for a premium during the selling season while the others wait, and then some of those go for a slight discount in the off-season.  It might be the pattern for every year, going forward.

San Diego Non-Seasonally-Adjusted CSI changes:

Reporting Month
M-o-M chg
Y-o-Y chg
January ’17
January ’18

The previous peak was 250.34 in November, 2005.

Inventory Watch

We have been having a surge in the market – the total pendings count went up a net 11% this week!  The new-listings count slowed down too – see below (it was almost tied last week).

The inventory change was almost all on the low-end. There are 14 fewer active listings today, and 15 more pendings.  In the area between La Jolla and Carlsbad, there are roughly 300,000 people, and today we only have 68 houses for sale listed under $1,000,000 – and NO houses under $665,000.


Have you joined the contest yet? Click here to register your guess of how many total listings we will have between Jan 1 and Feb 28 – the closest will receive four tickets to a Padres game!

NSDCC New Listings Jan 1 to Jan 22:

2016: 365

2017: 284

2018: 317

2019: 282

This will be the last week to enter the contest, or to revise your guess!

The 2019 total-listings count fell behind lately, which means potential sellers are in no rush to get their house on the market.  Unless it was painfully obvious to them that the market is under siege (there’s no sign of that), then the list prices are going to hold – or possibly rise in the coming months.  Three of the four inventory groups below have substantially higher list-price-per-sf than they did at this time a year ago!



Pin It on Pinterest