Chart Accuracy

Peter Darnell

New member
Greetings
I am an LRE for Daniel Gale Sothebys International Realty covering Brooklyn and Long Island. I recently created a Median Listing Discount Char for townhouses in five neighborhoods in Brooklyn. See attached. My sponsoring broker seemed astonished at the percentage change in certain neighborhoods. Am I missing something? as per the attached chart, the MLD for a townhouse in Fort Greene has increased by 19,600.0% from last year. So for example, if a townhouse was on sale for $3,900,000 in Fort Greene last year, a consumer might presume to pay this year they might pay 3,159,000? That does type of discount does not match what any of us have seen.
Thank you for explaining this
Peter Darnell
 

Attachments

  • Town House.pdf
    60.9 KB · Views: 3

woodside

Member
I don't think you numbers are correct. These number do not reflect what is happening over there.

How did you compute the median discount?

Are you comparing the median listing price of all townhouses in each neighborhood from this month to previous month or other months? The median price approach is a very flawed approach as it is influenced by the types of the properties in the market. e.g.. this month, you may have a very large property to sell, while last month there is only one property with a smaller size. The median price ratio can give you an unrealistic high price increase (,say, more than 1000% increase). This distortion will be more extreme when only very few properties are on sale when you were only concentrating on the town houses in each neighborhood

As a sanity check, please compare the price per square feet. Price per square feet has its own flaws as it ignores the location, layout etc, but it can be a first step to check what is going on.

The gold standard to check the price appreciate is always the StreetEasy price index, where it uses the sale price of the same property to compute the price appreciation by using some statistical models.
 

Noah Rosenblatt

Talking Manhattan on UrbanDigs.com
Staff member
Ok so, the reason is volume it seems - here is a chart for just Fort Greene % list disc from orig ask. https://www.urbandigs.com/marketwide-charts/median-listing-discount/fort-greene/townhouse/

Most months show between 1-6 sales per month in thelast 1yr or so. This is for sales that WE HAVE BK LISTING DATA FOR and can calc list disc. If we do not have listing side data for asking price, etc, we cant calc % disc for the closed sale.

Here I see 57 townhouse sales in Fort Greene in last 1yr - https://www.urbandigs.com/sold-fort...d=false&proptypes=TOWNHOUSE&start_date=1 year

A quick look shows a lot wihtout listing data: the no pics are a dead giveaway - so excluding those, you only got a very small sample size here. The reason for the % chg discrepency is there is 1 sale this last month + 1 sale in the 1yr ago period so it shows % chg from 0.1 to 19.7 % - not high enough. I would do a CMA and select all the listings in the last 6 months or so that have data -- when I do that check the % list disc is about 0.5%, see below: 1641057148417.png

1641056928781.png
 

David Goldsmith

All Powerful Moderator
Staff member
Some disjointed comments:
A) Fort Green is a fairly small neighborhood, especially since Clinton Hill got recognized as a separate area. The small number of houses for sale ==> small number of transactions ==> greater affect an outlier has on numbers.

B) I think you might be conflating discount from original ask on prior sales with discount from sale price on prior sale. Over the past few years while sales have been extremely liquid, one thing we saw certainly from 2018 to 2020 in many prime Brooklyn neighborhoods was houses getting listed at crazy numbers, not selling, and going on and off the market for years until they finally came on at prices the market would bare. So discount from original list price to sale price recently has had a number of data points which look like big discounts even if they sold over the final asking price.

But I have warned for decades about the dangers of extrapolating for the general to the specific. Buyers may see a story printed somewhere that "median listing discount is currently X%" so they look at 20 properties, find the one where the Seller elected a marketing strategy of pricing under market to induce a bidding war, and then want to buy it X% off of that number "because that's what the market is these days." That doesn't work any better than a Seller who prices 20% over market and then is disappointed by all the "bottom fishers making low offers."

C) If a townhouse was listed in Fort Greene in 2020 for $3.9M and it still hasn't sold in the most liquid market in history, then yeah, maybe it is only worth $3.2M. I think one of the big stories of 2021 is that it was a year where a lot of sellers finally got the message on where the actual market was. That's why we not only saw record deal volume, but at the same time saw huge numbers of Off Market. And I think it was because many agents were so busy that they didn't have time to "work" aspirationally priced listings so they were more willing to have frank conversations with their sellers and tell them "either we price it where it's going to sell or take it off the market."

D) There's a townhouse in Fort Greene which has been on the market for 562 days and hasn't had a price drop in over a year. So we see what happens to the outliers who don't fall under (C) above.
 

Peter Darnell

New member
Thank you, David,
I am just beginning to understand the nuances of the effectiveness of charts In UrbanDigs. I think Noah's comment above sums up how "median price discounts" are not a go-to filter to uncover trends and valuable information about specific neighborhoods.
I think one has to be very careful when extrapolating from the general to the specific.
Thank you again.

Peter Darnell
Licensed Real Estate Sales Associate
Daniel Gale Sothebys International Realty
89 Seventh Avenue
Brooklyn, NY 11217
 

Noah Rosenblatt

Talking Manhattan on UrbanDigs.com
Staff member
Some disjointed comments:
A) Fort Green is a fairly small neighborhood, especially since Clinton Hill got recognized as a separate area. The small number of houses for sale ==> small number of transactions ==> greater affect an outlier has on numbers.

B) I think you might be conflating discount from original ask on prior sales with discount from sale price on prior sale. Over the past few years while sales have been extremely liquid, one thing we saw certainly from 2018 to 2020 in many prime Brooklyn neighborhoods was houses getting listed at crazy numbers, not selling, and going on and off the market for years until they finally came on at prices the market would bare. So discount from original list price to sale price recently has had a number of data points which look like big discounts even if they sold over the final asking price.

But I have warned for decades about the dangers of extrapolating for the general to the specific. Buyers may see a story printed somewhere that "median listing discount is currently X%" so they look at 20 properties, find the one where the Seller elected a marketing strategy of pricing under market to induce a bidding war, and then want to buy it X% off of that number "because that's what the market is these days." That doesn't work any better than a Seller who prices 20% over market and then is disappointed by all the "bottom fishers making low offers."

C) If a townhouse was listed in Fort Greene in 2020 for $3.9M and it still hasn't sold in the most liquid market in history, then yeah, maybe it is only worth $3.2M. I think one of the big stories of 2021 is that it was a year where a lot of sellers finally got the message on where the actual market was. That's why we not only saw record deal volume, but at the same time saw huge numbers of Off Market. And I think it was because many agents were so busy that they didn't have time to "work" aspirationally priced listings so they were more willing to have frank conversations with their sellers and tell them "either we price it where it's going to sell or take it off the market."

D) There's a townhouse in Fort Greene which has been on the market for 562 days and hasn't had a price drop in over a year. So we see what happens to the outliers who don't fall under (C) above.
the value you bring to this forum David is incredible. Thank you!
 

Noah Rosenblatt

Talking Manhattan on UrbanDigs.com
Staff member
Thank you, David,
I am just beginning to understand the nuances of the effectiveness of charts In UrbanDigs. I think Noah's comment above sums up how "median price discounts" are not a go-to filter to uncover trends and valuable information about specific neighborhoods.
I think one has to be very careful when extrapolating from the general to the specific.
Thank you again.

Peter Darnell
Licensed Real Estate Sales Associate
Daniel Gale Sothebys International Realty
89 Seventh Avenue
Brooklyn, NY 11217
Yes, I wonder if there is a way for us to tweak the interface to simplify this somehow or help in cautiousness of interpretations when data is minimal, etc
 

Noah Rosenblatt

Talking Manhattan on UrbanDigs.com
Staff member
Noah,
Maybe displaying the total number of properties involved in reaching the final percentage would provide some value or lack thereof to the accuracy of the calculation.
noted...will keep it in mind as we shift design to chart updates in a week or so..thx!
 
Top