Two methods to get the scoop on appreciation in your neighborhood, for an existing home or an investment. Includes a twenty second and a fifteen minute version.
I’ve said many times before that real estate is hyperlocal; it explains why your aunt in Arizona and your brother in California have such different views of whether real estate is a good investment. Because of this, one of the most important pieces of data to understand when it comes to real estate is how much appreciation your individual market is experiencing. Not just your state or even your county, but down to the city and even individual neighborhood. The information you will find on this point is pretty abysmal. You might find a metro-area view. But what if, as is the case in many cities, the neighborhoods differ widely in price point and growth? You could ask your real estate agent, but I interviewed three agents and their responses on this front were lackluster. They could pull recent comps for me. If the Zillow city level view was too wide in scope, the comps list was too narrow. It didn’t cover a large enough time period to evaluate growth rate over time.
Fortunately, you can take the matter into your own hands. I present to you two tools: one which takes 20 seconds to search and gives a city view, and one which takes 15 minutes and uses direct MLS data to create a custom data set for any level of specificity you desire, all the way down to a single neighborhood or side of the highway.
20-Second Tool: Zillow City-Level Home Data
Google for “Zillow (Your City) Home Prices.” You want to find the page for your city where Zillow compiles all its market data, not the usual listings portion that makes up the bulk of the site. The search result should look like this.
The market page on Zillow has some neat information. You’ll see a region map to show you what data was included. Then you get an overall readout of the most recent year’s market appreciation.
The interface allows you some level of customization. As you scroll down, you will get to a section that allows you to sort on home type and number of bedrooms. This is important, as your market might be seeing huge appreciation for starter condos for young 20-somethings, but a more mature market for the single-family four-bedroom home you own. You can play with the toggles and run your mouse across the chart to see prices in each given month. With that data you can compute your own appreciation for the desired time period (i.e. 1 yr, 3 yr, etc.)
While I think the Zillow tools are a very strong starting point, I have several concerns about the analysis.
Neighborhood Individuality – Firstly, a city-level view often does not capture the true picture of a given neighborhood. Particularly if you live in a highly urban area, you know very well that two neighborhoods sitting next to each other can command very different prices and growth prospects. In the NYC area, two avenues can be the difference of being in the hood vs. a fancy pants nanny and stroller neighborhood.
Home Size – Secondly, the Zillow tool allows no distinction based on home size. In many areas, land is the most valuable part of the home. This tool doesn’t distinguish between 3-bedroom homes that are 1500 sq feet vs 3-bedroom homes that are 2400 square feet. This matters when you are computing home appreciation when the mix of homes is changing over time. Again taking highly dense areas as an example, you will probably notice homes being built in smaller and smaller footprints. Since land is so valuable, newer developments go from single-family homes to townhouses, or from towhouses to condos. Then it’s an average size shrinkage from 1500 sq feet to 1100 square feet. You can see this in Seattle. The price per square foot in Seattles nicest neighborhoods has almost doubled in five years. You will see a ton of new construction of smaller townhouses now. If the mix of inventory changes, this will skew your appreciation data.
In making a real estate decision worth several hundred thousand dollars, this level of granularity was not enough for me. That’s why I wanted to get access to direct MLS data and create my own data set. I show you how to replicate that process below.
15-Minute Comprehensive Tool: Redfin Raw Data
We are going to pull our data from Redfin. Redfin is a brokerage itself. It displays all of its data directly from the MLS, as it has access being a real estate broker.
Type In The Market You Want To Research
Here I’m going to look into a specific neighborhood I’m interested in in Jersey City, a city directly across the river and one subway stop away from Manhattan.
Zoom In On Your Target Area
The beauty of this technique, as I said before, is the level of customization you can use. We are going to zoom into one particular corner of Jersey City, the area known as downtown JC. This is important because Jersey City includes an extremely diverse set of neighborhoods. Downtown JC is highly gentrified and sells for over $700 per square foot. Then you have areas that are much less gentrified and go for $400 per square foot. We don’t want our data to be skewed by incorporating such different neighborhoods.
Here, I’ve used the plus and minus areas to zoom in.
Add Your Filters
Next, we want to go to the section entitled ‘More Filters’ near the top right hand side of the screen. Here you want to select Sales Records rather than active listings. You can get all the transactions in, for example, the past three years. You can also then narrow the search to homes which match what you’re looking for. In this case, I’ve narrowed my search to the following:
- 1 bedroom
- 500-750 square feet
- Downtown JC location
Download The Data
Finally, I scroll down the listings on the right hand side of the screen until I get to the buttons to turn to the next page. There, right about the page buttons you will see a selection to ‘Download All.’ This is the data goldmine you’re looking for. Click this and it will load all of the listings into a spreadsheet.
Note that Redfin limits the number of data points it will pull into your sheet to about 350. They do this because they don’t want third parties writing API’s to bulk download their data for commercial use. If you find you have more data points than will fit in one download, you can stitch together what you need by zooming into smaller sections, hitting download, and then eventually copy all the cells into one combined spreadsheet.
Note (08/04/2017): Some readers are reporting that they do not see a Download All button for their area. I’ve reached out to Redfin who has said they make an attempt to provide data for all cities, but in certain areas they are prevented by MLS rules from sharing that data publicly. If you do not see a Download All button, it is likely the area you are searching falls into this category.
Clean And Analyze Your Data
When you open the file, you see an overwhelming set of data. Don’t worry, you’ll be ignoring most of it. In fact, you’re really looking at one particular column: price per square foot.
You want to be able to sort it by the year in which the transaction took place. To do this, you can write a quick little new column that pulls only the year the property was sold. Column B shows the actual Sold Date. Create a column next to it and enter in the formula below. This says ‘take the right four characters from the Sold Date column. That would be the year it was sold.
You now have all the data you’ll need for the exercise in just two highlighted yellow columns:
In fact, because I find the presence of all those extra rows distracting, I’m just going to hide them all. You could also just copy and past these two columns into a new sheet.
We have one last step before you appreciating data will magically appear. Every data set requires some cleaning and prepping. In Redfin’s data, you will notice some fields are blank. We want to make sure these data points don’t skew our data set so we remove them from the calculation. To do this, we want to sort the data set by largest to smallest value on each of the two columns, and delete anything at the very bottom that has a blank entry.
You see how after the sort completes, there are a bunch of blank cells gathered together near the top for Year Sold? we’re going to remove those rows. Then we sort on the column $/Sq ft and delete any row which does not have a value for $/Sq Ft.
Finally, you can sort again by Year Sold. You know have a chronological list of every transaction in that neighborhood and what it sold for in $/Sq Ft. Simple take the average of each year. You can, for example use the formula ‘=median(___) and select all cells with the year 2016 to get the median. You can take the mean as well.
You then compare it the figure from one year to the next and you have all your appreciation figures!
Here they are for our test case, a 500-750 square foot 1 bedroom in the downtown JC area:
Keep in mind that this strategy works best for markets in which units are particularly uniform and/or expensive markets where most of the value is in the land. In order to compare all this data, we will be using metrics like square footage. If there’s a unique historical mansion on the market with detailed antique carvings, inlaid gold finishes, and its own pet bear in the backyard, it’s obviously not going to fit in on a price per square foot measurement with the rest of the cookie cutter houses in the new development down the street.
What is appreciation like in your area? Were you surprised by the results?