Finding Our Market: A Census Data Dive

[Chris Salvato]

I am hoping this post will help some others on here who are trying to make decisions on where to invest.

My wife and I are looking for the market we want to tackle, and this post will describe our process.  Part of our process included using my skills as a developer/data analyst to pull US census data for every county in the USA.  Here is a link to that data (sourced right from the US Census Bureau) for your own perusal:

https://cl.ly/3w0i0s3V2o3m

This data is free to the public but VERY hard to access. The government doesn't really know how to make anything easy.  For interested parties, the rest of this post is the story of how we are using this data, and continue to use this data as we try to find our market.

Back Story

A few months ago my wife and I decided to get into real estate investing and started doing a ton of research.  We determined that our end-goal is to have a portfolio of rental properties that generate $10k/month in cash flow within 5 years, with our stretch goal being $20k/month within 5 years.

To kick off our entry into real estate, we decided our first purchase would be a "house hack" where we live in one unit, and rent out the other(s) in a duplex, triplex or quad using an FHA loan on our first deal.

Right now, we are fairly mobile and can move just about anywhere.  We are looking at this as an advantage, because we can pick any city in the country to move to. But it's also been an Achilleas Heal because it's easy to paralyzed by options.

To narrow down our options, we want to move to a place with a better-than-average chance of success, that is also a place we really want to live (i.e. more on the urban side, less rural.  We don't like a rural life).

So rather than throwing darts at a map, we decided to look at data in the entire US, and set a criteria.  Any areas that met that criteria would be worth visiting, assembling a team, and moving.

To do that, I needed to look at US Census data.

Our Criteria - What are we looking for?

Since a big part of my career has been analyzing data, I know it's a lot easier to analyze data when you go in with a particular set of questions or criteria you're looking for. 

Here are the things we wanted to look for, which drove what data we needed to pull, and what we calculated:  (Maybe we are looking at the wrong things?  Input is welcomed...)

• 5 Growth
• Low value of property (so it's easier for us to enter the market)
• High cost of renting
• High percentage of units being rented vs. owned
  
• High percentage of renters renting property with 3x income
• Low vacancy rates
• Low unemployment rates 

Of these metrics, we couldn't get a few from the census...at least not yet.  Those we could get are bolded above.  Those we couldn't (or haven't found a way to get it yet) are not bolded.

Our Criteria - What are the numbers?

We went in not knowing what a good number was for a lot of the metrics we were looking at.  So we looked at the data set overall and figured what a good value was based on national averages and regions that we knew we liked.  That led us to the following numbers:

• 5 Year Growth > 0 (the bigger the better)
• Gross Growth > 10,000
• Total Population > 100,000
• Median Home Value < $350,000 (which means a good deal would be ~$245,000)
• Must be in a state we want to live in (e.g. TX, NY, PA, VA, MD, CO, CA, NC, NJ, HI, FL, GA)
• Rental Vacancy Rate < the national average + 1%
• Unemployment < the national average + 1%
• Low unemployment rates 

Navigating US Census Data

Naively, I thought US Census data would be easy to access.  That's kind of stupid, because nothing provided or done by the government is easy.

For anyone wanting to look at US Census data on your own, go to http://factfinder.census.gov/faces/nav/jsf/pages/guided_search.xhtml and poke around for a bit.  

(I am not really in the mood to type up a 10 page article on how to use the census. It took me about 3 days to figure it out.  If people are interested, I may write up a guide or create a video for you all.)

The long and short of it is that you pick the metrics you are interested in (say, total population) and the geographies you are interested in (say, counties, states or places/cities/towns/villages).  Then you can generate a report that has those metrics for that area.

Unfortunately, the US Census only has a few reports that pull various kinds of information together (e.g. Population and median home value).  If you're an uber-nerd like me, you need to pull the different data sets out for different metrics, and cross reference the data based on geographies...but that's a topic for another time, (and likely just for data nerds like me).

In the end, we were able to pull various data sets that shed light on various areas.  The attached file above is for counties, but we pulled county and ZIP code data, so we can get a higher level and more granular view at different areas.

From those data sets, we were able to pull/calculate:

• population in 2015
• population density
• gross growth
• growth 5 years
• percent white population
• median home value dollars
• median owner costs with mortgage
• median owner costs without mortgage
• median gross rent 2014
• percent units rented
• rental vacancy rate
• home value to income ratio
• mean income dollars
• unemployment rate 16 and over
• percent renting for at most one third income

Narrowing Down

When we ran all the US counties through the criteria only 66 of 3140 counties met our criteria.  Talk about narrowing down!

But 66 is still too many to look into at once.  We really need 1 area that we can focus on and blast through.  So our next idea was to take all of the counties that match these criteria and plot them on a map.

Once they are on a map we will be able to instantly see a few things:

1. Are these places we actually want to live?  (e.g. are they near cities or really freaking remote?)
2. Are multiple counties really a single area? (e.g. are we looking at 10 counties all around the Dallas area?  If so, we can just look at Dallas as a whole, and start our search in the counties/ZIPs that look most promising)

These are the maps we created:

Florida

Missouri

Georgia

Maryland

North Carolina

Virginia

Pennsylvania

New York

Colorado

Texas

After looking at these maps, and reflecting on where we would want to live, we narrowed it down to the following 5 (in order of personal preference):

• Denver Colorado
• The San Antonio -> Austin TX Corridor
• The Washington, DC to Richmond VA Corridor
• Southern Maryland
• Raleigh-Durham, NC

Next Steps

We are still discussing what our next logical step is...but we are fairly certain the next immediate step is doing a more qualitative analysis of the areas: reading business journals, reading local newspapers, investigating school and community development initiatives, investigating school quality, etc.

With that qualitative analysis, we will likely eliminate another 1-3 cities, then plan to build our team and network with local investors 

After that, we will build our team in those cities and plan a visit to build relationships and vet properties.

I hope that revealing part of our process helps at least one person, even if it's just learning something about your market from the HUGE CSV FILE I posted.

Do you agree with this process? Disagree?  Too much thinking?  Looking at the wrong stuff? Looking at the right stuff? :)