All Forum Posts by: Art Perkitny

@Zak Marinko

I think this may help you out:

Income by neighborhood in Chicago (2018) with chart on the right showing income trend for Humboldt Park:

Absolute population change by neighborhood in Chicago (2010 to 2018) with chart on the right showing population trend for Humboldt Park:

Date used is from the American Community Survey (ACS). Neighborhood data is computed using a propitiatory imputation algorithm on ACS data since the ACS does not provide data tabulated by neighborhood.

Hope this helps 

- Art 

@Raymond Northcutt, possibly, I would start with this resource:

 https://valleycounty.maps.arcgis.com/apps/webappviewer/index.html

You may be able to scrap the owner info from the site and cross reference with tax addresses on the assessor website. If the tax address is different than the property address then the building in question likely has an absentee homeowner

@Drew Sygit

Really appreciate it! 

While conducting my research I also came to realize how large the city of Detroit really is. I came across this graphic that puts the size into perspective. 

And yes, within all the cities in the study, there will be sub-markets that sometimes have dramatically different investment results than the city average. In the future I plan on running the same computation but at the zip code and neighborhood level to see how the results change.  

@Kalen Jordan

The American Community Survey provides vacancy data every year at granularities ranging from the state level all the way down to block groups, which are much smaller than neighborhoods. Using this data you can impute neighborhood rental vacancy rates using various geo-statistical methods. 

Here is an example for a part of Cleveland using block group

The 2nd image shows the vacancy rate at the neighborhood level. I used block group data to impute these values. 

Hope this helps! 

@Scott Trench your methodology to calculate top cash flow markets is very interesting.  

So much so, that I decided to try and replicate your study. 

I will include a link at the bottom of this post for anyone that wants to check out the study as well as details on the methodology used. 

Here is what I found

MSA Level Study

Based on my study, the top ten cash-flow markets based on unlevered CoCR , tabulated by MSA, are the following: 

• Muncie, IN Metro Area
  Jacksonville, NC Metro Area
  Lumberton, NC Micro Area
  Terre Haute, IN Metro Area
  Beckley, WV Metro Area
  Meridian, MS Micro Area
  Sumter, SC Metro Area
  Fayetteville, NC Metro Area
  Charleston, WV Metro Area
  Yuma, AZ Metro Area

As you stated in your earlier post, in your study, the Detroit MSA and Cleveland MSA do not even make it into the top 50%.

I have found this to be true as well, Detroit come in at 171/333 and Cleveland 234/333 for CoCR.

Another point of interest for me was your assertion that rent to price ratios don't correlate well with cash flow. I decided to compare the two as seen in the following graph:

With an R2 value of 0.244, the correlation is indeed rather weak. 

City Level Study

I was curious to see if these results would continue when I ran the same analysis, but at the city level. 

Tabulated by city, the top ten cash-flow markets based on unlevered CoCR are the following:

• Detroit, Michigan
  Youngstown, Ohio
  Hammond, Indiana
  Cleveland, Ohio
  Dayton, Ohio
  South Bend, Indiana
  Buffalo, New York
  Akron, Ohio
  Deerfield Beach, Florida
  Lansing, Michigan

The above results are significantly more inline with the cities most commenters have been suggesting as top cash flow locations. The reason for the drastic difference in results is rooted in a fundamental problem that arises in geo-statistics called the modifiable areal unit problem (MAUP). The wikipedia link below does a good job of explaining the issue, but in short, geo-data tabulated by large geographies, such as states or MSAs, tend to obfuscate metrics by averaging out the data over the whole geography. Essentially, the resolution is not high enough to capture pockets of high and low values. 

https://en.wikipedia.org/wiki/Modifiable_areal_unit_problem

Take the two graphic below as an example. The first shows the median household income tabulated by MSA and the second at the county level. 

Notice how the median income at the MSA level is 58k. However, when looking at the same area, but at the county level, we can clearly see that Wayne county has much lower incomes than say Oakland county. This averaging at the MSA level is what is causing Detroit to look less like a cash flow market than it really is. 

This can also be seen in the difference in median RE tax paid between the MSA and City level. 

Detroit-Warren-Dearborn-MI-Metro-Area

Detroit City

This lines up well with that @Travis Biziorek said in an earlier post regarding RE tax paid on his primary residence in Troy vs his investment properties in Detroit. The data suggest that the median per month RE tax cost should be $111 while at the MSA level it's $235

As with MSA data, I wanted to compare the rent to price ratio with CoCR values at the city level. 

This time the R2 value is 0.681, suggesting a moderate correlation between rent to price ratio and CoCR. In a follow up study, I'd be interested if the R value will continue to increase when the market granularity decreases. 

Top Cash Flow Market based on CoCR over time

I really like what @Russell Brazil said regarding looking at CoCR not just today, but in the future. Essentially, when rents go up in a market, your CoCR will go up since your cost (purchase of property) is fixed. This means yields will increase as time progresses. To compute the best markets for CoCR over time I took the difference between CoCR at 2011 home values and CoCR at todays values to get a delta value. When this delta is positive, the cap rates in the market you invested in have constricted. Meaning your yield is higher than what you could buy into at today's prices. Based on my study, the markets that have experienced the greatest increase in CoCR over time are the following: 

• Moreno Valley, California
  Deerfield Beach, Florida
  Pittsburgh, Pennsylvania
  Buffalo, New York
  Sunrise Manor, Nevada
  North Las Vegas, Nevada
  San Bernardino, California
  Surprise, Arizona
  Palmdale, California
  Spring Valley, Nevada

Here are the cities with the worst CoCR delta:

• Akron, Ohio
  Mobile, Alabama
  Boca Raton, Florida
  Bridgeport, Connecticut
  Youngstown, Ohio
  Roanoke, Virginia
  Toledo, Ohio
  Detroit, Michigan
  Dayton, Ohio
  Cleveland, Ohio

Interestingly enough these line up well with the top cash flow cities based on todays CoCR. All of these cities have negative CoCR deltas suggesting that the perceived risk in all these markets has increased. These results sort of echo @Jay Hinrichs and @Michael Ealy sentiment championing higher quality B/C assets over cheaper D/F properties. 

Issues and Improvements to Study

As with any study, there are things that can be done better or differently. Here are some things that I would like to change and/or improve in a future analysis.

Insurance data could be more granular. The data I am using to compute insurance premiums is averaged at the state level. I think that if the data was tabulated at say the zip code or lower level, the results would be changed. Things like crime rates affect insurance rates, for example, and state level data does not do a great job in capturing these nuances.

Certain locations have state wide ordinances that affect how property taxes are assessed. For example in California, Under Proposition 13, the annual real estate tax on a parcel of property is limited to 1 percent of its assessed value. This "assessed value," may be increased only by a maximum of 2 percent per year, until and unless the property has a change of ownership. This creates an issue since tax rates in my study are calculated on taxes paid by current home owners. Meaning, depending on how long they have help the property, rates will be a lot lower than if an investor bought an asset today. This also changes the CoCR over time computation since taxes in CA won't increase as quickly as other areas. 

A big issue and one that a few people have already commented on, Capex and maintenance data leaves a lot to be desired. I agree with @Will Barnard that using a fixed capex value for all locations creates an unrealistic scenario. As a result, in my study I ran my Capex and maintenance numbers as 10% of median contract rent. With that being said, I have included an adjustable parameter in my spreadsheet that allows the numbers to be re-computed based on a fixed ($250/month) capex and maintenance budget for all locations. I'm open to suggestions as to where I could get better data on this subject. 

Data on evictions would be a great addition to the study. These tend to be more expensive than general vacancy or turn-overs. 

Utilities, while commonly passed onto tenant is SFR, may alter results especially when considering MF.

Notes on Data and Link to Study

• City is Place from census and MSA is CBSA from census
• Place > 50k pop
• MSA > 100k pop
• Eliminated any that where missing key data points
• Data is based primarily on data from SFR

More info on data sources and methodology is located via the link below

Link to Study:

Top CashFlow Markets Study

@Chris Skutnik

Traditionally cap rates aren't used to assess residential real estate (less than 5 units), which I figure is what you are in the market for. 

With that in mind, we can still use cap rates to get an idea of the returns we should expect in a given market. The caps will be rather low compared to commercial properties in the same market due to how residential real estate operates. 

The map below shows cap rates for single family properties on long island tabulated by zip code: 

As you move west to east cap rates change from the low 1s to high 3s and 4s. 

These figures were computed using data from the American Community Survey and assume the 50% rule when calculating NOI.

Hope this helps! 

@Tyler D.

As others have mentioned, the number of 4-unit properties in Pittsburgh is very low. 

Data from the American Community Survey shows the following distribution of units by building size in the Pittsburgh market

11,240 units being reported as located within a three to four unit building suggests a total of about 3,200 buildings. 

That being said, if we take a look at more granular data for neighborhoods, we can at least identify one where the concentration of fourplexs in highest. The map below illustrates this: 

The Shadyside, Bloomfield, and East Liberty neighborhoods seem to have the highest number of 3 to 4 units. 

Here is the same breakdown as above for Shadyside as an example: 

I would focus my efforts on these areas if you are keen on procuring a four unit in the Pittsburgh market. 

Hope this helps! 

@Jason Appel

It's difficult to say exactly since these numbers will vary from market to market. With that in mind, look for areas with growth in population, income, and home values. 

Also, try to locate cities with job growth above 2% and crime rates that are falling. 

Low poverty and unemployment rates are also good indicators of a good area. 

Try to target sub-markets with incomes over 35k since anything under this will usually yield troublesome tenants and cause high rates of turn-over. 

Hope this helps! 

@Joseph Mejia

Not sure if you are planning on relocating to start your operation but It may be difficult to run a wholesaling business remotely.

That being said, it's worth your time to dig into the data to get a better understanding of the markets out there. To aid in this, I have included a list of datasets, that I've found in my research, which will provide you with a number of key performance indicators (KPIs) that you can use to evaluate markets at all levels.

Census Bureau: American Community Survey (ACS)

The ACS was launched in 2005 and is administered every year by the Census Bureau to roughly three and a half million households in the USA. This dataset contains over 18,000 data points, which are tabulated at the state, MSA, county, city, zip, and block level.

This survey contains hundreds of KPIs that investors should use when searching for new markets to invest into. Here is a list of a few that I find to be of most value:

• Population Total
• Home Values
• Household Incomes
• Poverty Rate
• Population Age
• Rental Vacancy Rate*
• Homeowner Vacancy Rate*
• Educational Attainment Rate (High School/GED & Bachelors)
• Number of Housing Unit
• Rent to Income Ratio*
• Rent to Price Ratio*
• Population on SNAPS (supplemental nutrition assistance program) percentage
• Property Tax Rate*
• Median Age of Buildings
• Number of Structures by Units (SFR, Duplex, Triplex, Quadplex, etc...)
• Median Rents by Number of Bedrooms
• Population Change (Births, Deaths, In-Migrations, Immigrants)

*note that some of the metrics are composites that are computed by comparing two or more variables from the ACS

A number of the variables above can also be broken up into cohorts, which will give you even more insight into how an area is doing. For example, an area may show a low median income, however when looking at the cohorts, it may show a bimodal distribution of income at both the lower and higher end. This could be an indicator of an area that is going thought the early stages of gentrification.

I'd also encourage you to look at the deltas, or change, for some of the indicators mentioned above. This will better help you get an understanding of not just where market is today, but also where it may be heading. You wouldn't want to buy into an area that is declining rapidly where the tenant quality is decreasing along with the area.

Bureau of Labor Statistics: Current Employment Statistics (CES)

The Current Employment Statistics (CES) program produces detailed industry estimates of employment, hours, and earnings of workers on payrolls every month. CES National Estimates produces data for the nation, and CES State and Metro Area produces estimates for all 50 States, the District of Columbia, Puerto Rico, the Virgin Islands, and about 450 metropolitan areas and divisions. Each month, CES surveys approximately 142,000 businesses and government agencies, representing approximately 689,000 individual worksites.

The CES is the gold standard when it comes to analyzing the job growth of a market. As an investor you will want to look at the number of jobs add/removed within a market. In addition to this figure, the CES provides more granular data for jobs tabulated using the North American Industry Classification System (NAICS). This system allows you to get a very detailed picture of how the local job market is composed. You will want to avoid markets where a single industry dominates the local economy in the number of jobs since these kinds of markets a incredibly susceptible to swings in the market.

Census Bureau: Building Permits Survey (BPS)

The purpose of the Building Permits Survey (BPS) is to provide national, state, and local statistics on the number and valuation of new privately-owned housing units authorized by building permits in the United States. The Building Permits Survey covers all "permit-issuing places," which are jurisdictions that issue building or zoning permits. About half of the permit-issuing places in the United States are surveyed monthly. The remainder of places are surveyed annually.

Using data from the BPS, investors can get an idea of the supply and demand dynamics within their market. Coupled with migration data from the ACS, this can be a very powerful dataset to use when attempting to predict future housing prices, vacancy, and absorption rates.

Federal Bureau of Investigation: Uniform Crime Reporting (UCR)

The Uniform Crime Reporting (UCR) program compiles official data on crime in the United States, published by the Federal Bureau of Investigation (FBI). UCR is a nationwide, cooperative statistical effort of nearly 18,000 city, university and college, county, state, tribal, and federal law enforcement agencies voluntarily reporting data on crimes brought to their attention. UCR data is published annually by the FBI.

UCR data is reported as either violent or property crime, which then break down further into eight possible sub-crime categories. Investors will want to ensure that the areas they are investing into have low rates of crime, or are at the very least improving over time.

Consumer Financial Protection Bureau: Home Mortgage Disclosure Act (HMDA)

The Home Mortgage Disclosure Act (HMDA) requires many financial institutions to maintain, report, and publicly disclose loan-level information about mortgages. HMDA data is the most comprehensive source of publicly available information on the U.S. mortgage market. HMDA data is released annually.

Investors can look at the types of loans being given, weather the property is used a primary residence or investment, and the demographic information of the borrower. It's possible to also tabulate the data by the type of property (residential or commercial). This data is tabulated at the census tract level, which allows for a tremendous about of granularity when conducting market research. Investors should use this data to identify areas where other investors are buying as well as to get a better understanding of the demographics of buyers in the area.

Census Bureau: Decennial Census

This is the census that most of us are familiar with. It has been administered to all households in the country every decade since 1970.

Investors should use this data set when more granularity or long term trend data is desired than what the ACS offers. However, one should be leery of the data once a few years have passed since the census was taken since demographics can change rather quickly in small areas.

Honorable Mentions:

The datasets below are ones that I have come across in my research, but have not yet had the chance to fully go through. Nonetheless, they will likely provide investors with even more actionable data.

• Internal Revenue Service: Statistics of Income Migration Data (SoI Migration)
• Census Bureau: Value of Construction Put in Place Survey (VIP)
• Census Bureau: Current Population Survey - (CPS)
• Department of Housing and Urban Development (HUD): Survey of Construction (SoC)
• Census Bureau: Survey of Market Absorption of New Multifamily Units (SOMA)

Remember that real estate is hyperlocal and it's important not only to find a market that posses mostly positive indicators but also a sub-market that has the same positive characteristics. Locations with poor indicators will attract poor tenants and the quality of the tenant base is often more important than property you select itself.

I think it's more important than ever for investors to look at the data as we reach the end of the longest bull market in US history. Investors that allocate into places with good market indicators will fare much better in the looming recession and will avoid making an extremely costly mistake.

By looking at the data you will be leaps and bounds ahead of the competition.

Let me know if you have any other questions, hope this helps!

@Farzan Setayesh

@Farzan Setayesh

Not a local investor, but here is a data driven look at the market conditions in McAllen Texas

Two of the most important market performance indicators, population and median home values, have been increasing steadily for the past decade   

Rents have followed a similar trend 

The chart to the right shows median household incomes over time, which have also been increasing. 

Lastly, education rates are up somewhat since 2010 while poverty rates are down.

SNAPs (supplemental nutrition assistance program) rate is up a bit however. 

All in all, the data seems to agree with what you read. The key performance indicators suggest that this is a good growth market with a lot of future potential. 

Hope this helps!