Introduction
Traditionally, commercial multifamily investors would generate deal flow via two primary methods: agent relationships and direct-to-owner physical mail campaigns. Once a property is brought into focus by either method, a study of the property's income-expense figures is performed to determine current net operating income (NOI), as well as gauge capacity to execute 'value-add', raising NOI and increasing the property's valuation.
I hereby describe an automated methodology to gauging commercial multifamily 'value-add' at scale across hundreds, or even thousands of off-market properties. Such data-driven analysis, albeit preliminary in nature, may constitute a momentous improvement over the prevalent exclusive single-property income-expense sheet analysis.
Focus of the article is on methodology and an automation mindset, though data sources, as well as approach to sourcing the data and its analysis may vary. Implementing this methodology can be technical and may require knowledge of coding. Usage of vendor commercial multifamily data, and no-code web scraping tools, combined with spreadsheet analysis may alleviate the need for coding.
The Rise of Python Coding, Internet Languages and Web Scraping
While there are various approaches to data sourcing and analyzing, Python coding as an approach has been growing in popularity immensely over the past decade.
Some common applications of Python coding include:
Web development
Data science, including machine learning, data analysis and data visualization
Scripting
An article for StackOverflow from 2017 titled 'The Incredible Growth of Python' states: "Python has a solid claim to being the fastest-growing major programming language”.
Additionally, understanding of internet technologies, including the practice of web scraping, has made it increasingly feasible to source massive data online, also bringing up legal questions and the need to source internet legal guidance.
Commercial Multifamily Valuation, Realizing Value-Add, Properties Data
Property valuation methods that can be used in valuating commercial real estate including commercial multifamily properties include: Income Capitalization Method, Sales Comparison Method, Cost Approach, Value per GRM, Value per Door, Cost per Rentable Square Feet. Income Capitalization Method is the prevalent and most common method, where the value of a property is derived as its net operating income (NOI) divided by a market capitalization rate (cap rate).
To illustrate how property 'value-add' is derived under the Income Capitalization Method, here is the net operating income (NOI) calculation for a sample multifamily property, with a formula for % value changing with NOI. Under a given market capitalization rate, a percentage increase in NOI results in a similar percentage increase in property value.
Sourcing "off-market" properties via agent relationships and direct mail campaigns has been a chosen path to finding "deals" for commercial multifamily investors, as compared to on-market commercial listings. Off-market properties are though typically not analyzed or compared at scale.
Scaling the Property Analysis Process, Off-Market: Data Sources, Methodology
Data Sources may include vendor-purchased commercial multifamily properties data or own data gathering, subject to legal guidance. Vendor-purchased data may not include properties rental info or income-expense insights.
Approach is to derive insights into individual properties’ income, expense using publicly available rental listings data. One large public repository of commercial multifamily rental listings data is CoStar’s Apartments, though there may be others as well.
Currently, there is little analysis of off-market multifamily financials done in a data-driven scalable fashion. Individual income-expense sheets are supplied to a purchaser / operator, and analyzed on a singular basis. There is limited capacity to harnessing off-market multifamily financials at scale for obtaining preliminary insights.
Commercial multifamily rental listings data may serve to provide insight into a property’s income and expense components relative to other properties in its immediate neighborhood:
Rent levels and unit mix
Other income: one-time and recurring fees including pet-related, parking, storage, administrative, application etc.
Occupancy rates
Utilities expense: Gas, Water, Electricity, Heat, Trash Removal, Sewer, Cable, Air Conditioning etc.
Below are the data and modeling steps hereby assumed.
Data Steps:
Source public data online, subject to legal guidance OR vendor-purchased properties data + rental listings data
Perform automated modeling on top of the data
Produce readable summaries
Modeling steps:
Compute estimated value-add (NOI increase % = property value increase %)
Rank properties within target markets, or across markets.
Final vector ranking to serve as predictor of future rankings of actual value-adds realized (predicting order, rather than exact magnitude of the value-add).
Data Sourcing and Modeling, Analytics
Sample Web Scraping Process:
URLs extraction: Performing search and extracting full set of listing URLs. Note: May use template input search parameters.
Contents extraction: Extracting page contents based on elements HTML, collecting data into a table, and combining multiple searches.
Output of data feed
Computing Property Analytics:
Compute Building class, utilize input parameters for Utility weights, Improvement categories and Field data type classifications. Building class a function of both built and renovated ages, Input dictionary files for assumed input parameters
Compute market table data frame aggregating average rent, average square feet, average bath counts across bedroom sizes, by neighborhood and building class
Compute fees table data frame aggregating one-time and recurring fees across town and building class combinations
Count units availability, Compute Unit weights and mix, Current rents, Potential rents, Rent increase %, Occupancy %
Compute Current other income, Potential other income, Other income increase %, Utilities expense improvement %
Final calculation for after-improvement value (NOI / value-add) %
Output feed can be generated providing preliminary analysis and ranking of thousands commercial multifamily properties across multiple regions. It gives estimates of:
Building class, unit mix, utilities % of total expense, current vs. potential rents and other income, occupancy and final contribution of these to after-improvement-value (NOI / value-add) %
Summary Analytics, Supplementary Data
Summary analytics by region can be computed off the data:
Commercial multifamily property counts at different unit size ranges, % of buildings in class A, B, C, D, average % utilities paid, average % occupancy
Top tier (high percentile): rent change %, other income change %, utils change %, AIV (NOI / value-add) change %
Can further add statistics for: Unit Mix, average recurrent and one-time other income, average rents
Output feed gives summary across multiple U.S. regions:
Commercial multifamily data and analytics, can be supplemented with owner information, such as full name, physical address, and other tax records information. Such can be sourced from:
Vendor-purchased data such as Reonomy, ProspectNow etc.
Extracted from County property tax search resources online
Conclusion, Validating the Model, and Applications
Ranking vector of commercial multifamily properties by % value add is expected to show sufficient correlation vs. subsequent % value-add realized on same properties, post buyer purchase and resale.
There are challenges, as plenty of commercial resale data over several years needs to be collected. A backwards study, assuming data availability, may be more suitable.
Model provides for preliminary analysis of a large volume of commercial multifamily data at scale. It allows to rank properties within specific markets, as well as across markets, better informing direct marketing campaigns, and overall investor focus. Potential applications include:
Determining top properties across markets
An easy study of unlimited markets inventories
A preliminary scalable automated look at buildings income-expense potentials
Data and Legal Guidance
Data sourced from online resources or in an automated fashion may be subject to terms of service contractual obligations. Data purchased from vendors may also have limitations for its specific use or share-ability.
It is advisable to consult an attorney specializing in internet law, intellectual property and online contractual obligations.
This article is for educational purposes to building an automated scaled process for off-market multifamily analysis. It does not include legal advice or prompt violating contractual or other legal requirements.
About the Author
Stefan Tsvetkov is the founder of RealtyQuant, a company that brings data-driven and quantitative techniques to the real estate industry. On a mission to add massive industry value through education, investment, technology, and analytics.
Financial engineer turned multifamily investor, analytics speaker, and live webinar host. He holds a Master's degree in Financial Engineering from Columbia University, and during his finance career managed ~ $90 billion derivatives portfolio jointly with colleagues.
Featured on multiple Podcast and Webinar events including InvestUp, Best Ever Real Estate Show, Discovering Multifamily etc. Organizer of Finance Meets Real Estate live webinar series.