The Real Estate industry contributes close to three trillion dollars to the total US GDP. That share continues to grow each year as is indicated by the graph below. While real estate investments can range from low risk/low return to high risk/high return, investors should anticipate a long investment cycle. Due to the volatility in the real estate market, particularly after the 2008 financial collapse it has become critical to minimize the real estate investment losses. Investors need to carefully research the real estate market to mitigate the risk. The risk factors can potentially comprise of location of the property, demographic and economic factors as well as several physical property attributes such as property size, number of units, unit size, amenities, parking spaces etc.
Many organizations in the real estate investment management business struggle to integrate their internal data points and external/market data that could provide even richer sets of data for analysis. These organizations mostly rely on traditional reporting methods which only provide rear view or historical trends of their investments. Beyond providing historical trends, effective and comprehensive real estate analytics should reveal in real-time what is happening now and what will happen in the future. Though many analytical factors are shared across different asset classes (single family, multi-family, commercial etc.), the following paragraphs focus mostly on multi-family rental properties analytics.
The valuation of a property is one of the critical initial steps towards deciding whether an investment will yield positive or negative results. At minimum, in order to determine the potential value of a property, investors input multiple facets of real estate information into a financial model that compiles the data and generates various results. These results determine if investing in a particular property will yield positive results within the investment time horizon. When considering to invest in a particular property, the following preliminary data should be collected:
• Physical Property Details: This includes information like size, number of units, parking spaces etc.
• Financing Information: This would include the loan details (interest rate, amortization, loan terms etc.) that you would acquire to pay for the property.
• Purchase Details: The price you will have to pay to purchase the property plus any additional upgrades you would require to make the property rental ready.
• Income: This includes the monthly income that the property would generate
• Maintenance Expenses: Includes the cost of maintaining the property, managing the property, property taxes, insurance, and other costs related to general maintenance and upkeep of the property.
• Rent Roll: An itemized list of current residents by each unit and the amount of rent collected from each resident in a multi-family property.
• Tenant Profile: Includes things like credit history, credit score, employment history, family size, education, past rental history and more.
• Current and past operating statements: These statements include itemized list of income received from a rental property and the expenses incurred to maintain that property.
In addition to the specific property data mentioned above it is equally important to keep an eye on the current market environment including economic and demographic trends. In many cases the investment process starts with an analysis of the broader market to determine the feasibility of a location. Usually job growth markets along with favorable demographics contribute towards stronger and long-term tenant demand for apartments. For example a location with rising employment opportunities, income growth, and young adults (25-34 year-olds) living with their parents has a tendency for a pent-up demand for apartments. Below are few of the external data points that should be analyzed individually as well as in correlation with each other to determine the location viability of a property:
• Unemployment Rates/Income Statistics
• Consumer Debt as a Percentage of Disposable Income (Debt Barrier)
• Age Demographics
• Apartment Supply (multifamily permits)
• Vacancy Rates
• Rent Growth Statistics
• Urban Living Trends
The Following are few examples of graphs showing comparisons/correlations between different data points. Exhibit-3 below shows the correlation between gross income growth and vacancy rates. It is easy to see that the vacancy rate and the gross income move in opposite directions meaning that as the gross income increases the vacancy rate goes down (e.g. higher apartment demand). This type of analysis can also be performed at the local market level. For example in the specific MSA (metropolitan statistical area) where the subject property exists.
Exhibit-9 below shows the rent growth in different metro areas throughout the USA. Rents continue to increase the most in markets in California and Seattle. By having both macro and micro level market knowledge an investor can make an informed decision towards investing in a property.
Once a property or set of properties has been acquired it is important to monitor the performance of individual properties or portfolio of properties. There are two types of analysis that can be performed:
This type of analysis allows you to monitor your investment from different perspectives or dimensions. For example you may want to see the current market value of your investment(s) in different regions of the country or at a more granular level such as the MSA (metropolitan statistical area). You may also want to compare your average rent with the market rent over a certain time period. Another example of multi-dimensional analysis would be comparing the occupancy with the market occupancy in a local or broader market. The multi-dimensional analysis provides the ability to explore your data interactively.
Benchmarking is another example of multi-dimensional analysis where an organization would want to compare their investments with the industry performance. This is the type of analysis where quality external data plays an important role in the overall data management and analytics strategy. One such data source is NCREIF (National Council of Real Estate Investment Fiduciaries) which maintains an index that measures the performance of income producing real estate. Institutional investors still rely on this index to benchmark their investments. Investors usually want to compare their investment allocations against the index by property type and regions.
Data Mining/ Advanced Analytics
Data Mining is the process of finding hidden patterns in a set of historical data. This technique is used for both exploratory and predictive purposes. There are many statistical algorithms available which can be used for data mining. Few of the commonly used statistical algorithms are explained in the following paragraphs.Association Rule Mining finds associations and correlation relationships between several data points. Association rule shows attribute value conditions that frequently occur together in a given data set. For example price of a property could be associated with other attributes like year of sale, property type, location etc. This type of analysis can potentially uncover a hidden association in a large data set that may have not been visible through traditional reporting techniques. Special attention can be paid to those attributes that have a strong correlation with the property price.
Regression is a statistical algorithm that explains how a set of independent variables impact a dependent variable. In the case of a property, one of the dependent variables is the renewal probability of a lease, where as independent variables are the different attributes related to the property and external economic/demographic factors. The renewal probability ranges between 0% and 100%. The data collected is assumed to be thousands of lease renewals over a period of time from various MSAs (metropolitan statistical area). This type of analysis helps in providing a better understanding of issues that impact the probability of lease renewals and therefore can help in retaining tenants. The following are examples of a few independent variables that can be fed to the regression algorithm:
Number of Stories
Vacancy (an economic variable indicating vacancy rate in a sub market)
Employment (another economic indicator showing employment rate in a sub market)
Demographic factors related to tenants
Clustering is another data mining technique that groups similarly situated objects into a cluster. The goal is to identify sets of clusters with a high similarity within a cluster and low similarity between clusters.
This technique is very useful for segmentation purposes. For example in the case of property leases, tenants can be segmented into different clusters based on their shared characteristics. Segmentation can help in identifying different attributes of tenants which may result in long term or multiple lease renewals vs. the ones which terminate their leases pre-maturely. By analyzing the contributing attributes one can devise strategies to retain tenants or help in creating more effective and targeted marketing campaigns to attract the best renters.
Clustering can also expose trends that otherwise may not have been obvious through traditional reporting methods. For example, the lease renewal rate could be very high for a particular property leading one to conclude that operations are running efficiently. However, a clustering exercise could expose a cluster or group of highly profitable tenants who have been terminating their leases recently. This group of tenants may have been looking for certain amenities that are not available at the property at the moment. With this knowledge the property management can evaluate whether it is economically feasible to have those amenities built at the property in order to retain these high profit renters as well as attract new tenants with similar profiles.
With so many different data sources and volumes of data available it is challenging to collect, integrate, store, and manage all the data. But as long as real estate organizations are focused and set clear goals for their data management and analytics initiatives the opportunities are endless in data analysis and predictive analytics. Companies who are successful in combining different and varying data sources together to create multi-faceted analytics across these boundaries are positioning themselves to have the most unique competitive advantage in the marketplace.
For more information about implementing an effective real estate analytics program; Contact Us