In multifamily real estate, selecting good rent comps is the foundation of a sound analysis. If you select good rent comps, you know how to adjust your asking rents and concessions to match the competition. You know if a potential acquisition may be undercharging or overcharging for the market. Most importantly, you know how much your asset is really worth, as rents are (perhaps obviously) very strongly correlated with market value.
But there are many different perspectives on what constitutes a good rent comp, and sometimes these differing perspectives can be the difference between correctly identifying an underachieving asset or one that is overcharging for its market.
This article explains how a few simple mathematical techniques can be used to identify and support rent comps for any multifamily property. Hopefully, it can help you close more deals with less stress.
Objectively Determine the Similarity of Your Comps
A simple way to identify and support comps is to come up with a method to define the “similarity” of nearby properties to your subject, and distill the similarity of each nearby property into a percentage value that can be used to filter down to the “best” comps. This percent similarity can be composed of several different factors — and different people will put different weights on each of these factors. However, as long as you have sound reasoning behind your choices, this approach will usually help justify your selections.
In the example below, we have 10 properties which we think might be comparable to the subject The Uptown Regency, located at 5050 N. Sheridan Rd., for which we have data on the physical distance from the subject property, year built, number of units, unit mix, advertised rents, square footage, and building and unit amenities.
Most wouldn’t argue that each of these variables are important to consider for a comp — but the importance of each is probably up for debate. Not a problem though… we can start by just weighting them all equally. A quick note on this:
In real estate, the old adage “location, location, location” holds true. However, location is really just a proxy for many additional variables including demographics and proximity to transit. In this case, the distance in miles from the subject should be a sufficient proxy to represent these location variables in the analysis.
The next step is a simple ranking exercise — rank each comp in terms of “closeness” on each of these variables, and sort them in a table. Now, the parts that are a bit tricky are the unit mix and the amenities packages. For the unit mix, the solution is actually not too difficult, just break out each detail separately.
If you rank everything on a scale of 1 to 10, with 10 being the MOST similar, your table should look something like this:
In this case, it’s best to break out each unit type, the square footage for that unit type, the rent, and the rent per square foot for that unit type separately. This way you can get a true apples-to-apples comparison for each unit type.
The amenities are a bit more subjective because the best way to account for similarity on these dimensions is with some human judgment. Sure, you could break out each individual amenity and verify what each comp and the subject has and does not have — but really it’s a bit easier to look at photos and come up with a “similarity rank” on your own (most real estate people are pretty good at this).
Very simple. Now, what’s the intuitive next step? Just add them up! The highest total will be the most similar comp, and the lowest will be the least similar. Check it out:
Now, you may be wondering “what if I don’t want to weight everything equally?” That’s a great question. In the city, we’ve found that proximity has a very big impact, whereas in the suburbs, things like year built and amenities are more important. If you want to weight each parameter differently, a simple way would be to just add a percentage value to the end of each row that totals to 100. The question you’re answering here is “What portion of the overall similarity of each comp is represented by each line item?” You can choose any breakdown you want as long as you have good reasoning for it. When you’ve finalized your breakdown, just multiply each total by the percentage to get the totals for each row. Here’s the same table, but with weighting for each variable called out and with items sorted by the sum column on the right:
The beauty of this approach is that if you can agree with all parties involved in the analysis/transaction on the weighting of these variables, then you don’t have to argue for the selection of comps. Any new comps you may want to add can also be added to the table and analyzed in this way — facilitating a true apples-to-apples comparison.
This is similar to the way Enodo’s comparable property detection algorithm works, except we automate the data collection and the calculation of similarity.