Doing A Comps Analysis
Republished from April 2011 -- A: I can see it already, I'm going to get a lot of haters and strong opinions on this discussion. That's all fine. This site has always been about open debates and discussions and everyone is entitled to their own opinion; including you and me. This discussion is my opinion on how to do a Manhattan Comparable Sales analysis.
In my opinion, this is the most important aspect of a buyer brokers job these days. Understanding how to analyze past comparable sales, apply adjustments for time/views/reno/etc.., interpret current market & submarket conditions, and come up with a fair market opinion trading range for a target property to me, is the meat and potatoes of the buyers broker process. Lets get right into it.
The biggest mistake brokers make when doing a comps analysis is --> CHANGING THE VARIABLES!
AS YOU INTRODUCE MORE VARIABLES, YOU DEGRADE THE QUALITY OF THE COMPARABLE SALES ANALYSISThe key to a quality comparable sales analysis is to resist the temptation to focus on recency at the expense of relevance. What I mean is, if you have a highly relevant same-unit comparable sale that closed 14 months ago and disregarding it to focus instead on a smaller, lower floor sale that closed 3 months ago. Here are some other common mistakes:
Situation #1: Do a comparable sales analysis for a co-op, by comparing it to a different condo sale!
The Problem: How do you know how to quantify the premium that condos take in over co-ops in the open marketplace? I certainly don't know. Then you got two different buildings which means two different apartment layouts, exposures, etc. This is like comparing apples to chocolate; its not even the same kind of food and the analysis will be completely degraded!
Situation #2: Keep property type the same but look outside the building where the target unit is, and use a comparable sale from another building. Usually this is done when attempting to justify a higher price. The in-building comps say one thing, the seller or broker wants more, so you ignore relevant sales and look outside the building for better ones.
The Problem: How do you quantify the differences between the two buildings and how any one buyer will perceive and bid up one difference over another? When comparing two buildings you will notice:
a) varying service levels in the bldgs
b) varying amenities in the bldgs (storage, gym, roofdecks, sublet policy, pet policy, board purchase requirements, etc.)
c) varying building financials, reserve funds and assessment history
d) varying levels of MCI (major capital improvements)
e) varying apartment lines, economies of scale in running each building
f) at times, varying school district zones
How do you adjust for these things?? The answer is you can't! Nobody knows on a quantitative level how one buyer values one building's characteristics over another. There are no ways to measure and adjust for these things. You can't use the argument that "its a recent sale and worthy" and then ignore all the other variables. That is why you should always STAY IN THE SAME BUILDING and keep those above noted variables fixed as much as you can!
Situation #3: Analyzing a different apartment line or size in same building when you have more relevant sales to focus on.
The Problem: When you change an apartment line or size, you must adjust for it. Size is easy enough assuming its a condo and you have the stated SFT. It gets trickier for co-ops where the total size is either missing completely or estimated by the listing agent. Using a different apartment line is a bit tougher. One line may enjoy open S/W views, arguably one of the most desirable views in the city without specifying a need for river or park views. Another line may have direct North views, indirect sunlight reflecting off the neighboring building with views of mechanicals on rooftops. So even changing the line (while keeping the property type the same) may have an affect on your comps analysis.
THE SOLUTION --> Make every attempt to keep the variables constant! Same unit sales are always preferred, which by default would mean, same line sales in the same building are the next best option. Only stretch to smaller or larger units in-building if you have to; knowing full well that studios tend to trade for less per sft than 1brs, and 1brs tend to trade for less per sft than 2brs, etc.. -- in short, larger units do not have as much supply as studios and 1brs and the demand that chases them values them as such.
Moving on. Unless you have an unchanged same unit trade to analyze you will have to make a few adjustments; lets assume you don't have a same unit comp. Every single analysis between two different apartments will be exposed to the following variables:
1) Time - how is the market today and how has the market changed since the contract was signed for the selected COMPARABLE-A? I always use a 3% +/- range when adjusting for time as this is an imperfect science. The key is not to be scared to play around with an older, yet highly relevant sale since we now have a way to track Manhattan real estate price action.
The Streeteasy Repeat Sales Condo INDEX is a wonderful tool to use when trying to figure out how to adjust for time! There are very good reasons why Streeteasy ONLY used:
a) condo sales
b) repeat sales only
As Streeteasy states in the Condo Index Methodology: "A repeat sales transaction-based index allows for an "apples to apples" approach and is more like a stock market index as it tracks price changes of the same properties (or in the case of the stock market, the same stock) over time. Since this approach compares literally the same properties, errors or biases created by variables like location, size, age, and quality, are minimized."
YES! This is 100% consistent with the comps analysis approach I am trying to discuss today. I would rather analyze a same line comp that is 2 years old than go outside the building and expose the analysis to multiple variables by using a non-relevant sale that occurred a month ago.
2) Renovations - how has the COMPARABLE-A been renovated compared to the target apartment we are trying to come up with a fair market value range for? This is a imperfect science and the best we can do is utilize all information/photos we have available and estimate what it might cost to put the target apartment in the same condition.
I have seen classic 6's get a $500,000 renovation and a $250,000 renovation and I couldn't tell the difference with the cheaper work; both apartments looked amazingly renovated with high quality materials and labor.
Similar to valuing apartments with very large terraces, there is a law of diminishing returns when an apartment sees a very expensive renovation relative to its total size. If you have two 1,700sft classic 6's, one does a $500,000 reno while the other gets a $250,000 reno, the apartment with the cheaper work will likely recoup a higher percentage of their 'all-in cost' than the higher one. There is no chart on this, its what I find from working in this market and seeing how buyers bid up for renovations. Chances are there are elements of the $500,000 renovation that don't match what the new buyer will pay a premium for; the law of diminishing returns. Of course this changes as the size of the apartment increases.
3) Floor Difference - using a floor multiplier, how much higher does a 15th floor unit trade over say a same line unit on the 4th floor? The key here is balancing a) the gap between floors and, b) drastic difference in views to come up with the right multiplier. Usually between $10,000 - $25,000 a floor is used for this adjustment. I would read the link provided above for a more detailed look at how to value higher floor units versus a lower floor unit.
4) Size Difference - if there is a size difference but you deem the COMP-A close enough to use for your analysis, then use the PPSF of the sold unit and adjust for size to get it on par with the target apartment.
TARGET-Apt 2BR/2BTH: 1,150sft asking $1,300,000
COMP-A 2BR/2BTH: 1,240sft sold for $1,325,000 or $1,069/sft
90sft difference X $1,069 = negative $96,210 adjustment for size
5) Other Adjustments - think special features like terraces or fireplaces. This gets a bit trickier depending on the quality and utility of the special feature.
Now think about doing the above five adjustments & adjusting further because you decide to use a different building, with different amenities, for a unit with a different layout, exposure and views! That is crazy! Yet I heard it all before. The broker likes to use the condo comp for a coop analysis because the condo comp sold only 3 weeks ago and is the most recent reflection of today's market. Ok, except that is really not true. Yes, the sale may have been recorded a few weeks ago but its highly likely the deal was signed 2-3+ months ago. Even these should go through a minor time adjustment. Why expose a comps analysis to 10+ variables that can't be adjusted for when you can keep all other variables mostly constant and simply adjust for time using SE's repeat sales condo index?
Its quite important that I conclude this discussion on the following note. Nobody has any right to deny a seller the right to "test the market" and wait for their price! Doing a comps analysis starts out with science and real data and ends with art and adjustments. The more you do using this format, the better you will get at it over time. The goal is always to minimize the number of adjustments and to stay in the same building as the target apartment - preferably the same line, same configuration, and as close in proximity to the floor that the target unit is on.
But sometimes, what the comps tell you and what the seller's bottom line is are two different things. Frustrating, yes. But the seller has every right to test the market for themselves, especially if they went above and beyond improving the unit while they owned it. In the end, time will tell if the market can produce a bid high enough for the seller to accept. In the end, the market (the bid) dictates value, not the broker or the seller.
Finally, brokers are not appraisers! Yet our job is that much more difficult because we do not have an executed contract of sale with a contract price on it when asked "how much is this apartment worth in today's market?". You get three appraisers into a unit with no contract price and I guarantee that you'll get back 3 different values on what the place is worth! Knowing where the bid comes in is everything! Since we don't know, it's best to use the above approach discussed and learn with time how to apply each adjustment for time, floor, renovation, size, and special features. Data availability is your worst enemy here; specifically, those buildings with so few sales that you are forced to analyze outside buildings as comparables.