Automated Valuation Models
One of the most crucial fiduciary duties for a real estate agent is to inform their client of a proper value for a home they are considering for purchase.
It helps craft a more compelling offer, and ensures a client knows at what point he or she may be overpaying.
Assessing a home’s value is a little bit “art”, a little bit “science”.
It involves an intimate knowledge of local inventory, the rhythm of the local market, comparative pricing among different properties, the state of lending and more.
Hence, choosing a local expert active in the local market is key.
That leads us to discuss automated valuation models, or AVMs.
AVMs are popping up many places on the internet, most notably on popular listing aggregators.
For a specific property, AVMs output a “home value” or “suggested asking price” based on numerical and geographical factors, like location, square footage, bed and bath count, and historical comparables in the area.
What AVMs do not take into account are interior condition, differences in “flow” or room layout, light, view, and other more idiosyncratic factors. In NYC, these factors grow when introducing co-op policy differences, making AVM values even less applicable.
While AVMs may get a buyer in the ballpark on a price estimate, a broker’s estimate based on detailed local expertise could foot out 10-20% higher or lower. This can amount to tens or hundreds of thousands of dollars.
As with other elements in real estate, technology must be married to expertise in order for it to be actionable for your offer.