The Importance of Underwriting. Is Automation Effective?

[Jack Pasmore]

Hey BP Community!

I’m diving into the multifamily-apartment investment space, and one thing has become crystal clear: if you can’t underwrite a deal, you’re flying blind. Every great multifamily investor I’ve spoken with emphasizes this skill, and I see why.

On the other hand, underwriting can be time-consuming and prone to human error. I have been underwriting my own deals and wonder Why am I spending hours crunching numbers manually when technology can streamline the process? I’ve been exploring tools like Google Sheet calculators that are coded for accuracy and speed, allowing individuals to focus more on their strategies, acquisitions, and capital raising.

So here’s my question to you:

• How did you master underwriting multifamily deals?
• Do you believe that automation the way forward for scaling and efficiency?

I’d love to hear your insights. And if you’ve built or used a killer underwriting tool, feel free to share, I’m always looking to improve and innovate!

[Greg Scott]

My initial reaction was "hell no".

Algorithms are usually poor estimators of actual value.  Anyone that has looked at Zillow's Zestimate can tell you that an algorithm can give you a general idea, but it is just aggregating past data and spitting out a number. 

Take a house that sold at a low price twice in the past.  The algorithm has no idea if the problem as a dysfunctional layout or just the fact that it is dated and nobody has spent the money to update the property.  One problem may not be feasible to fix, and the other is easy to fix.

Does the apartment complex have a unique feature that makes it more desirable?  I've seen properties on a lake or near a park or popular trail that will consistently get higher rent.  Meanwhile, the adjacent property with a view of an industrial park will consistently get lower rents.  An algorithm will probably tell you that undesirable property is an amazing "value add" opportunity.  But in that case, you can't fix what is broken.

I've not even seen a great source for knowing exactly what the competition charges.  For example, an algorithm can't compare a property that charges a flat fee for utilities vs one that charges based on use.  We have one property that charges a flat fee.  The property closest to ours charges based on use.  However, they didn't insulate that building well so in winter residents consistently get heating bills over $200/mo.  Many of the residents there eventually move to our property. You find this out by talking to people, speaking with past residents, people in the industry, or mystery shopping the property.

Using technology to pull in data so you don't have to manually type it in, is helpful, but other than that, I would not trust automation.  So, on second blush, I would change my answer to "definitely, hell no".