AI in Real Estate: Understanding Local Markets

Using AI in real estate to understand local markets

Hi there, Mo here šŸ‘‹

I hope you enjoyed last week’s edition of AI in Real Estate. I got a few emails for me to go more indepth so here we are.

The tool I’m using is Claude.ai – I like it because it allows you to search the web and can upload PDF’s.

I asked Claude, ā€œwhat can you do with AI in CREā€ and this is what it spit out…

Here are a few ways AI can influence commercial real estate:

- Market analysis - AI can analyze large amounts of data on demographics, economic trends, transportation patterns, etc. to help identify promising locations for new commercial developments. It can also track supply and demand dynamics in different markets.

- Property valuations - AI models can estimate property values more quickly and accurately than traditional methods. This helps set listing prices, assess property taxes, and evaluate investment opportunities.

- Predictive analytics - By analyzing data on past performance, AI can generate forecasts for occupancy rates, rental income, operating expenses, etc. This helps inform investment and management strategies.

- Automated deal analysis - AI can rapidly process leases, sales contracts, loan documents etc. to extract key terms and evaluate financial impacts. This streamlines due diligence.

- Smart building technologies - AI optimizes HVAC, lighting, security systems in commercial buildings to improve efficiency and reduce costs. It can also analyze usage patterns to better understand tenant needs.

- Chatbots & virtual assistants - AI can automate customer service and leasing interactions to be 24/7 responsive. It can also provide tenants/buyers with personalized recommendations.

- Risk assessment - AI can evaluate potential risks across portfolios spanning credit, operations, regulatory compliance etc. This enables preventative risk management.

In summary, AI is transforming commercial real estate by automating tasks, generating strategic insights from data, and improving efficiency across the board. Its applications will likely continue expanding in the future.

Today, we’re going to be talking about market analysis. I do this with Costar reports, broker reports, and even free BLS data sources to comprehend what the status quo of a market is.

The first step: arranging your data

With AI, it’s important to have great quality data. Some popular data vendors in the CRE space are…

Some solutions (like Compstak and Costar) have property financial data for lease + sales comps, ownership information, and some loan data. Some of them are specific to lease and sales comps only, some are focused on debt information, some are focused on buying.

The CRE data landscape is quite vast and sometimes can be contradictory. Always try to reach out to brokers as well – brokers are ā€œboots on the groundā€ and will have direct data for lease and sales comps in a particular submarket.

Lots of the brokerage houses will have their own in-house databases.

Second step: Upload the data into AI

I’m using Claude because it can upload PDF’s. This is very important as you may need login creds and can export data to analyze.

In this exercise, I’m uploading Costar quarterly market reports, market analytics, and an underwriting report.

The third step: Understand your objectives

I am looking at a retail property right now and want to understand where the market is in terms of comps.

Understand if you want a market analysis, property analysis, crime analysis, job outlook, transit outlook, supply/demand analysis.

The data is all out there, but ultimately it comes down to ā€œwhat questions are we trying to answerā€.

With the right prompting comes the right answers.

My prompt:

I want to understand what the retail market looks like in Phoenix. What is the current outlook of the market, what do comps look like.

What else would you add to my question?

Claude:

Here is an overview of the current retail market conditions in Phoenix:

Market Overview

- The Phoenix retail market encompasses over 147 million square feet of space across 6,800+ properties.

- Vacancy rate is currently 3.8%, down from 4.2% last year. This indicates a tightening market with limited availability.

- Over the past 12 months, 132,000 SF has been absorbed while only 48,000 SF delivered to the market. Positive absorption shows healthy tenant demand.

- Asking rents average $29.22 per square foot, up 7.9% year-over-year. Rents have been growing steadily in recent years.

- Neighborhood and community centers make up the largest share of retail inventory in Phoenix. These cater to everyday shopping and dining needs.

Competitive Properties

- 9 comparable neighborhood centers in N. Scottsdale show asking rents from $14 to $28 per SF, with availability rates around 10.5% on average.

- Recent leasing deals at top properties range from 1,750 SF to 10,800 SF over the past year. Typical months on market is 85.

- Sales of similar retail assets averaged $304 per SF over the past 12 months. Cap rates averaged 6.2%, indicative of investor demand.

To further analyze the market, I would look at consumer spending patterns, new housing growth, and traffic counts around prospective sites. We could also break down market conditions by specific submarkets and retail categories to hone in on opportunities. Let me know if you need any additional details!

Notes

I like to add ā€œwhat else would you suggestā€. Although human logic should guide the machine, the machine can source new ideas on things I may be overlooking.

From the data, it shows 9 comparable properties with a range in comp set. I would detail the compset down from Costar into the different markets and property types.

There’s many different ways to slice and dice the data but you want to make sure it’s good quality and comparing ā€œapples to applesā€.

More Prompting

The beauty with AI is that you should prompt it more to get the details you want. AI can suggest more things and you can use it to extract more information.

For this exercise in particular, I used wanted to understand consumer spending patterns, new housing growth, and specific market conditions by submarkets.

Claude’s response

Absolutely, here are some additional details on those aspects of the Phoenix retail market:

Consumer Spending

- Compared to the national average, Phoenix residents spend more on groceries, health care, entertainment, and insurance.

- Spending is lower than average for apparel, personal care, education, and pensions/social security.

- On a per household basis, top spending categories in Phoenix are food at home ($8,418), owner-occupied housing ($20,040), health care ($6,718), and gas/vehicles ($10,624).

Housing Growth

- Maricopa County saw over 24,000 single family permits issued in 2022, up 2% from 2021. Strong housing expansion continues.

- Top growth areas include Buckeye (+43% permits in 2022), Goodyear (+36%), Casa Grande (+35%), and Queen Creek (+28%).

- Housing drives retail demand, suggesting opportunities may exist in these rapidly growing suburbs.

Submarket Performance

- The East Valley submarket has the lowest retail vacancy at 2.5%. Average asking rent is $27.55 per SF.

- The Southwest Phoenix submarket has struggled more with 4.3% vacancy and only 1.3% rent growth last year.

- Grocery-anchored centers average just 2.1% vacancy across the metro currently.

- Shopping centers are tight at 1.3% vacancy with rents up 9.4% annually.

I can provide any additional details on consumer profiles, trade areas, sales volumes, etc. for certain retail categories or locations as needed. Please let me know if you would like me to expand on any specific part of the market overview.

My response

Is there demand for retail within the Phoenix metro? SHould I buy something with $13-15 PSF in-place rents?

I have a specific property in mind and want to get a bird’s eye glance at what the market is without having to calculate a bunch of things. Claude seems like a great assistant for it…

Claude

There does appear to be healthy demand for retail in the Phoenix metro area currently, based on the following factors:

- The metro's population has grown over 9% since 2017, expanding the consumer base for retailers. Job growth is also strong, fueling retail spending.

- Vacancy rates across most retail property types are in the 2-4% range, indicating limited available space options for tenants. This suggests pent up retailer demand.

- Net absorption has been positive, led by neighborhood centers absorbing over 55,000 SF in the past year. Openings are being leased quickly.

- Asking rental rates are up nearly 8% annually on average, as tight conditions allow landlords to push rents. Properties with below-market rents have upside.

- Investors are showing confidence in Phoenix retail. Cap rates have compressed to the low 6% range due to competition for assets.

Regarding the $13-15 PSF in-place rent range, this does look to be below current market rates for most areas and retail property types in Phoenix. Well-located neighborhood centers for instance are averaging over $29 PSF asking rent.

That said, there could still be value in acquiring assets with upside potential from bringing rents to market over time. The key considerations would be location, visibility, tenant rollover timing, and options to add value through renovations or repositioning. But the market fundamentals appear generally supportive of Phoenix retail currently.

Let me know if you need any other specific data points to assess potential retail investments in the metro area. I'm happy to provide additional details on sales comps, cap rates, consumer profiles, etc. for certain submarkets or retail categories to help with your analysis.

Final Thoughts

Hope you enjoyed this edition of the newsletter. I’m constantly experimenting with AI and think it would make a good little segment of the newsletter. Please let me know if this is helpful to you!

Also if you like Phoenix retail, respond to this email and let’s set up a time to chat. I may have opportunities in the future.

PS: Please fwd this email to anyone who may find it useful. It helps a lot!

Thanks,

Mo