EP 396 Is This 128-Unit Apartment Worth Buying?

APPLE PODCASTS | SPOTIFY

How we use an AI Investment Committee to review multifamily deals faster, uncover questionable assumptions, and decide which opportunities deserve a closer look.

Can I tell you something that might surprise you about real estate investing?

Sometimes, the most valuable thing you can do when a new apartment deal lands in your inbox is decide not to spend another week analyzing it.

I know. Not exactly the glamorous side of building a real estate portfolio.

But hear me out.

When you’re buying multifamily properties, there is no shortage of information to review.

There’s the offering memorandum. The rent roll. The trailing twelve months of financial statements. Renovation budgets. Property taxes. Insurance. Financing assumptions.

And that’s before you’ve even walked the property.

My sister Nancy and I have spent years building our real estate investment business, and one thing we’ve learned is that the challenge isn’t always finding another opportunity.

It’s having enough time to figure out which opportunities are actually worth pursuing.

So we started asking ourselves a different question.

What if we could review more deals without having to build a much bigger investment team?

That’s one of the reasons we built AIREI, our AI Real Estate Investing Operating System.

And recently, I sat down to walk through a real multifamily offering in Arizona, using our AI-powered underwriting process.

The property had 128 apartments, was built in 1972, and came with four important documents.

What happened next is a perfect example of why AI can be such a powerful tool for real estate investors—and why you still need to know what you’re looking at.

Let’s get into it.

First, Meet the 128-Unit Apartment Building in Tucson

The property we reviewed was an older, Class C multifamily community in Arizona.

Built in 1972.

128 apartments.

And a history of capital improvements that included HVAC replacements, fencing, flooring, appliances, and other property upgrades.

Now, here’s something important about buying older apartment buildings.

An older property isn’t automatically a bad investment.

In fact, depending on the purchase price, there can be opportunities that make sense.

But the older the building, the more attention we pay to potential deferred maintenance.

Because the price you pay to acquire an apartment building is only part of the equation.

You also need to understand what it will cost to own, maintain, improve, and eventually sell it.

And that brings us to the first step.

Step 1: Upload the Documents Before You Start Making Assumptions

For this walkthrough, we uploaded four files into AIREI:

• The offering memorandum (OM)
• The trailing twelve months of operating statements (T-12)
• The current rent roll
• The seller’s capital expenditure history

If you’re newer to multifamily investing, think of these documents as four different pieces of the same puzzle.

The offering memorandum tells you how the seller is presenting the opportunity.

The T-12 tells you what the property has actually earned and spent over the past year.

The rent roll tells you about the apartments, tenants, and rents.

And the capital expenditure history helps you understand which improvements the seller has already made.

Here’s the challenge.

Not every seller organizes their financial information the same way.

One property might categorize an expense as repairs and maintenance. Another might list it under unit turns. A third might use a completely different name.

And when you’re manually entering everything into spreadsheets, those differences take time to reconcile.

That’s one reason we’ve continued refining AIREI’s document-reading process.

At the time of this demonstration, we’d gone through approximately 22 versions of our parsing tool.

Why so many?

Because real estate financial statements aren’t standardized.

And if the system can’t interpret a document correctly, we need to know.

During this demonstration, AIREI scanned the four files in roughly two minutes.

My complete first-pass process typically takes me about 10 to 15 minutes, depending on the complexity of the deal and how much I need to investigate.

But here’s the important part.

Reading the documents faster doesn’t mean you should trust every number automatically.

We still check that the information ties back to the source documents.

Because faster underwriting is only valuable when the analysis is reliable.

Step 2: The Seller’s Expenses Looked a Little Too Good

This is where things started getting interesting.

As AIREI organized the financial statements, one number caught my attention.

The property’s reported operating expenses represented approximately 31% of gross potential rental income.

And my reaction?

That seems low.

Really low.

For this particular older property, I wasn’t comfortable assuming that expense level would continue without further investigation.

Now, that doesn’t necessarily mean the seller did anything wrong.

But it does mean we need to understand how expenses were categorized.

For example, the seller’s capital expenditure history included items such as flooring replacements, appliances, and other improvements.

Some expenditures may properly qualify as capital improvements.

Others may be ordinary operating expenses, depending on their nature and the applicable accounting treatment.

Why does that distinction matter?

Because moving an expense below the operating-income line can make the property’s net operating income, or NOI, look higher.

And NOI plays a major role in how commercial real estate is valued.

So when an expense ratio looks unusually low, I want to know why.

Are some expenses missing?

Are recurring costs being categorized as capital improvements?

Has the seller recently completed major repairs that won’t need to be repeated?

Or is the property genuinely operating more efficiently than comparable buildings?

These are the questions that matter.

A beautiful-looking NOI isn’t enough. You need to understand what went into it.

Step 3: We Started With a $10 Million Purchase Price—But It Was Just a Placeholder

Here’s something I want every investor to understand about first-pass underwriting.

You don’t always need every answer before you begin.

For this walkthrough, I didn’t have a confirmed asking price or whisper price available.

So I entered $10 million as a starting assumption.

Not because that was the property’s verified value.

Not because I believed it was the right offer.

Simply because we needed a number to begin testing the investment.

Then we started building out the rest of the assumptions.

For renovations, I used an initial estimate of $5,000 per apartment.

Across 128 units, that gives us a preliminary renovation allowance of approximately $640,000 before further adjustments.

And again, that’s not a contractor’s estimate.

It’s a starting point based on experience.

Once we tour the property, review the seller’s renovation history, and obtain bids, we can refine the numbers.

Because there’s a big difference between what you think a building needs from behind your computer and what you discover when you’re standing in front of it.

Step 4: Why a Property Tour Can Change Your Entire Underwriting

Can we talk about the things that don’t always show up in a spreadsheet?

The roof.

The parking lot.

The drainage.

The plumbing.

The condition of the exterior.

These are the things I pay attention to when I first walk a property.

And no, a property tour isn’t a replacement for a professional inspection.

But it can tell you whether the deal deserves further investigation.

For example, replacing an entire roof can cost hundreds of thousands of dollars.

Major parking lot repairs can be expensive.

Drainage or foundation problems can completely change your renovation budget.

Even something as ordinary as tree trimming can add up quickly across a large apartment community.

And then there are the improvements the seller has already completed.

If the seller recently replaced HVAC systems or installed fencing, that’s useful information.

But we still need to verify the scope, condition, and remaining useful life of those improvements.

The goal is to avoid paying twice for the same work—or assuming a problem has been solved when it hasn’t.

This is one of the reasons I don’t believe AI should replace physical due diligence.

AI can help organize the questions. But someone still needs to walk the property and find the answers.

Step 5: The Rent Roll Didn’t Tell the Same Story as the Offering Memorandum

Here’s where I had to stop and investigate.

When we imported the property’s unit mix, the system initially grouped some apartment types together.

But the offering memorandum distinguished between classic and renovated apartments.

The rent roll didn’t clearly make that same distinction.

And that created a problem.

Because a classic one-bedroom apartment and a renovated one-bedroom apartment might have different rents.

If you combine them, you can distort your understanding of current income and future rental potential.

So I went back to the offering memorandum.

I reviewed the apartment types.

I separated classic and renovated units.

And I compared the current rents with the seller’s projected rents.

One renovated one-bedroom category, for example, showed current rents around $1,348 with a projected rent of $1,500.

A renovated two-bedroom category showed approximately $1,653 in current rent and a projected rent of $1,800.

Could the property achieve those increases?

Maybe.

But that’s not something we can assume simply because the seller included the numbers in a brochure.

We need comparable properties, actual leasing data, and a market study to support the projections.

And this is a perfect example of why AI is a tool—not the final decision-maker.

The system organized the information.

It helped us identify where the categories didn’t line up.

But I still needed to interpret the documents and make corrections.

Sometimes the most important underwriting question isn’t whether the numbers add up. It’s whether you’re comparing the right numbers in the first place.

Step 6: We Tested the Financing, Not Just the Purchase Price

Once we had a preliminary operating picture, we moved into financing.

For this demonstration, I modeled a floating-rate bridge loan.

That meant considering the interest-rate index, the lender’s spread, and the cost of an interest-rate cap.

At the time of the walkthrough, the illustrative all-in interest rate was approximately 6.66%.

We also modeled a rate cap, loan term, extensions, and working capital reserves.

Some of those figures were placeholders rather than lender-quoted terms.

Why spend time on this?

Because a property can look attractive before financing and much less attractive afterward.

Interest expense affects cash flow.

Reserves affect the equity you need to raise.

Renovation costs affect your total investment.

And loan maturity can create refinancing risk.

I also explained why an assumable loan isn’t automatically a great opportunity.

Imagine finding a property with a very low existing interest rate.

Wonderful, right?

But what if that loan matures in three years?

Now you have another question to answer.

Will the property’s income and market conditions support refinancing when that loan comes due?

That’s why we look at the entire capital structure.

Not just the headline interest rate.

Step 7: We Changed the Purchase Price to See What the Deal Could Support

This is one of my favorite parts of using AI in the underwriting process.

Instead of spending hours rebuilding a spreadsheet every time we want to test a different purchase price, we can adjust an assumption and review the updated results.

During the demonstration, I changed the illustrative purchase price from $10 million to $6 million.

That’s a substantial difference.

And at that lower modeled price, the preliminary analysis showed approximately:

• 1.88x investor equity multiple
• 20.88% projected investor IRR

Now, those are model outputs—not actual investment returns or verified expectations.

And they depended on assumptions that still needed investigation.

But here’s where the analysis became especially useful.

The equity multiple was just below our 1.9x target.

The property was also outside our preferred investment market.

And it was built in 1972, whereas the investment mandate we were testing generally preferred properties built in 1990 or later.

So even with an attractive-looking projected IRR, the investment didn’t neatly fit our criteria.

And that matters.

Because when you’re evaluating real estate opportunities, you need to know the difference between a deal that looks interesting and a deal that fits your investment strategy.

Step 8: We Asked Our AI Investment Committee to Challenge the Deal

This is the part of the process that gets me excited.

After completing the preliminary underwriting, we convened our AI Investment Committee.

Instead of simply asking AI, “Is this a good deal?” we use a structured review to challenge the assumptions and identify what needs to be verified.

And in this example, the committee recommended passing on the opportunity under the assumptions and investment mandate we had entered.

But what I found most useful wasn’t the recommendation itself.

It was the questions behind it.

The analysis flagged several things that needed more support.

Could the property really increase gross potential rent from approximately $2.04 million to $2.2 million?

Could the operating expenses stay at the projected levels?

Was the insurance assumption reasonable?

What would a third-party market report tell us about occupancy, concessions, and comparable property sales?

And would the financing actually be available on the terms we modeled?

Those are meaningful questions.

Because they force us to examine what must happen for the investment to succeed.

And that’s where I believe AI can add real value.

Not by pretending to know the future.

But by helping us identify the assumptions that deserve our attention.

Step 9: The Most Important Number Might Not Be Occupancy

Toward the end of the walkthrough, we talked about another issue that’s especially important in multifamily investing.

Economic vacancy.

Because here’s something that can surprise newer investors.

A property can be 90% occupied and still struggle to collect its rent.

Physical occupancy tells you how many apartments are occupied.

Economic vacancy reflects lost rental income from factors such as vacancies, concessions, and unpaid rent.

Those are not the same thing.

We’ve seen situations where property managers were focused on filling apartments, but the quality of tenant screening and rent collections needed improvement.

And that’s a problem.

Because owning a building full of residents doesn’t automatically mean the building is generating the income you expected.

For the Tucson property, economic vacancy was around the mid-teens in the historical figures we reviewed.

We modeled gradual improvement rather than assuming the problem would disappear overnight.

But whether that improvement is achievable depends on the market, tenant quality, collections, property management, and competition.

That’s why we’re continuing to build more analytical checks into AIREI.

For example, understanding how many residents moved in during the most recent few months can raise useful follow-up questions.

Has occupancy improved organically?

Are concessions driving leasing?

Are recent tenants paying consistently?

You can’t answer all of those questions from a single occupancy percentage.

And that brings us back to the bigger lesson.

The Real Advantage of AI Isn’t That It Makes Decisions for You

It’s that it can help you reach the important decisions sooner.

Before we built AIREI, a large part of underwriting involved repetitive work.

Downloading documents.

Entering numbers.

Reconciling financial statements.

Building scenarios.

Preparing investment committee materials.

All of that work still matters.

But not every step requires the same amount of human attention.

By using AI to help organize documents, extract information, compare assumptions, and prepare the initial analysis, we can spend more time investigating the things that could actually change an investment decision.

In this Tucson walkthrough, the system helped us move from raw documents to a preliminary investment committee review in a fraction of the time a fully manual process might require.

But we also discovered places where the data needed clarification.

We corrected the unit mix.

We questioned operating expenses.

We used provisional financing assumptions.

And we identified information that would need to be verified before moving forward.

That’s what a useful first-pass underwriting process should do.

It shouldn’t give you false confidence.

It should give you a clearer picture of what you know, what you don’t know, and what to investigate next.

What I’d Want You to Remember Before Analyzing Your Next Multifamily Deal

If there’s one thing I hope you take away from this walkthrough, it’s this:

A deal doesn’t become a good investment because AI says the numbers work.

The numbers have to be grounded in reality.

The rent increases have to be achievable.

The expenses have to be supportable.

The financing has to make sense.

And the business plan has to be something your team can actually execute.

AI can help you test those assumptions faster.

But you still need judgment.

And sometimes, the smartest investment decision is simply to move on to the next opportunity.

Because your time, your capital, and your attention are all limited.

The goal isn’t to buy every deal that comes across your desk.

It’s to have the capacity to identify more of the right ones.

Want to See How We Use AI to Analyze Multifamily Deals?

If you’re a real estate operator or family office looking for ways to review more opportunities without growing your investment team at the same rate, we’d love to show you what we’re building.

Nancy and I are hosting a free live masterclass called How to Use AI to Buy More of the Right Multifamily Deals.

We’ll walk through how we use AI to organize deal documents, identify potential underwriting issues, challenge assumptions, and prepare questions before making an investment decision.

Because the goal isn’t simply to underwrite faster.

It’s to create the capacity to evaluate more opportunities while protecting the quality of your decisions.

Join Our Free AI × Real Estate Masterclass

AI can do more of the analytical work.

You still make the judgment.

Frequently Asked Questions

Can AI underwrite a multifamily real estate deal?

AI can assist with initial multifamily underwriting by extracting financial data, comparing operating statements, analyzing rent rolls, and testing assumptions. Important calculations and findings still need to be verified against source documents.

How long does it take to analyze a multifamily deal with AI?

In this walkthrough, AIREI scanned four property documents in roughly two minutes. The broader first-pass underwriting process typically takes approximately 10 to 15 minutes for Palmy, depending on the deal and the amount of review required. Full due diligence takes considerably longer.

What documents do you need to analyze an apartment building?

Common starting documents include an offering memorandum, trailing twelve months of financial statements, and a current rent roll. Capital expenditure histories, financing information, and additional property records can help refine the analysis.

What is an AI Investment Committee?

An AI Investment Committee is a structured analytical process that reviews a potential investment from different perspectives, such as underwriting, risk, financing, and investment criteria. It helps organize findings and challenge assumptions but does not replace human decision-making.

What is the difference between physical occupancy and economic vacancy?

Physical occupancy measures how many units are occupied. Economic vacancy measures rental income lost through vacancies, concessions, unpaid rent, and related factors. A property can have high physical occupancy while experiencing significant economic vacancy.

Can AI replace a multifamily real estate analyst?

AI can accelerate repetitive analytical tasks, but experienced professionals remain important for verifying information, interpreting market conditions, evaluating risks, and making investment decisions.