How AI is Quietly Changing Multifamily Financial Modeling
Over the past year, most conversations about artificial intelligence in real estate have focused on marketing automation, leasing tools, or property management platforms.
However, there is a quieter transformation beginning to happen in another area of the industry:
Financial Modeling & Investment Analysis.
For multifamily investors navigating a complex refinancing environment, the ability to evaluate financial scenarios quickly is becoming increasingly important. Higher interest rates, tighter lending standards, and shifting capital markets mean analysts are often asked to evaluate dozens of potential outcomes before making a portfolio decision.
AI tools are beginning to help accelerate that process by improving how analysts interact with them, rather than simply replacing their financial models.
The common theme that seems to be shaping up across all industries is that AI should be viewed as a co-pilot/assistant and not a replacement (at least at this stage in the game - we'll worry later about what AI will be able to do autonomously 20 years from now). So the good news for now is that human analysts are still very necessary in the industry, so let's now look at how we can best interact with our AI tools in an already turbulent time for commercial real estate.
1. Model Auditing and Error Detection
Even experienced analysts occasionally encounter modeling issues such as:
- Broken cell references
- Inconsistent assumptions across worksheets
- Circular references
- Sensitivity tables that fail to update correctly
Traditionally, identifying these issues requires time-consuming manual review. However, AI copilots can assist by reviewing formulas and flagging potential inconsistencies much more rapidly.
For example, analysts can paste formula logic or sections of model structure into tools such as:
- ChatGPT
- Claude
- Manus
- Excel Copilot
These tools can help verify formula logic, assumption consistency, and calculation flow between worksheets. By no means does this replace model review, but it can function as a second set of eyes, particularly in complex underwriting models. (For those of us who know the joy of being up late at night trying to fix a model before a presentation the next day, this is welcome assistance!)
2. Rapid Scenario Generation
In the current refinancing environment, multifamily analysts often need to evaluate multiple capital structure scenarios quickly.
A typical refinance analysis may require testing:
- Different interest rate assumptions
- Varying loan-to-value levels
- DSCR constraints
- Partial paydowns
- Loan extensions
Traditionally, analysts have had to adjust inputs manually across multiple worksheets.
AI-assisted workflows can accelerate this process by helping analysts generate scenario frameworks, sensitivity structures, and comparative outputs.
For example, an analyst can prompt an AI tool to help structure a scenario table such as:
| Interest Rate | LTV | DSCR | Loan Proceeds |
|---|
This allows teams to evaluate a broader range of refinance outcomes more efficiently. We still need to understand how to go through these refi scenarios and recognize any mistakes the AI could have made (often subtle, so it still pays to look closely at its reasoning and data sourcing), but nonetheless, this process is much more efficient this way.
3. Portfolio-Level Risk Analysis
Many investment firms still evaluate refinancing decisions asset by asset, but in practice, refinancing risk often exists at the portfolio level. For example:
If multiple loans mature within a two-year window, capital allocation decisions across the portfolio become interconnected.
AI-assisted workflows can help organize and interpret:
- Multiple asset refinance scenarios
- Capital requirements across the portfolio
- Timing of loan maturities
- Recapitalization options
Instead of analyzing assets independently, asset managers can begin identifying portfolio-wide capital constraints earlier.
4. Investment Memo and IC Preparation
Another area where AI tools are quietly improving workflows is investment documentation. (Oh yeah - if you know, you know - this was always one of those blood pressure-spikingly fun tasks to work on, and usually with very limited time.)
Preparing investment committee memoranda often requires translating complex financial models into clear investment narratives. AI tools can assist analysts by helping to summarize modeling outputs, structure investment committee memos, and identify key risk variables.
Rather than replacing analytical judgment, this instead helps analysts produce clearer investment narratives more efficiently. (Also, spending less time on very early morning Zoom meetings with the managing director making last minute corrections, questioning our life decisions).
5. The Real Limiting Factor: Data Quality
Despite the promise of AI-assisted modeling workflows, one reality remains unchanged:
Real Estate Data Quality Still Determines Analytical Accuracy.
Even the most advanced tools depend on reliable inputs such as:
- Rent rolls
- Operating expenses
- Capital expenditure forecasts
- Debt terms
- Market data from large commercial databases which can still be missing cap rates, accurate sale prices, and so much more
For many firms, improving internal data organization may provide greater benefits than any individual technology tool. This can be tedious but very worthwhile, as someone who has done this several times (once during a large merger between two CRE firms - ugh!). Its best to create at least an annual (if not quarterly) audit-and-update process to review and add or correct data at least going back a few years (then the older data won't normally be changing, so the more you do this, the easier it gets). With your databases squeaky clean and reliable, you'll get far more out of using AI tools with said databases.
The Bigger Shift
The role of the real estate analyst is evolving. Financial models (and analysts) remain central to investment decision-making, but the tools surrounding them are becoming more powerful.
Analysts who combine traditional modeling skills with modern analytical workflows may gain a significant advantage, particularly in a market environment where refinancing strategy, capital allocation, and operational performance are tightly connected (such as the Bay Area).
AI will not replace real estate analysts, but it may significantly enhance the speed, clarity, and scale at which investment decisions can be evaluated.
As the next phase of the multifamily cycle unfolds, the most effective investment teams may not simply have the best models, but the best analytical workflows.
In a future post, I plan to explore how analysts can build AI-assisted refinance stress-testing models to evaluate loan maturity risk across entire portfolios. Stay tuned for that!
Useful link: check out Adventures in CRE here (non-affiliated; just a fantastic source for financial modeling ideas and help)
Disclaimer
The content on this blog is provided for informational and educational purposes only. While TSZ Enterprises makes every effort to ensure accuracy and usefulness, the material is not tailored to your unique circumstances and does not constitute professional, legal, medical, financial, or tax advice.
Information on this site does not create a professional-client relationship between you and TSZ Enterprises. If you require personalized guidance, please seek the services of a qualified professional in the relevant field.
All use of the blog’s information is entirely at your own risk. TSZ Enterprises expressly disclaims any liability for any damages or losses resulting from your reliance on the content provided here or on third-party links. While we aim to keep content current, we make no guarantee of completeness or accuracy.
Comments
Post a Comment