Orlando MSA - Urban Growth Forecasting & Market Overview

Key Context

• Orlando submarket population is rapidly growing, with a 3.03% CAGR since 2020 and forecasted to exceed 1,200,000 residents by 2025, while median household income has increased 7.74% annually since 2020.

• Brightline aims to raise $400M for its Orlando–Tampa rail extension; Universal files permits for “Project 915” expansion at Epic Universe; Camping World Stadium’s $400M renovation moves through entitlement; and speculation continues about The Boring Company’s possible I‑Drive tunnel.

• Epum’s Urban Growth Forecasting model uses machine learning to analyze over 300 spatial, regulatory, and economic variables at the parcel level, flagging areas with historical development patterns to identify likely residential growth zones in the Orlando submarket.

Submarket Overview

Population & Median Household Income

Largest Local Employers (Private Sector)

Recent Orlando MSA Announcements

• Brightline moves to secure $400M in private activity bonds for its Orlando–Tampa extension, a major infrastructure play that could reshape regional land use patterns. The alignment will run from Orlando International Airport into downtown Tampa, signaling future transit-oriented development opportunities along I‑4.

• Speculation surrounds The Boring Company eyeing a tunnel beneath Orlando’s I‑Drive corridor, similar in concept to Las Vegas’s “Loop” system. If advanced, this project would connect major hubs, including the Orange County Convention Center and the Universal parks.

• Camping World Stadium’s proposed $400M renovation enters the entitlement pipeline, including expanded capacity and new premium spaces. The city is also actively seeking new naming rights partners, positioning this as a catalyst for reinvestment in the West Lakes Opportunity Zone and surrounding districts.

• Universal files new permits for “Project 915” at Epic Universe, just months after opening day. The filings suggest vertical construction—likely retail, dining, or entertainment—within the 750-acre development footprint, signaling ongoing phased entitlements and long-range expansion planning.

An Intro to Urban Growth Forecasting (UGF)

Last year, Epum build a machine learning model to forecast the likelihood of any parcel being developed into a residential use within the next 3 years (2024 - 2027) in the Orlando MSA. This model focused primarily on raw land parcels but evaluated developed parcels as well.

How UGF works?

Epum’s Urban Growth Forecasting model is built to predict where residential demand and development are likely to take place with a (non-exclusive) focus on raw land parcels. The model is trained on decades of historical development data to find patterns where housing got built and under what conditions. From a high-level, it uses a two-step process:

1. The machine learning model isolates the geographic and regulatory features that have historically made parcels more likely to be developed. These include fundamentals like proximity to transportation infrastructure, school data, favorable zoning overlays, adjacent development activity, topography, parcel size, and more.

2. The machine learning model scans every parcel in the current market (the Orlando MSA) for those same feature combinations, flagging areas that align with past high-growth conditions. Epum’s early version of this model (UGF-1) utilized satellite imagery and over 300 spatial, regulatory, and economic variables designed as “features” for the machine learning model.

Historical Performance & Conclusion

Epum’s first Urban Growth Forecasting model (UGF-1) was tested on pre-2024 historic data to simulate real future predictions. For housing, where only 3.07% of parcels were developed during the test period, traditional accuracy metrics aren’t meaningful. Instead, we measure “top-k% accuracy”: how effectively the model identifies the parcels that are the most likely to be developed.

In the top 1% of predicted parcels, 80.9% actually saw housing development, rising to 93.2% accuracy at the top 0.5%. For the general development of raw land parcels to any sort of real estate use, model accuracy reaches 84.1% and 93.5% at the top 1% and 0.5% thresholds, respectively.

In conclusion, Urban Growth Forecasting is not only feasible but also effective and ripe for utilization within the real estate industry. Epum’s first model (UGF-1) showed promise and our current iteration (UGF-2) is much more accurate and expansive. Urban Growth Forecasting as a field of science is niche with few researchers focused on advancing the state of the art, but it has the potential to change how real estate investment and municipal governance decisions are made around the world. The Epum R&D team will continue to advance the state of the art and publish its progress when possible. Please feel free to reach out to Epum’s Chief R&D Officer, Marvin Mc Cutchan (https://www.linkedin.com/in/marvin-mc-cutchan-78a216129/), to discuss any collaboration opportunities within industry or academia!

(Bonus Content) Recently Approved Housing Projects in the Orlando MSA

• 154 unit Mixed Use Development on ~10 acres at 5449 S. Semoran Blvd., Orlando, FL

  • Approved/Decision Date: July 15th, 2025

  • Type: Site Plan Approval

  • Status: Approved

• 1000 unit Mixed Use Development on ~11 acres at 520 Callahan Dr., Orlando, FL

  • Approved/Decision Date: July 15th, 2025

  • Type: Rezoning

  • Status: Approved with Conditions

• 168 unit Single-family Housing Development on ~21 acres at 2400 South Goldenrod Road, Orange County, FL

  • Approved/Decision Date: April 24th, 2025

  • Type: Rezoning

  • Status: Approved with Conditions