How one CRE giant is finding undervalued assets in unlikely areas

Global real estate investment manager BGO ($89B AUM) has materially improved its investment performance by implementing a proprietary AI-driven data science model. This model, developed through backtesting two decades of internal deal data, revealed that local market fundamentals, rather than broad property pricing or national economic trends, are the primary determinants of real estate outperformance. BGO cites a Las Vegas industrial development where their model accurately predicted significantly higher rents ($9/sqft vs. $5.88/sqft underwritten) than conventional research, underscoring the competitive advantage gained through advanced analytics in identifying undervalued opportunities and optimizing investment decisions.

Analysis

BGO, a global real estate investment manager with $89 billion in AUM, is achieving material outperformance by deploying a proprietary AI-enhanced data science model. This model, developed by backtesting two decades of the firm's own deal history, has identified hyper-local market fundamentals as the singular determinant of investment success, effectively displacing the influence of national economic trends and property pricing in their strategy. A prime example of the model's efficacy is an industrial development in Las Vegas, where it predicted explosive growth, leading BGO to underwrite rents at $5.88 per square foot. The investment subsequently achieved rents in the $9 per-square-foot range, a significant outperformance that conventional research models had failed to anticipate, labeling the deal as mediocre. The model's success stemmed from identifying a key logistical arbitrage: companies could save approximately 60% on total costs by relocating from the expensive Inland Empire to Las Vegas, despite a longer drive, due to lower rents, taxes, and labor. While this data-driven approach has yielded strong returns in other regions like Florida and the Rust Belt, the firm's co-CEO acknowledges its limitations, noting it cannot predict idiosyncratic events such as a major employer abruptly leaving a city.

Key Decisions for Investors

• Investors should prioritize allocating capital to real estate managers who can demonstrate a deeply integrated and proprietary data science capability, as this appears to be a significant emerging source of alpha generation, capable of producing non-consensus investment theses.
• The success of BGO's Las Vegas investment highlights the potential in secondary markets that serve as cost-effective alternatives to primary hubs; investors could screen for similar opportunities by analyzing total cost arbitrage, including logistics, labor, and taxes, rather than just rent comparables.
• While the AI-driven approach shows significant upside, the acknowledged model risk from unpredictable events reinforces the need for diversification and suggests that quantitative strategies should be complemented with qualitative risk assessment.