Evolving for real estate 3.0

As real estate enters a more demanding, fundamentals-driven era, performance is no longer a function of rising tides, but of skill, discipline and execution. Investors must rethink the use of data and organizational design to succeed in the era of ’real estate 3.0,’ explains Larissa Belova, Head of the US Real Estate and Chief Investment Officer, Real Estate.

What is real estate and how is that expected to change over the next cycle?

Larissa Belova: The prior real estate cycle was unusually forgiving. Investors benefited from a powerful combination of historically low interest rates and sustained cap rate compression. In many cases, performance was driven less by active decision making and more by exposure to the right sectors at the right time. Take industrial. The extraordinary demand created by e-commerce meant that simply owning warehouse assets – often regardless of location or building – generated strong returns. Cap rate compression alone accounted for a meaningful portion of total performance during that period.

Looking ahead, that environment is behind us. Investors will not be rescued by falling interest rates or cap rate compression. We believe the next decade will be defined by fundamentals: asset selection, underwriting discipline and the ability to drive sustainable cashflow growth. In that sense, the market should become far more discriminating, and potentially more rewarding for genuinely skilled property investors.

What does that mean for how value is created within real estate portfolios?

Larissa Belova: Value creation is becoming much more operational and granular. Coming back to industrial as an example, the sector has matured dramatically, especially relating to the modern physical building requirements. It is no longer enough to buy a ‘good’ warehouse in a strong location. You have to ask different questions: is the building designed for the next generation of tenant needs? Can it support automation, robotics and the additional power required to run those systems? These factors increasingly separate assets that have pricing power and can drive net operating income growth from those that will fall behind.

At the same time, onshoring and deglobalization are changing demand drivers. Industrial is no longer just about imports moving through ports. We are seeing growth tied to domestic manufacturing, data center development and related supply chains. That is creating demand in markets such as Texas and parts of the US Midwest, alongside a continued need for infill logistics near population centers. The result is significant dispersion between markets, submarkets and individual assets. That dispersion is where alpha now lives.

What are the implications for investment organizations themselves?

Larissa Belova: To extract value in this environment, firms need to operate more like true owners and operators. That requires deeper vertical integration, real expertise at the property level and a customer-service, tenant-focused mindset. Asset management cannot be an afterthought; it has to be central to the investment process.

Large, scaled managers may have an advantage here. They can afford specialized teams, engineers and data infrastructure. They can invest in talent and systems that allow them to scrutinize every line item, actively engage tenants and manage capital expenditures efficiently.

How does technology factor into this new value-creation model?

Larissa Belova: Data and technology are no longer just productivity tools that create efficiencies. The real value comes from integrating human judgment with machine driven insight and embedding that into a repeatable, disciplined investment process.

We think in terms of a ‘human plus machine’ process. Predictive analytics allow investors to move from reactive, deal-driven behavior to proactive, strategic decision-making. With the right tools, firms can anticipate market shifts, target specific submarkets and underwrite investments with a much clearer understanding of what drives returns.

What role does data play in asset underwriting and portfolio construction?

Larissa Belova: One of the most important shifts is deconstructing returns and sensitizing risk. As the days of cap rate compression driving returns are in the rear-view mirror, instead, we now ask: what needs to happen operationally at the property level to achieve base-case returns across a variety of capital markets scenarios? That means developing tools and models that allow us to be granular and objective about assumptions on rent growth, occupancy and expense control, and then testing those assumptions against history. Have rents ever grown at that expected pace in this asset class or in this submarket, and under what conditions?

In certain cases, markets and sectors are evolving rapidly, especially the specialty sectors like industrial outdoor storage, seniors housing and self-storage, where data is fragmented. In addition, demographics, technology and secular changes are moving at a pace where history, while useful, may be less informative as a predictor of the future. That is where machine learning can help us gather tangible, real-time insights to sensitize demand drivers and outcomes, to better understand our upside and, more importantly, our downside.

This approach forces discipline. It replaces instinct alone with evidence-based insight.

How is the role of data and research evolving inside real estate investment firms?

Larissa Belova: Historically, research in real estate was often descriptive rather than prescriptive. It produced market views, but not always actionable recommendations. Research departments also tended to sit in silos within real estate organizations, rather than being embedded in the day-to-day decision making of investment teams. Further, data was fragmented, processes were manual and portfolio construction was highly discretionary.

Today, research must form part of the foundation of the investment process. The goal is to deliver data- and model-driven recommendations that directly inform sourcing, underwriting and portfolio construction. When done well, this can create consistency, potentially improve performance and allow firms to better customize portfolios.

How can investors and managers successfully implement a data-driven investment strategy?

Larissa Belova: In the real estate industry, ’data driven’ is easy to claim and hard to operationalize. The difference between a firm that uses data and a firm that wins with data is integrity – the ability to trust inputs enough to move capital, change strategy and stand behind the results with investors.

Most analytics programs do not fail because the models are unsophisticated, they fail because the underlying data is fragile: exported manually from portals, assembled in spreadsheets, revised in e-mail threads and quietly reshaped as it moves from property systems to reporting decks. That workflow does not just slow you down, it introduces hidden drift in definitions, timing and completeness. And once you have drift, you do not have a signal; you have a story.

You want to create an environment where data runs through a disciplined interface with clear definitions, and where data is pulled on a repeatable cadence, captures incremental changes and validates inputs automatically.

You also want a clear audit trail. This creates the kind of operating reliability that supports higher-conviction decisions. Alpha ultimately comes from better decisions that are faster, more consistent and closer to the truth. Data integrity is the prerequisite.

It also requires cultural alignment. Data has to be part of the firm’s DNA. That does not mean replacing experience or judgment. It means using objective criteria as a check on anecdotal evidence. Every investment decision should be grounded in the best available data and drawn from clean, transparent sources.

Education is also critical. You can have the best models in the world, but if the organization does not trust or use them correctly, they will not matter. Portfolio managers need to understand what is in the models, why they are pointing to certain markets or submarkets, and how those insights connect to real-world fundamentals.

How important is alignment between data-research teams and investment professionals?

Larissa Belova: There has to be close collaboration and a real partnership between investment and research teams. That alignment is as much about organizational design as it is about technology. Ultimately, it is about driving a culture of accountability, collaboration and innovation. At UBS Asset Management, our research and data teams sit under my CIO umbrella, and they are crucially integrated into our portfolio decision-making, both top down and bottom up.

Why is granularity such a recurring theme in today’s real estate investing landscape?

Larissa Belova: Granularity is a main source of competitive advantage. Markets are not monolithic. And the way we think about diversification and risk in a real estate portfolio is changing, based on a more precise understanding of property return drivers. Saying that we like California or we like the West Coast is meaningless in isolation. Performance can vary dramatically by submarket and by asset characteristics.

With the right scale, talent and data infrastructure, firms can scrape hundreds of sources, back test forecasts and identify very specific pockets of opportunity. Instead of making big, blunt calls, they can direct teams to focus on a small number of high conviction submarkets where the data supports potential for outperformance.

How does this level of insight change portfolio construction and client outcomes?

Larissa Belova: It enables a more thoughtful conversation about how real estate fits into the broader portfolio. With better risk metrics and deeper insight into performance drivers, investors can articulate what real estate is actually doing – how it contributes to income, diversification and risk adjusted returns. Firms that can answer these questions with data, clarity and conviction will be better positioned in the real estate 3.0 era.

To extract value in this environment, firms need to operate more like true owners and operators.

Larissa Belova