"How can you not be romantic about Commercial Real Estate?" — probably Brad Pitt
In 2002, the Oakland Athletics did something that scandalized baseball's old guard. With one of the lowest payrolls in Major League Baseball, general manager Billy Beane — guided by the statistical philosophy of Bill James — assembled a team that won 103 games. They didn't do it by trusting scouts who loved a player's "good face" or a pitcher's "natural throw." They did it by asking a deceptively simple question: what actually wins baseball games? We believe commercial real estate is having its Bill James moment.
The Game Within the Game
Baseball's genius, once James cracked it open, was that it had been generating perfect data for over a century and almost nobody was using it correctly.
Sound familiar?
Commercial real estate sits on an extraordinary mountain of transactional data — cap rates, rent rolls, lease expirations, vacancy cycles, absorption figures, price-per-square-foot histories — going back decades across every submarket in every major city. And yet the dominant mode of analysis in most investment committees, portfolio reviews, and acquisition calls remains: does this pencil? The question that James forced baseball to confront was clarifying to the point of cruelty: what is the object of the game? In baseball, it's runs. In commercial real estate, it's total return — income plus appreciation, compounded over time, risk-adjusted against your cost of capital. Everything that doesn't connect to that number is a story. Everything that does is a signal.
So what is the commercial real estate equivalent of "he gets on base"?
It's rent growth relative to vacancy in a defined submarket, correlated against supply pipeline and demographic data such as employment density and income changes. It's a building in a micro-location where demand drivers are structurally intact and new supply is physically or politically constrained. It's not "great bones" or "irreplaceable location" offered without evidence — it's a property whose cash flow trajectory, when mapped against comparable assets over comparable cycles, demonstrates durable outperformance. It's the boring, unglamorous, repeatable signal hiding inside decades of transaction data that many have access to and almost nobody systematically uses — and those who do rarely run real analysis on it. We believe we have done this, using data from MSCI Real Capital Analytics.
1 — Finding Your Comparable
Every serious CRE analysis starts with finding comparables. The comp is to real estate what the box score is to baseball: the basic unit of meaningful information. But just as baseball misread box scores by obsessing over batting average while ignoring on-base percentage, real estate practitioners routinely misread comps by selecting them on the basis of convenience, familiarity, or — most dangerously — motivated reasoning.
A truly rigorous comparable selection process is disciplined and almost mechanical. Asset type, vintage, submarket, building class, lease structure, tenancy profile — all of these filters must be applied before a comp enters the analysis. This is far more than even the best analysts can do effectively by scrolling CoStar and zooming in and out for hours on end, but our deep neural network model is pretty good at it. The goal is not to find transactions that support a valuation. The goal is to find transactions that test it. That distinction, small as it sounds, is the difference between analysis and advocacy dressed as analysis.
2 — Mapping the Trends That Actually Matter
Once you have clean comparable data, the next level of Moneyball thinking is hyper-specific trend-mapping — not just where an asset is priced today, but where the underlying forces are pointing. This means tracking key metrics across asset type, price range, and submarket with the same rigor that a baseball analytics team tracks pitcher fatigue curves.
In practice, this looks like: what has transaction volume done in this submarket over three cycles? What is the relationship between deliveries and vacancy in this asset class? When interest rates moved 200 basis points in prior cycles, how long did it take for cap rates to reprice, and by how much? These are not exotic questions. They are the commercial real estate equivalent of asking what a left-handed pitcher's ERA is against right-handed batters in the seventh inning of close games. The data exists.
| Metro | Pairs | Bottom quartileQ | Median | Top quartileQ | Spread (pts) |
|---|---|---|---|---|---|
| Nashville | 577 | 7.4% | 15.0% | 46.2% | 38.8 |
| Miami / South Florida | 1,762 | 4.6% | 11.1% | 31.5% | 26.9 |
| Charlotte | 741 | 6.7% | 14.6% | 30.9% | 24.2 |
| Boston Metro | 806 | 4.5% | 9.5% | 24.1% | 19.6 |
| Phoenix | 1,957 | 7.0% | 15.0% | 25.8% | 18.8 |
| Seattle | 2,012 | 3.9% | 8.8% | 22.6% | 18.7 |
| Minneapolis | 659 | 2.1% | 6.6% | 20.7% | 18.6 |
| Salt Lake City | 445 | 6.5% | 11.4% | 25.1% | 18.6 |
| NYC Metro | 7,947 | 1.3% | 6.5% | 18.5% | 17.2 |
| Raleigh / Durham | 597 | 5.7% | 10.4% | 22.9% | 17.2 |
| Atlanta | 2,357 | 6.4% | 12.9% | 22.9% | 16.5 |
| Chicago | 1,976 | 2.4% | 7.3% | 18.9% | 16.5 |
| DC Metro | 1,152 | 3.0% | 7.6% | 19.1% | 16.0 |
The question is whether you're asking it. RCA data, for instance, provides exactly this kind of longitudinal, cross-market, cross-asset-class visibility. It allows an investor to situate any individual asset or portfolio within the broader performance universe and know not just whether a building is performing, but whether it is performing well relative to what was achievable given its type, its vintage, its geography, and the macro environment in which it operated. That is the difference between knowing your batting average and knowing your batting average adjusted for ballpark, era, and competition. One is a number. The other is useful information.
For example, the figure above draws on 8 years of transaction history — long enough to span more than one full cycle. That lookback matters: one of the questions we're most interested in is whether these ranges hold their shape through periods of market stress, or whether spreads blow out precisely when investors need pricing precision most. Vintage deserves its own mention here. Markets dominated by older stock tend to price with tighter spreads, plausibly as decades of trading history simply make them better understood. Younger, faster-growing markets carry wider dispersion, and therefore more room for the disciplined analyst to find mispricing. Disentangling how much of a market's spread is explained by vintage versus fundamentals is a question we will return to in a future article.
3 — Portfolio Management as a Living Process
Here is where most commercial real estate organizations stop. They do the acquisition analysis, they find reasonable comps, they build a five-year model with a terminal cap rate assumption that feels defensible, and then — almost universally — they shift into a mode of stewardship rather than active analysis. The asset goes into the portfolio. The model goes into a drawer. The market keeps moving.
Bill James would find this maddening. The whole point of systematic analysis is that it is continuous. Baseball teams don't build a roster in March and then stop analyzing the data in April. They are running live models on every player, every opponent, every situation, constantly.
Continuous, forward-looking portfolio management in CRE means treating the portfolio not as a collection of assets but as a collection of positions — each of which can be held, added to, reduced, or exited, and each of which should be continuously evaluated against the available opportunity set. It means having live visibility into rent-to-market ratios across every lease in the portfolio. It means tracking where you sit in the supply cycle for each submarket exposure. It means knowing, in real time, whether the assumptions embedded in your original underwriting are being confirmed or refuted by the data that has accumulated since acquisition.
4 — The Future Standard
And then there is the vision that should be genuinely disruptive to the industry: evaluating a commercial real estate portfolio the way Jane Street evaluates its trading book.
The world's premier quantitative trading firms do not have assets that are "performing fine." They have positions with live P&L, live Greeks, live risk metrics, and a culture of instantaneous accountability to what the market is actually saying versus what the model predicted. Every position is either making sense or it isn't, and the organization has the analytical infrastructure to know the difference in real time.
Applying that standard to commercial real estate is, for most of the industry, genuinely shocking to contemplate. What would it mean to know, at any given moment, the mark-to-market value of every asset in a portfolio, the implied return on each relative to current alternatives, the risk-adjusted contribution of each position to overall portfolio performance? It would mean the end of holding an underperforming asset because the quarterly report doesn't require you to acknowledge it yet.
It would also mean the beginning of a new kind of alpha generation, because the investor who has this visibility when others don't is the Billy Beane of the 2000s — finding value that the market has mispriced because the market is still using the old lens.
The Scoreboard Has Always Been There
Bill James didn't invent baseball statistics. He just insisted that the ones which actually mattered be used honestly and continuously. The data was always there. The runs were always what mattered. The only thing that changed was the willingness to follow the numbers wherever they led, even when they contradicted the consensus, even when they made powerful people uncomfortable.
Commercial real estate has its scoreboard. It has its runs metric. It has, in platforms built on data like that provided by MSCI RCA, the equivalent of the statistical infrastructure that transformed baseball. What it needs now is its Billy Beanes — the industry professionals brave enough to listen.
Data insights derived from MSCI Real Capital Analytics data.