Think back to third grade. Was there a U.S. map in your classroom, maybe with state lines and capitals? Now imagine other versions of that map you have seen. Topographical, with miniature Appalachian and Rocky mountains. Satellite maps showing dense green forests in the east, deserts in the west. And you’ve probably seen colored geological maps of the U.S., which look a little like bad finger paintings.

Regardless of complexity, any map with data represented on it falls into the Geographic Information Systems (GIS) discipline, and we all have become so used to seeing GIS maps in the media we almost never stop to think about how they’re created. They all start with true underlying geography and connect the spatial world to related information like climate, terrain, or geology, and to less-inherently related information like politics, health, economics or defense.

Fast forward to today’s age of big data and big computing, and GIS can take the form of deep micro-location analysis, yielding granular insights into real estate utilization and value. Recent research explored the potential of this methodology, studying the performance of retail assets in a single small British city between 2010 and 2023.¹

The authors eschewed zip codes and neighborhoods, instead mapping the city of Bracknell (pop. 65,000) in 250-square-meter micro-blocks. They measured retail property valuations (as a proxy for demand) and retail square footage (supply), eventually creating detailed three-dimensional colored maps showing property performance over time (by color) and strength (by height). Values shifted substantially over time, often in locations that conventional submarket logic would have ranked as interchangeable. Supply barely changed at all, unsurprisingly.

Their key finding: Performance varies “substantially” within an MSA, and even within very small micro-markets. Comparable real estate assets exist in real estate markets whose “boundaries do not necessarily align with physical or administrative divisions such as streets [or] postcodes…. As a result, two similar properties on the same street may belong to different market segments, while properties in separate streets may share similar market characteristics.”

The 3D model below illustrates the point. There are significant performance differences between specific micro-locations pre-COVID, with the dark red being areas of positive change in real estate value, white being negative, and pink being flat.

The novel methodology introduced here feels in some ways intuitive, like something we should have been doing for a long time, but this level of GIS sophistication is new. Publicly available property-level data - values in this case, but rents, occupancy and other metrics could work - is hard to get, especially for long time periods. And once you have it you also need the GIS skills to create the kinds of micro-blocks the research suggests you need to examine to see meaningful and actionable performance trends.

This is going to complicate your day: If your acquisition analysis starts at the submarket level, you may be averaging away the very signal that determines whether a deal works.The research calls for dis-aggregating city- and sub-market-level metrics and assessing market boundaries in a data-driven, dynamic way, “rather than relying solely on fixed geographic demarcations.” The work affirms the diversity of demand patterns that can occur in small areas, even on a single city block. That kind of signal in a market targeted for investment would be invaluable on its own and could help drive more accurate predictions.

Notably, the study examined retail performance before and after COVID, which to that industry was a significant shock. This examination is important, as it reveals an authentic reflection of risk in these assets. If a retail location’s performance stays flat or even increases after such a shock, as some did, one would argue retail in those resilient locations deserve a higher market valuation. The observed norm, unsurprisingly, was poor resilience, “suggesting most locations… struggled to recover post-COVID.” But the recovery was not uniform; it varied dramatically even within very small geographic areas, much like the pre-COVID performance above.

It’s important to say that this paper is about a methodological experiment, not a causal study. Even if the “why” behind the paper’s observations was left to future researchers, the experiment succeeded, as it revealed there may be much more complexity to demand than traditional real estate investors currently appreciate. This offers sophisticated investors new ways to gain an edge in their acquisitions and developments.

In our experience as humans trying to be good at our investment jobs, we carved markets into neighborhoods, submarkets and zip codes because we needed to simplify an impossibly complex system. But demand for real estate doesn’t organize itself that way. As this study shows, performance varies meaningfully at a level far more granular than our mental models allow, to the point where two buildings on the same street can behave like they belong in entirely different markets. Blunt frameworks, like those we’ve used, risk turning what we think is informed investing into something much closer to randomness.

That’s an uncomfortable thought, especially in a business that prides itself on experience and pattern recognition. But it’s also the opportunity, and papers like this illustrate how much signal still sits buried beneath the surface. The tools are getting better, the data is getting richer, and the investors willing to challenge their own assumptions can gain an edge.

Who: Sonic Fire Tech

What: Using sound waves for fire suppression.

The Sparkle: With one of the Haystack team moving to SoCal, we’re hearing more about innovative fire suppression and even watching people put sprinkler systems on their roofs (to extinguish incoming embers). But nothing has been as interesting as Sonic Fire Tech, an in-home fire suppression system that uses no water or chemicals and that’s safe for pets and people. Sonic uses “targeted infrasound waves… that disrupt the combustion process.” We don’t know what that means but we like the vision, and Sonic is just now seeing its first field installations.

Ivy League graduates plus graduates of four other schools - University of Chicago, Duke, Stanford and MIT - collectively make up fewer than one in 200 college graduates, but they occupy a much higher percentage of our country’s leadership positions. New research reveals how often you can expect to find one of these graduates in various successful sub-groups.  Hat-tip to The Atlantic for a great related (paid) article. 

Research Link: https://opportunityinsights.org/wp-content/uploads/2023/07/CollegeAdmissions_Paper.pdf

Editor’s Note: The Real Estate Haystack believes in sharing valuable information. If you enjoyed this week's newsletter, subscribe for regular delivery and forward it to a friend or colleague who might find it useful. It's a quick and easy way to spread the word.