About
Brad is the founder and editor of Thesis Driven (www.thesisdriven.com), a publication…
Articles by Brad
Activity
39K followers
-
Brad Hargreaves shared thisNew York is about to build its first Hong Kong-style megatower complex: 4,000+ apartments, 115 stories, 3 towers, 1 city block. The city just kicked off public review for 100 Gold Street, right across from the Brooklyn Bridge, and if it gets built it will be one of the largest residential projects anywhere in the Western world. Today the site is a nine-story 1960s city office building. GFP Real Estate, the firm behind the 25 Water Street conversion, was picked to replace it, and the plan has grown a few times since the original RFP, which called for about 1,000 units, to 3,700 and now 4,000, with 1,000 of those income-restricted. There's no city subsidy, just 485-x and the market-rate units carrying the affordable ones... a big part of why the unit count keeps going up. What I find most interesting is that there's almost no American comp for this. Stuy Town and Co-op City have more units, but they're spread across dozens of buildings and a lot of acreage. Even the biggest single apartment buildings in the country are a fraction of this size. To find something that really looks like 100 Gold, you have to go to Hong Kong, where thousands of homes stacked on top of a shared podium with retail, transit and amenities is just how a lot of people live. The towers over Kowloon Station are the obvious reference point, and the 100 Gold plan, with its public gym, senior center and plaza folded into the base, is heading in that direction. There are also few operational comps. Nobody in the US has leased up and run 3,000 market-rate apartments on one block at once. How do you phase absorption without flooding your own submarket? How do you staff it, and what do elevators, trash, package rooms and move-ins look like at the scale of a small town? Those are the problems Hong Kong operators have been solving for decades but don't really appear in US multifamily. It's also a reminder of why public land matters. There's no assemblage here and no holdout neighbor, just a full city block the city already owns, which is exactly the kind of site where this scale becomes possible at all.
-
Brad Hargreaves reposted thisBrad Hargreaves reposted thisAI is changing the economics of real estate. At FUTURE, Brad Hargreaves shares a perspective on what that shift means for the industry and what it takes to build stronger performance across a real estate business. His advice to operators: think through how AI will shape economies of scale in the sector and private equity money, and how to respond from a tech strategy perspective. #appfoliofuture #rpm #AI Show Less
-
Brad Hargreaves shared thisAmerica's public schools are losing students fast, and the worst of it is yet to come. Enrollment peaked at 50.8 million in 2019 and is already down to 49.6 million. Because enrollment lags fertility by years, it's projected to hit 46.9 million by 2031 and keep falling from there. The losses aren't spread evenly. New York, Illinois, and California accounted for 44% of the national decline from 2019 to 2024, with NYC losing 11% of its students, LA 16%, and Chicago 21%. The ten states that held flat or grew added about 80,000 students combined, against a national loss of 1.4 million. That puts cities in a tough spot, because fewer students means less revenue but costs don't fall with it. A school with one music teacher can't drop to 0.8 music teachers when enrollment falls 20%. And at the same time, states from Missouri to Texas are letting seniors freeze their property taxes, which chips away at the source of roughly 36% of all K-12 revenue. So districts are closing schools. Denver closed ten in 2024, Philadelphia approved 17 this year, and St. Louis is weighing 22 of its 62. It rarely saves much money (a Stanford study of California closures found $447 per pupil in savings almost fully offset by $433 in lost revenue from students who didn't follow), but it does put buildings on the market. For real estate, that's interesting. Old schools have a lot going for them as residential conversions: operable windows on multiple sides, high ceilings, corridors already sized for egress, and a repeating classroom module that lays out well as one and two bedroom units. The gyms and auditoriums are the hard part. The school-to-multi pipeline is still small, around 9,300 units at the start of 2026, but RentCafe has it as the fastest-growing category of adaptive reuse. I'd be careful cheering too hard, though. The more I study fertility, the more I think it's cultural, and one of the best predictors of whether young people have kids is whether they're around kids. A lot of them spend their 20s and 30s in exactly the cities where schools are emptying out fastest. Plenty of these conversions will pencil over the next couple of decades. But cities should be careful not to replace every school with an apartment building and every playground with a dog park. Full letter on the K-12 crunch is linked in the comments.
-
Brad Hargreaves shared thisA solo real estate operator can't out-hire a national platform's research team, but this Thursday we'll show how one can out-see it. Our case study is Lionel: 30 years old, five years at a larger firm, now out on his own and raising for a foothold in his hometown of Nashville. He has no analyst bench, just Claude and a plan to spot the opportunities nobody else is looking at. Over 90 minutes we'll build him three connected pieces: • An interactive visualization that turns public price and transaction data into a real picture of how Nashville has actually moved, not how a national report says it has • An AI wiki that helps him test his thesis against the data and sharpen it into something worth defending • A research agent that keeps gathering and filtering the news, listings, and signals tied to that thesis on a recurring schedule, long after the session ends What you walk away with is a system that keeps surfacing the few developments worth acting on, which submarkets, which sellers, which shifts, before everyone else catches up. It's market-agnostic, so you can point it at your own geography and product type the same afternoon. No coding required, and if you can form a view on a market, you can build this. Market Research with Claude. This Thursday, October 1, 12:00 to 1:30 PM ET on Zoom. Everyone who registers gets the recording, slides, and skill files - whether or not you make it live. Link in the comments.
-
Brad Hargreaves shared thisI just wrapped up 3 days at Blueprint in Vegas, and I have one surprising takeaway: a lot of VCs will get bailed out over the next 24 months. That's because there's serious appetite for M&A in proptech right now, which I didn't expect at all. It felt like every other company I sat down with was looking to buy startups. And it wasn't just the big PE groups, either. Plenty of them were operating companies looking to make strategic acquisitions that plug into what they already do. These weren't dumpster divers hunting for cheap acquihires of overfunded, underperforming companies. This was genuine strategic interest, with a willingness to pay for it. As far as I can tell, there's one reason for this: data moats. Everyone has figured out that features aren't a defensible moat in a world where AI can build them for you. Distribution helps, but it isn't a moat either. And nearly every software company I talked to is terrified that the frontier models will get good enough that their products are no longer necessary. So where's the moat? It's data that the frontier models don't have: construction projects and bidding, maintenance requests and tenant complaints, building systems breakdowns, leasing inquiries and tour bookings. All the data cruft these unprofitable startups have been generating for years is now gold, and the incumbents are willing to pay real money for it. For a lot of funds sitting on proptech portfolios that looked stuck a year ago, that's a very welcome way out. Who do you know is actively looking to buy?
-
Brad Hargreaves shared thisBrad Hargreaves shared thisPLANSCAPEARCH 2026: RADICAL OPTIMISM Exploring the Next Generation of Urban Infrastructure MEET THE SESSION 02 PANEL Who turns urban ambition into reality? The second conversation at PlanScapeArch brings together four perspectives on cities, development and investment: Brad Hargreaves | Thesis Driven Entrepreneur, writer and founder of a research platform focused on housing, urban development, proptech and real estate investment. Tom Wright | Regional Plan Association President & CEO of RPA, where he has helped lead major regional planning initiatives and conversations about the future of metropolitan New York. Román Viñoly | Rafael Viñoly Architects Architectural and development leadership spanning design, operations, finance and strategic planning. Aida Stoddard | MAG Partners LP More than 25 years working across construction, design and real estate development in New York. Together, they'll bring multiple sides of the city-building equation into one conversation. What does optimism look like when it has to pencil out? Friday, October 2, 2026 9:00 AM–5:00 PM 6 CEUs In-Person + Virtual Center for Architecture 536 LaGuardia Place, New York City PlanScapeArch is the annual Joint Symposium co-organized by ASLA-NY, AIA New York | Center for Architecture, and American Planning Association NY Metro Chapter, with ULI New York as Promotional Partner. Special thanks to our sponsors Kimley-Horn and ECOncrete. Registration is now open. Join us October 2. https://lnkd.in/eJd7RdP4 #PlanScapeArch #RadicalOptimism #UrbanInfrastructure #PublicRealm #Architecture #LandscapeArchitecture #UrbanPlanning #UrbanDesign #NYC
-
Brad Hargreaves shared thisPhilippe Kjellgren spent two decades reviewing the world's best hotels, and his complaint was that most of them could have been anywhere. He has now co-founded a brand on the opposite premise, with a 135-acre private island in the Grenadines as its flagship. The category barely exists yet. Soneva coined "barefoot luxury" in 1995 and runs just three properties thirty years later. Habitas and Zannier are brands rather than platforms. Meanwhile the private island resort market hit $7.6 billion last year and is projected to pass $15 billion by 2033, with top Caribbean resorts holding $1,000 to $3,500 a night. Demand was never the constraint! Branson bought Necker in 1979 for $180,000 against a $6 million ask, which looks like a steal until you realize it's 74 acres of scrub with no water, no power, no dock, and a government deadline to develop it in four years. The build cost $10 million, and Necker now goes for more than $100,000 a night. That is the island equation: Land is a rounding error, the operation is the asset, and success or failure depends on whether the developer can run a jobsite where every bag of cement arrives by barge. Desalination and power generation come with every project, hurricane risk means self-insuring, and no bank will lend against an undeveloped island. Limestone Capital is solving this by building (and pre-selling) villas rather than betting on massive hotels. Buyers finance the construction before the resort has an operating history. The Swiss firm behind Voaara runs more than €1 billion across four funds and 12 hospitality platforms. Palm Island is 30 keys and 30 branded villas on a $155.5 million budget, with $49.5 million of the residential piece funded through pre-sales. Savills counts 910 branded residence developments globally, up from 323 in 2015, and branded product carries a 36% price premium in the Americas. Villa sales fund the build, management fees become recurring revenue, and owners become repeat guests. Kjellgren argues the place should tell the architecture what to be. Whether that survives the same template being run in Costa Rica, Mexico, and the Arctic is worth watching. Paul Stanton dug into the full Limestone model in Thesis Driven. Link in comments.
-
Brad Hargreaves shared thisOn November 4th I'll be on stage at EliseAI's leasing and marketing summit in San Antonio. I'm giving the market outlook, but it's remarkable how short the shelf life of that sort of thing has gotten. A market outlook used to be an annual exercise. You set your assumptions in January, and by August most of them still held up well enough to plan against. Now the prospect journey, the tooling, and the supply picture in any given market are all moving faster than ever. Figuring out which of those shifts are structural and which ones are temporary is a big part of the job right now. Navigating it all is a big part of what I'll be covering on stage. November 4-5, San Antonio, free to attend. Come find me if you're going.
-
Brad Hargreaves shared thisSpec office doesn't pencil anywhere in America right now. We modeled 18 markets to find out when it does, and only two get there by 2028. The demand side has been recovering for two years. Net absorption just notched its ninth straight positive quarter, asking rents are growing at their fastest pace in six years, and vacancy is falling in 19 of the top 25 markets. But almost nobody is building spec. Over the past year we count four spec projects that broke ground, three in metro Dallas and one in Manhattan, and that one only happened because a ground lease deadline forced BXP's hand. Everything else starting today has an anchor: Jones Day at Knox & McKinney, Amex at Two World Trade Center, two law firms at 725 12th Street in DC. Per a JLL broker in Dallas, these buildings need 40% or even 50% preleasing just to get a construction loan, which means they aren't really spec at all. The gap between operational recovery and capital formation comes down to math. So we built the math. Hard costs from Rider Levett Bucknall's 2026 prime office ranges ($705 a foot in Manhattan, $348 in Uptown Dallas), trophy rents from real comps, land estimates, and a required yield on cost of the local exit cap plus 350 basis points of development risk. BXP tells investors it targets 8% or higher on projects that are already mostly preleased, roughly 300 over a 5 cap. A spec building has to clear that hurdle, and then some. Brickell and Manhattan cross the line first. Uptown Dallas and West Palm Beach follow by the end of 2030, with Austin, Nashville, and DC later in the decade. Six markets, including Boston, Denver, Houston, and Atlanta, never pencil inside a ten year window, because absorption is too weak to outrun construction cost inflation. Penciling isn't the same as funded, though. The first wave of spec office gets capitalized with low-leverage private money, because institutional capital moves slowly on gradual shifts and overreacts once someone else goes first. Nobody wants to be the person bringing a spec office deal to the investment committee. Office was once real estate's trophy asset. It may now have to climb the institutional ladder the same way surf parks and outdoor hospitality did. Full model and market-by-market timeline in the letter, link in the comments.
Brad Hargreaves replied to a comment
13h
Chelsea Kneeland imagine a rainy Friday night and 4,000 people all order Doordash simultaneously
Brad Hargreaves replied to a comment
13h
James Brechlin 3% vacancy is 120 vacant units at any given point in time. In prime summer months you're turning an entire 300 unit building in a weekend
Brad Hargreaves replied to a comment
13h
Reid Kaufmann perm lenders don't often syndicate, but in this case...
-
Brad Hargreaves reacted on thisBrad Hargreaves reacted on thisThe most overlooked arbitrage play in real estate: short-term lending on a single family home. The loans generate equity-like yields from a senior, asset-backed position. And it's all thanks to Blue Owl. There's a lending gap in real estate where money is scarce and demand is surging, and almost nobody is filling it. It's the market for loans under $6 million to the small builders and renovators who make and fix up America's homes. Here's how it got this way: 1/ Private credit cracked Investors asked for $10 billion back from two Blue Owl funds in six months, and Blue Owl capped payouts at 5%. Too many of those loans were backed by a company's future earnings, not anything you can sell. 2/ Banks stepped back, and the institutions stayed at the top Banks' share of commercial real estate lending fell from 43% to 24% in two years. The insurers, REITs and private equity firms that stepped in only want $140M-$260M bundles of loans. Big buyers need big, uniform pools. 3/ Small loans got ignored A $2 million loan on one house is too small and too different to bundle, and too costly for banks to underwrite 4/ Demand is surging The U.S. is short 4 to 5 million homes, and 40% of the housing stock predates 1980. Builders need fast cash to build and fix. That's the perfect storm: huge demand, shrinking bank supply, and institutions that can't be bothered. These are called residential transition loans (RTLs). A builder borrows for 6 to 24 months to build or renovate, then sells or refinances. The sale repays the lender. Because few lenders compete, the lender sets the terms: • 8.5%-10.5% interest, plus 1%-2% fees • First lien, so the lender is paid back first from the property • Loans capped under 70% of property value, so the builder's cash takes losses first Personal guarantee from the borrower That's an equity-like yield from a senior, asset-backed position. The arbitrage is simple: go where institutions won't. A4 Capital Partners is doing it. Our full deep dive is in the comments.
-
Brad Hargreaves liked thisBrad Hargreaves liked thisIt seems like overnight that nearly every real estate GP that I know has made the transition from real estate --> real assets --> to infrastructure. Is it just me, or has every real estate firm suddenly become a real assets firm and/ or an infrastructure company in the last 18 months? There's nothing wrong with that, and there are some very smart people who have made the pivot, but I also know of some that are changing their pitch to get allocation from a different pocket. As we all know, investors are very, very savvy, and they'll sniff out a marketing pitch from reality very quickly. For those who think they're going to get a quick win when changing the headline of their pitch deck, they're in for a rude awakening. For those with deep operational expertise and infrastructure chops, the wind is at their back. To all my friends in real estate, think of this as a generational buying vintage, and let's spend our time educating investors on what the role of real estate is in a portfolio of alternatives. As "traditional" real estate investors expand their focus, those that stay hyper focused will benefit. Is this obvious or am I just saying the quiet part out loud?
-
Brad Hargreaves liked thisWas prepping this today -- it's going to be a good oneBrad Hargreaves liked thisLast Friday, our CTO Daniel Kronovet and I spent the day in Nashville. In an afternoon he rebuilt the Market Research "department" for a regional value-add industrial firm. For $2,500/month in Claude credits and subscriptions. People are tired of AI pro tip posts. Including me. But the efficiency leaps and cost savings happening everyday are still mind-blowing. There are opportunities to cut operating costs by 50-99% everywhere. In one afternoon, Daniel helped them build a research system in Claude: • Two agents that read market news and build a 5-minute daily podcast • A custom site to pressure-tests deals against the investment theses • Live local pricing and transactions, on a tv in the office • An automated comp tracker that pulls from flyers and OMs in inboxes • A one-page market memo that drafts itself before every IC meeting A summary of what he taught them: 1/ Join datasets on the parcel ID Most firms look at one dataset at a time. Have Claude merge deed transfers, assessor rolls and zoning maps on the parcel number. Then you can see who bought what, what they paid per foot and what they're allowed to build. 2/ Build the dashboard as code, not a slide Ask Claude to write the dashboard as a small web app instead of a chart in a deck. Next month, drop in fresh data and every chart, submarket filter and trend line rebuilds in seconds. 3/ Turn your theses into a living knowledge base Keep each thesis as its own file inside a Claude Project, so Claude always has your strategy in memory. Every new data point gets tagged as supporting or contradicting a thesis. Each week, Claude flags which theses are getting weaker. 4/ Plug agents into your real sources Connectors (MCP) give agents direct access to your inbox, drive and news feeds. When a broker OM arrives, Claude parses it into a structured comp table: address, SF, clear height, price and cap rate. 5/ Run agents on a schedule with a scoring rubric Agents run at 6am, score every article 1-10 against your buy box and surface only the 8+ items. Each item cites its source and links to the thesis it affects. 6/ Turn your best analyst's judgment into a skill Write down once how your best person underwrites a submarket. Save it as a Claude skill, and every agent applies the same logic every time. Daniel is teaching this live on Thursday. 90 minutes. No coding required. Link in comments.
-
Brad Hargreaves liked thisBrad Hargreaves liked this78 days ago I had a baby. In "normal life" days, that sounds like a lot. But in the days when you become responsible for another tiny and tall human (99th percentile anyone!?) and struggle with wild physical and mental health issues postpartum? It’s a lifetime. I'm slowly, but officially getting back at work. Exhibit A of how that's going 👇 I sent an email to myself that simply said: "remember." Remember *WHAT* Lissy!? The email doesn't say. There's no crumb or clue. Not even in the timestamp. I have not a trace of recognition in my brain of what this was for. So... I will never know what I needed to do, or why I needed to do it. For whomever I let down, didn’t send the document, or forgot to respond to… I am SO sorry. If any of you can figure out what I was trying to remember (or make a good guess), you'll officially make my weekly wins list AND I'll gift you a gift card for a 15-minute nap (in this economy, this is just about the most valuable thing you can get)... Real talk: the last 11 weeks and 1 day have been the hardest of my life. But I'll share more about that soon on my substack. That said: it's good to be back! As my postpartum therapist put it: I'm good at being this (working gal) version of myself. So for those thinking I’m crazy for being here...it's TOO SOON!? PIPE down! Enough with the parental-industrial-judgement-complex. Believe it or not: this is an incredibly grounding exercise... writing to a ton of faceless hardworking, professional strangers about my innermost thoughts and experiences, vulnerabilities OH…and my obsession with mental fitness 😎🤗 I've spent the last decade studying the brain, and I've never experienced this level of change in my own (that I can remember…OY 😅). I can't finish many thoughts. Can't find the right words. I experience fog so thick I need a freakin’ flashlight… The primary thing I have been able to focus on is one very squiggly, crunchy, munchy, and bunchy human named Magnolia James (aka Maggie). And yet? I've never been more alert at 3am. Never more focused on the priority (not priorities) that actually matters each day. Never more ready for action at a moment's notice. My postpartum brain is wild, and I'm making a YouTube video all about it, for anyone who's been through it or loved someone through this jungle 😶🌫️😶🌫️😶🌫️ As horrifically cheesy and perfect as it sounds for a Linkedin post: Mental Fitness (and my people) are everything to me and have been at the core of how I'm working through all of this. Ever more excited to share what I'm learning, and how I'm growing, with all of you. Stay tuned for more reflections on the brain, working parenthood, juggling two conflicting things at the same time, and "remember"... Welcome back to... me? 😬😍 And Welcome to the big, special, and wild world my Magnolia flower, my Maggie James, my MJ … my CRUNCHY! Pics to come soon. I missed y’all!!! #MentalFitness #LosingmyMYNDY #WorkingMama #PostpartumBrain
Experience & Education
Recommendations received
4 people have recommended Brad
Join now to viewOther similar profiles
Explore more posts
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content