Systematic Real Estate · Fund I · A vehicle for family offices
Confidential working draft · September 2026 · Not an offering

Manufacture the opportunity.

We map where Mexican cities are going, find what is missing there before it is priced, build it with tenants who have already proven the demand, and sell it to the people who wish they had seen it first.

01

The thesis

Monterrey is breaking into hubs faster than its services can follow. The gap between the two is the fund.

Nearshoring put the plants on the periphery. Six municipios on the edge of the metro now hold 92 manufacturing sites of more than 250 workers each: Zuazua, García, Pesquería, Salinas Victoria, Ciénega de Flores, El Carmen. Industrial land there is already priced. The layer around it, housing, clinics, schools, credit retail, food, is not.

Congestion makes those hubs self-contained. One in eight Nuevo León workers already commutes more than an hour each way; another 27% sit between half an hour and an hour. TomTom ranks Monterrey among the world's most congested cities, with 89 hours a year lost at rush hour. Past that threshold a city stops being one market and becomes a dozen.

Obsolescence is the other side of the same movement. Plazas, old naves in the built-up ring, highway motels and vacant Infonavit blocks were built for a population that moved. Their owners still price them on the past.

5.26Mpeople, Monterrey metro
25hubs, by where activity ends
1.04Mpeople in 8 periphery hubs, schooling under 10 yrs
92plants of 250+ workers on the periphery
13%of NL workers commute over 1 hour
89 hlost per driver per year at rush hour
02

The system

Opportunistic funds wait for a broker to call. This one runs a pipeline, and the pipeline is what investors see.

Map

Where the gaps are

Each metro is cut into hubs by where daily trips end, not by municipal lines. A bill of services says what a hub of a given size and income should contain. What is missing is a ranked list of questions.

Signals

When a gap is opening

Chain openings, business registrations, job postings, listing prices and days on market, permits, water concessions, plant announcements. Refreshed on a schedule, diffed against last quarter.

Playbooks

What we know how to build

A short list: refurbish an obsolete asset, put a credit tenant on a corner, add the civic layer to an industrial node. Each with a standard model, a partner set, a dominant risk and a kill rule.

Field

Every gap gets a verdict

Two site visits at 06:30 and 18:00, ten structured interviews, one call to the target tenant's real estate team, and a written answer to "why hasn't anyone done this?" Dead, watch or pursue.

Any single dataset is buyable. The edge is the combination, and the discipline of turning every gap into a verified yes or no.

The other half is a query. When a broker calls with a plaza in Escobedo, the same map answers in minutes whether it fits: which hub it sits in, what that hub is missing, whether the asset converts to it, and which constraint kills it. The fund stays opportunistic without being naive.

03

Proof of concept · Monterrey

Thirty days, free public data, one metro. The map already ranks 25 hubs and says what each one lacks.

Every business in the metro (INEGI DENUE, 193,686 units with sector and size), the 2020 census at block-group level (2,185 AGEBs), and the official polygons went onto a hexagonal grid. A watershed on jobs, retail and population cut the metro into 25 hubs. A bill of twenty services, from OXXO to cold storage, was counted against each hub's catchment. The whole run takes under a minute and repeats for any Mexican metro by changing one config line.

Then we tested our own thesis against the chains. OXXO goes wherever people are; its density is flat across income. Coppel goes where people have credit. Bodega Aurrera marks where income drops off. Nine hubs carry the first two signals without the third: the chain-validated medium band, two million people, 39% of the metro. In those nine we asked which services that normally follow purchasing power, gyms, cinemas, private schools, clinics, have not arrived, and whether the market is already closing each gap.

9 of 25hubs validated by OXXO + Coppel
2.03Mpeople in them
24lag gaps · market closing them
4structural gaps · nobody building
49%of the business registry is younger than 2020
Appendix A · interactive analysis · hover any hub or cellOpen full page ↗

A gap is a question, not an answer. Every gap above threshold gets the field protocol before a peso is committed. The numbers above come from a static map; the signal layer that sees growth, not just stock, is Phase 1.

04

Where we play

Medium-income hubs where the chains have already confirmed the customer and the next layer has not followed.

Structural gaps first

  • Discount supermarket · Santa Catarina, Trabajadores structural
    69 stores where peers have 80; less than one opening a year.
  • Private primary, credit retail · San Nicolás, Predio Aldape structural
    71,000 people, none of either, nothing registered since 2020.
  • Cinema · San Nicolás, Industrial Los Parques structural
    An established hub of 72,000 with no screen.

Lag gaps, where timing is the edge

  • Private clinic, gym · Apodaca, Prados de la Cieneguita lag
    282,000 people, the most affluent medium hub, clinics at 7 of 16 expected. At the current pace the market takes 28 years to close it.
  • Credit retail, discount super · Guadalupe, Avante lag
    462,000 people, both below peer density, both filling slowly.
  • García, Valle de Lincoln one rule away
    97,000 people with no cinema and no private primary. Misses the medium band by 0.15 Aurrera stores per 100,000.

Why medium and not low: the low band is where the plants are, but income has to be there before services can be. The medium band is where the chains have already done our demand study for free.

05

Playbooks in Fund I

PlaybookWhat it isCycleDominant riskKill ruleSleeve
Refurbishment (flagship)Buy obsolete assets at old-use prices where the data shows a different population is already there; reposition to what the hub needs. Plaza to clinic and school; nave to last-mile; motel to operator housing.18 to 36 moTitle, structuralTitle unresolved in 6 months; structural cost above 40% of price~50%
Pad + credit tenantCorners on corridors the chains are entering. Solve title and permits, lease to OXXO, Autozone, pharmacy, QSR. Small, fast, repeatable.12 to 24 moTenant says noTenant real estate team declines; not entitled in 9 months~30%
Node centroThe missing civic layer for an industrial node: clinic, business hotel, food, services. Pre-leased to the operators the plants already use.24 to 48 moAbsorption paceAnchor not signed in 12 months~20%
Workforce housingRental housing inside the 30-minute ring of a cluster, with a shuttle. Only with two-plant demand.24 to 36 moSecurity, financingSingle-plant demand; security verdict redopportunistic
Constraint arbitrageLand that is cheap because of water, ejido title or permits.36 to 72 moPolitics, timeExcluded from Fund I: too long for a five-year vehicle. Fund II.
06

Fund I · MXN 200 million

Small on purpose, and built for family offices. Six to eight positions, two fast playbooks, one slow one, and a funnel that shows work every quarter.

Shape

  • Commitments MXN 200M target from six to ten family offices · first close MXN 75M · tickets MXN 10 to 40M
  • Term 5 years: 3 investing, 2 harvesting, plus two one-year extensions
  • Positions 6 to 8 · MXN 15 to 40M equity each · asset-level bank debt up to 50% loan-to-cost, so MXN 300 to 350M of projects
  • Vehicle Mexican trust (fideicomiso) or SAPI. Private, unlisted, no pension money: one document, one governance table, an investment committee where the families sit
  • Family rights co-investment on any position above MXN 40M of equity · first look on every exit, so a family can take a finished, leased asset onto its own balance sheet at the institutional price · full data-room access to the map and the funnel
  • GP commitment 3 to 5% · economics 2% on committed then on invested, 20% carry over a 10% preferred return with catch-up (illustrative, set with counsel)

Targets and underwriting floors

  • Net IRR 18 to 20% in pesos · multiple 1.7 to 1.9x
  • Yield on cost at least 13% at stabilisation, against institutional exit cap rates of 9 to 10%: a development spread of 350 bps or more on every deal
  • Pre-lease at least 60% of income signed before capex starts; anchor tenant signed before land closes on node centro
  • Exit FIBRAs and institutional core buyers, or the families themselves through the first-look right. The return is the spread between "nobody has looked here" and "obviously this needed to exist"
  • Currency peso leases, peso capital. No FX in the base case

The benchmark stack

Three bars, each answering a question a family office actually asks. What does the money earn doing nothing. What does it earn in the liquid version of the same asset class. And what does the family already earn owning finished real estate directly, which is the honest alternative to any fund.

BarReferenceTodayWhat Fund I commits to
Risk-freeCETES 28 days, Banxico auction of 22 Sept 2026 · reference rate 6.50%6.15%Preferred return of 10%, roughly CETES plus 400 bps. Carry earns nothing below it.
Liquid alternativeS&P/BMV FIBRAS index, one-year total return to August 2026+14.9%Performance benchmark: FIBRAS index plus 500 bps. If a listed FIBRA beats the fund over the term, the illiquidity was not paid for.
Doing it directlyStabilised, leased institutional-quality assets a family can buy today: cap rates in pesos, the same rate the fund underwrites its exits at9 to 10%The fund earns its keep only on the spread above that: yield on cost of at least 13% at stabilisation. The family ends up with the building it could have bought, at the cost of the land plus a risk it did not take itself.

Why not a fixed 15% hurdle: a pesos hurdle that ignores CETES rewards the manager for rate cuts and punishes them for hikes. Tying the preferred return to a public rate keeps the incentive honest across a five-year cycle. Why FIBRAS plus 500: the exit buyer is a FIBRA; if the fund cannot beat its own buyer's return by a margin, the family should own the buyer. Why the direct-ownership bar: most families in the room already own real estate; the fund has to be worth more than one more building.

07

What the families see

The report is the funnel, not the deals. A quarter with no acquisition is still a quarter of visible work.

Hypothesisgap or mismatch above threshold, no budget
Field-validatedprotocol complete, verdict pursue
Controlledoption or LOI, title review started
Entitledland use, water, CFE confirmed
Buildinganchor tenant signed, project equity
Exitedstabilised, sold to institutional buyer

Monthly

  • Count per stage, conversion rates, median days per stage
  • Kill reasons, in writing, with the "why hasn't anyone done this" answer for every dead hypothesis
  • Broker inbound scored against the map: how many calls, how many fit
  • Every family sees the same page, live: the map, the funnel, the field verdicts. No quarterly letter that says less than the data

Quarterly

  • Map refresh: hubs, top gaps, top mismatches, chain openings since last quarter
  • Per deal: yield on cost against underwriting, pre-lease, budget, dominant risk status
  • One new metro brought to Phase 0. Saltillo and Querétaro first
08

Risk ledger

RiskWhere it bitesBuilt into the playbook
Title and tenureEvery periphery site; ejido landTitle review at control stage; ejido specialist on the bench; kill at 6 months
SecurityCorridors, some northern and eastern nodesField verdict includes security. Walk away, do not price it
WaterAnything north of MonterreyCONAGUA concession registry (REPDA) in the data stack; concession confirmed before entitlement
Tenant concentrationAnything tied to one plantTwo-plant minimum or shuttle to a cluster; credit tenants on pads
Politics and permitsMunicipal governments change every three yearsDiversify municipios; entitle early; local partner per city
Thin dataThe map sees stock, not growth; DENUE flattens plant sizesPhase 1 signals (job postings, plant announcements, listings); field protocol on every gap before capital
Small fundConcentration; one bad deal is 15% of the fundPads and refurbs cap single-asset equity at MXN 40M; no asset above 20% of commitments
09

Next twelve months

Before first close

  • Q4 2026 Field protocol on the five top gaps in Monterrey; two tenant real estate teams saying yes on paper
  • Saltillo and Querétaro at Phase 0, same pipeline, same page
  • Bench signed in Monterrey: developer-operator, notario, ejido specialist, water and CFE consultant
  • Vehicle, terms and tax structure with fund counsel

After first close

  • H1 2027 One pad and one refurbishment under control; first anchor lease signed
  • Weekly signal layer live: chain openings, listings, job postings, permits
  • First funnel report with real numbers to the families
  • 2028 First exit to an institutional buyer; Fund II conversation on the platform, not the deals
The map and the process are the moat. The deals are what they produce.