The Mortgage They Cannot Afford

• Silicon Valley Bank’s March collapse sends shockwaves through an already-constrained global PropTech funding ecosystem
• GPT-4’s March launch ignites the most consequential AI conversation the property sector has ever experienced
• Jamaica mortgage rates reach decade highs as the BOJ sustains its inflation-fighting rate cycle
• PropTech venture capital declines sharply; Caribbean-focused property technology startups face a funding drought
• Jamaica’s property prices hold their ground despite affordability stress, underpinned by the structural supply deficit
• NLA land registry modernisation advances but full digital implementation remains years away from completion

The second quarter of 2023 was defined, for the global technology and finance world, by two events in March that together encapsulated the tension between the old financial order and the emerging digital one. On March 10, Silicon Valley Bank — the lender of first resort for a generation of technology startups and the institutional backbone of the global PropTech funding ecosystem — collapsed under the weight of a duration mismatch between its long-dated bond portfolio and its short-dated deposit liabilities. On March 14, OpenAI released GPT-4, a large language model that demonstrated professional-grade capability across a range of knowledge-intensive tasks that practitioners in real estate, law, finance and a dozen other fields had previously considered immune to automation.

These events were not related in any direct causal sense. But together they captured something important about the moment: the financial infrastructure of the old technology economy was under stress, while the intellectual infrastructure of a new one was being assembled at a pace that outran both regulation and institutional preparation. For Jamaica’s property sector, the quarter’s lessons were therefore simultaneously cautionary and galvanising. The funding pipeline that might eventually have supported Caribbean PropTech ventures was more constrained than ever. And the technology those ventures would be built around was more powerful than it had been a month before.

SVB and the PropTech Funding Winter

Silicon Valley Bank’s failure was, in the United States, quickly contained by federal deposit insurance and the rapid intervention of the Federal Deposit Insurance Corporation. But its effects on the technology venture capital ecosystem were not fully contained. SVB had not merely been a bank that lent to technology companies; it had been the institutional hub around which the US startup ecosystem — from the earliest seed rounds to the latest pre-IPO growth financings — had organised its banking relationships. Its failure shook confidence in the broader startup banking model and, more immediately, disrupted the working capital arrangements of thousands of companies whose deposits were temporarily inaccessible.

For PropTech specifically, SVB’s collapse was doubly significant. The bank had been a meaningful lender to PropTech companies at various stages of their development, and its failure removed a source of venture debt that complemented equity financing in the capital structures of many technology businesses. More broadly, the SVB crisis reinforced the narrative that had been accumulating since 2022: that the era of abundant, cheap, risk-tolerant capital for growth-stage technology companies was over. For Caribbean markets hoping to attract PropTech investment, the message was clear: the window of opportunity that the pandemic-era capital surge had briefly opened had closed, and any serious ambition to build regional property technology capability would have to be funded primarily from domestic and regional sources rather than international venture capital.

GPT-4 and the AI Inflection in Property

GPT-4’s release in March 2023 marked a qualitative shift in the AI conversation rather than merely an incremental improvement on its predecessor. The model’s demonstrated ability to pass bar examinations, medical licensing tests and financial analyst assessments at performance levels competitive with human professionals changed the terms on which the property industry had to think about AI’s eventual role. These were not tasks that the industry had previously believed AI could competently perform. The realisation that they could be performed, with appropriate domain context and professional supervision, had direct implications for property practice across the full chain of a transaction: valuation, legal due diligence, mortgage underwriting, lease review, market analysis.

Within weeks of GPT-4’s release, the leading North American PropTech companies were integrating its capabilities into their products. AI listing tools became significantly more capable, generating property descriptions with a specificity and a quality that removed any reasonable doubt about their commercial utility. AI document review tools could now process title chains, mortgage documents and commercial lease portfolios with a thoroughness and speed that no human team could replicate at equivalent cost. For the Caribbean practitioner watching this development from a market with no equivalent data infrastructure, the appropriate response was neither panic nor dismissal. It was the recognition that the technology was arriving and that the question of when it would arrive in Jamaica was now primarily a question of data readiness and regulatory preparation.

Jamaica’s Mortgage Market: A Decade High in Pain

The Bank of Jamaica’s rate hiking cycle, which had begun in late 2021 in response to global inflationary pressures that were quickly transmitted to Jamaica’s import-dependent economy, reached its most consequential phase for the mortgage market through the second quarter of 2023. Commercial mortgage rates for qualified borrowers had moved above 9 percent for most lenders, with some products reaching 10 percent or beyond for borrowers at the margins of credit qualification. The impact on the NHT’s mortgage portfolio was visible in the softening of new lending volumes and in the lengthening of application-to-disbursement timelines as affordability assessments required more careful navigation.

The structural consequence of sustained high mortgage rates was a bifurcation of the property market that was becoming more pronounced with each passing quarter. At the top of the market — the high-net-worth segment, the diaspora investor, the established portfolio holder — the rate environment was uncomfortable but manageable. At the bottom — the aspiring first-time buyer, the NHT contributor hoping to translate years of mandatory savings into a mortgage, the young professional pricing their first home — the market had effectively closed. The bifurcation was not merely an economic phenomenon; it was a social one, with implications for wealth distribution and mobility that would outlast the rate cycle that produced it.

Prices Hold, Supply Doesn’t Move

Against the backdrop of mortgage market stress, Jamaica’s property prices demonstrated the structural stubbornness that made the island’s market a consistent outlier in global affordability analyses. Where markets with abundant supply and highly leveraged buyer populations had corrected sharply in response to rising rates — Canada, Australia, parts of the United States — Jamaica’s prices held, in most segments and locations, within a few percentage points of their recent peaks. The explanation was straightforward: the supply deficit that had been accumulating for decades did not correct in response to interest rate changes. The 150,000-unit housing shortfall did not diminish because mortgages became more expensive. If anything, it deepened, as the deteriorating economics of affordable construction further reduced the pipeline of new supply.

The market’s response to this combination was a gradual but unmistakable shift in investor behaviour toward a hold-and-rent logic. Property owners who might otherwise have sold into a declining market chose to retain their assets and redirect them into the rental market, where the demand displaced by the purchase market’s effective closure for first-time buyers was creating yield opportunities that offset the carrying cost of holding. The rental market strengthened. Rental yields in Kingston’s urban segments moved higher. And the pool of long-term renters who had been priced out of purchase — sustaining landlord returns but building no equity of their own — grew quietly and with consequences that no quarterly review could fully capture.

The NLA’s Long Game

Jamaica’s National Land Agency continued its programme of land registry modernisation through the second quarter, advancing the technical groundwork for what would eventually become the e-Titles system. The procurement process that would culminate in the selection of a technology partner was proceeding through its evaluation stages. The legal framework — the amendments to the Registration of Titles Act that would give electronic titles the same legal standing as their paper predecessors — was under development. The NLA’s geospatial capabilities were being progressively enhanced through investments in GIS infrastructure and satellite imagery resources that would eventually underpin both the e-Titles system and the broader data environment that AI-powered property analytics required.

The pace of this work — measured, institutional, proceeding through procurement and legal processes that had their own logic and their own timelines — was inevitably frustrating to those who understood the urgency of the digital infrastructure challenge. The technology that Jamaica was building toward had, in markets that had already implemented it, demonstrated its capacity to transform the transparency, efficiency and accessibility of property markets in ways that benefited buyers, sellers, lenders, and the government agencies responsible for tax collection and planning. Every quarter that the implementation was delayed was a quarter in which Jamaica’s property market operated at a lower level of efficiency than it was capable of achieving. The work was progressing. It was not progressing fast enough.

Outlook: The AI Era Begins Whether Jamaica Is Ready or Not

The remainder of 2023 will unfold in the shadow of two certainties: interest rates will remain elevated longer than most market participants would prefer, and AI’s capability in property-relevant tasks will continue to advance regardless of the pace of Caribbean adoption. The question for Jamaica’s property community — practitioners, developers, lenders, regulators, and the government agencies that set the institutional framework — is not whether to engage with either of these realities but how to engage with them in ways that position the sector advantageously for what comes next.

The mortgage market will ease when rates ease, and the structural demand that has been compressed by affordability constraints will release into transaction activity when it does. The AI tools that are today available only to the most sophisticated and data-rich property environments will, as they are commoditised and as Jamaica’s digital infrastructure matures, become accessible to practitioners in every segment of the market. The trajectory is clear. The urgency is to prepare, to build, and to ensure that when the moment of acceleration arrives, Jamaica’s property sector is not discovering the need for digital foundations that should have been laid years earlier.

Follow Jamaica Homes on Youtube @jamaicahomes and Instagram @jamaica_homes and on Facebook @jamaicahomes Send us a message or email us at onlinefeedback@jamaica-homes.com or editor@jamaica-homes.com