Listed Real Estate Asia, KT Capital Group

Jacob Cheng

Jacob has 7 years of professional buy-side experience in real estate investment and listed equities.  Jacob worked as a portfolio manager/analyst in Global Listed Real Estate team in AMP Capital and Macquarie Asset Management.  Prior to that, Jacob has spent 5 years in APG asset management, in the listed real estate team.  As one of the largest RE funds in Asia, he co-managed EUR6bn investment portfolio with the team with a focus on Hong Kong and China property.  He led the USD200m cornerstone investment in Wanda Commercial Properties IPO (the largest mall operator and developer in China).  Prior to this, Jacob has 3 years of experience in investment banking in Bank of America Merrill Lynch, covering equity, debt and M&A transactions across real estate, consumer& retail and TMT sectors.  Jacob specializes in fundamental bottom up research.  Aside from work, Jacob has a personal interest in spirituality and personal development.  He practiced meditation and has published a book about meaning of life in 2016.

Listed Real Estate Asia

American Tower: Monetizing the AI Edge Data Center Revolution

Keppel DC REIT as the Primary Beneficiary of the Global AI Inference Pivot

The AI Inference Supercycle: Which SG REITs Can Survive the 100kW Upgrade?

KT Capital Group

Principal investments and advisory firm that focuses on connecting the CIS countries with East Asia.

Principal investments and advisory firm

Thematic (Sector/Industry)

United states Real estate / Thematic (Sector/Industry)

<ul>
<li>AI inference demands sub-10 millisecond latency, forcing massive compute workloads away from centralized hyperscale campuses to distributed edge data centers.</li>
<li>Five-Year grid interconnection delays make edge sites critical bypass valves, positioning tower operators like American Tower to monetize pre-zoned, power-ready land.</li>
<li>A core pair trade going long American Tower and short Keppel Infrastructure Trust captures edge growth while hedging against elevated 44.2% gearing and finite-life concession cliffs.</li>
</ul>


AI inference demands sub-10 millisecond latency, forcing massive compute workloads away from centralized hyperscale campuses to distributed edge data centers.
Five-Year grid interconnection delays make edge sites critical bypass valves, positioning tower operators like American Tower to monetize pre-zoned, power-ready land.
A core pair trade going long American Tower and short Keppel Infrastructure Trust captures edge growth while hedging against elevated 44.2% gearing and finite-life concession cliffs.


Edge Inference: Profiting from AI Decentralization and Grid Constraints