DiLuPa Family Portfolio — Investment Thesis 2026

✨

AI-Driven Consumer Innovation MED

$5,913,040 · 8.3% of portfolio

DUOL	$2,345,090	3.3%
ZETA	$2,020,291	2.8%
SHOP	$1,547,660	2.2%

IR 2026 Theme 02 — MED conviction. Software platforms embedding AI to unlock pricing power. DUOL: AI tutors driving DAU growth and subscription ARPU — personalized learning at scale. ZETA: AI marketing cloud displacing legacy ad-tech incumbents; CDP platform wins enterprise. SHOP: AI merchant intelligence (Sidekick, predictive analytics) driving GMV attach rates.

📱

End-Game AI / On-Device HIGH

$1,963,180 · 2.8% of portfolio — UNDERWEIGHT vs IR 2026 target 10–15%

AAPL	$1,000,025	1.4%
GOOGL	$963,154	1.4%

IR 2026 Theme 03 — HIGH conviction. The device+OS+personal-data layer owns the ambient AI interface. AAPL: 1.5B+ active devices, Apple Intelligence, Neural Engine, iPhone supercycle. GOOGL: Search+Gemini+Android+Cloud — most capable multimodal AI stack. Both significantly underweight at 2.8% vs IR 2026 10–15% target. Priority add for this portfolio.

🤖

AI Infrastructure & Platform Leaders HIGH

$4,863,418 · 6.9% of portfolio

NFLX	$1,889,878	2.7%
META	$1,052,547	1.5%
NVDA	$1,042,496	1.5%
AMZN	$878,497	1.2%

Mixed IR 2026 themes. NVDA: AI Infrastructure Theme 01 (HIGH conviction) — GPU backbone, CUDA moat. META: AI Consumer Innovation Theme 02 — Advantage+ AI targeting driving CPM expansion, 3B+ users. AMZN: AWS hyperscaler (AI Infra adjacent) + IR 2026 Gap Analysis pick. NFLX: Streaming Theme 11 (LOW conviction) — profitable but lowest-priority per IR 2026 framework. Missing from AI Infra: MU (HBM3e), SNDK (NAND storage), CRWV (AI GPU cloud) — all named explicitly in IR 2026 Theme 01.

🌎

LatAm Digital Finance HIGH

$22,643,380 · 31.9% of portfolio — EXTREME CONCENTRATION

NU	$20,784,000	29.3%
MELI	$1,859,380	2.6%

IR 2026 Theme 05 — HIGH conviction. NU: world's largest digital bank by user count, 110M+ customers, cross-sell barely begun (credit, insurance, investments). MELI: $200B+ annual payments TPV, logistics network moat. NU at 29.3% is extreme single-stock concentration — IR 2026 caps LatAm at 4–7%. Governance priority: trim NU on strength, redeploy into missing HIGH conviction themes.

⚕️

Healthcare & GLP-1 Revolution MED

$9,130,144 · 12.9% of portfolio

OSCR	$5,800,032	8.2%
LLY	$1,856,446	2.6%
NVO	$1,473,666	2.1%

IR 2026 Theme 06 (GLP-1) + Gap Analysis (healthcare disruption). LLY: tirzepatide (Zepbound) capturing obesity + diabetes — next-gen retatrutide in pipeline. NVO: semaglutide + oral pill approval in progress. OSCR: health insurance disruption via value-based care — at 8.2% this is still heavy with regulatory headwinds. Consider trimming OSCR and redeploying to reinforce LLY/NVO.

💵

Fixed Income & Cash

$12,903,234 · 18.2% of portfolio

FIXEDINCOME	$12,701,700	17.9%
CASH_DEPOSIT	$201,534	0.3%

Ballast and dry powder at 18.2%. At portfolio scale ($70M), this represents ~$13M earning 4–5% in fixed income. IR 2026 recommends 5–10% cash/gold. Opportunity cost is significant — redeploy into missing HIGH conviction themes (Nuclear, Data Centers, AI Infra) over 6–12 months as positions are identified.

📊

Managed Portfolio

$13,416,616 · 18.9% of portfolio

MANAGEDPORTFOLIO

$13,416,616

18.9%

Externally managed diversified portfolio. Verify that managed portfolio's internal allocations don't duplicate or offset the IR 2026 theme gaps — particularly Nuclear, Defense, and Data Centers which are missing from the self-directed allocation. If managed portfolio already covers these, gap severity is reduced.

🔀

Alternative Assets

$154,689 · 0.2% of portfolio

ALTERNATIVE

$154,689

0.2%

Alternative asset exposure (private equity, hedge funds, real assets). Minimal allocation at current levels.

🔍 IR 2026 Framework — Theme Coverage

How the DiLuPa self-directed equity allocation maps to the 12 Investment Research 2026 themes · Updated Feb 18, 2026

IR 2026 Theme	Conviction	DiLuPa Exposure	Coverage
🤖 AI Infrastructure & Compute	HIGH	NVDA (1.5%) — missing MU, SNDK, CRWV, AVGO	⚠ UNDERWEIGHT
✨ AI-Driven Consumer Innovation	MED	DUOL (3.3%), ZETA (2.8%), SHOP (2.2%), META (1.5%)	✅ COVERED
📱 End-Game AI / On-Device	HIGH	AAPL (1.4%), GOOGL (1.4%) — below 10–15% target	⚠ UNDERWEIGHT
⚡ Nuclear Renaissance	HIGH	0% — no CEG, CCJ, SMR, NLR/URA ETFs	❌ MISSING (HIGH)
🌎 LatAm Digital Finance	HIGH	NU (29.3%), MELI (2.6%) — extreme NU concentration	⚠ OVERCONCENT.
💊 Healthcare / GLP-1 Revolution	MED	LLY (2.6%), NVO (2.1%), OSCR (8.2%)	✅ COVERED
🔩 Copper Supercycle	MED	0% — no FCX, SCCO, COPX ETF, CPER ETF	❌ MISSING (MED)
🔥 Natural Gas / Energy Transition	MED	0% — no EQT, CTRA, LNG, FCG ETF	❌ MISSING (MED)
🛡 Defense & Geopolitics	MID	0% — no LMT, RTX	❌ MISSING (MID)
₿ Crypto & Digital Assets	MED	0% — no COIN, BTC ETF, ETHA	❌ MISSING (MED)
🎬 Streaming & Content	LOW	NFLX (2.7%) — LOW conviction per IR 2026	⚠ LOW PRIORITY
🏗 Data Centers & Infrastructure	HIGH	0% — no VRT, EQIX, DLR	❌ MISSING (HIGH)

⚠️ Portfolio Gaps vs IR 2026 Framework

NU concentration (29.3%) — risk management priority #1 — Single largest position at nearly 30% is extreme. Any regulatory shock, Brazil FX depreciation, or credit cycle in LatAm could impact materially. IR 2026 caps LatAm at 4–7%. Trim NU on strength above $19; redeploy proceeds into missing HIGH conviction themes. Target: reduce to 20–22% initially, then 15% over 12 months.
🏗 Data Centers — VRT / EQIX / DLR (0%) — HIGH conviction gap — The physical AI layer is entirely absent. At $70M portfolio scale, this is a $7–10M sizing gap at IR 2026 target allocation. Vertiv (VRT) is the picks-and-shovels for every GPU cluster; liquid cooling TAM expanding 2.5x by 2028. EQIX and DLR have multi-year locked-in leasing backlogs.
⚡ Nuclear (0%) — HIGH conviction gap — AI data centers require 24/7 carbon-free baseload power. Nuclear is the only solution. CEG (Constellation) signed Microsoft Three Mile Island deal — the clearest hyperscaler demand signal. CCJ (uranium miner) benefits from long-term contracted pricing at $80–100/lb. NLR and URA ETFs (IR 2026 recommended) provide diversified nuclear/uranium exposure at scale.
🔩 Copper (0%) — MED conviction gap — Zero commodity exposure despite electrification mega-trend. Copper demand from AI data centers, EVs, and grid infrastructure creating structural supply deficit by 2030. COPX ETF (diversified miners basket) and CPER ETF (direct copper price exposure) are IR 2026 recommended vehicles — appropriate for portfolio scale without single-mine risk.
🔥 Natural Gas (0%) — MED conviction gap — AI power demand creates structural floor for gas prices. LNG (Cheniere) is the highest-quality LNG exporter with $7–8B contracted annual EBITDA. EQT/CTRA for domestic gas production. FCG ETF (IR 2026 recommended) for diversified exposure. At portfolio scale, a 2–3% allocation = $1.4–2.1M.
₿ Crypto (0%) — MED conviction gap — Bitcoin ETF institutional flows (BlackRock IBIT doing $500M+ daily inflows) provide a structural demand floor. COIN as regulated exchange earns fees regardless of which crypto asset wins. At portfolio scale, even a 2% allocation provides meaningful digital asset diversification.
🤖 AI Infra incomplete — only NVDA (1.5%) — NVDA is correctly held but per IR 2026, the AI Infrastructure theme explicitly includes MU (HBM3e memory — every Blackwell GPU requires it), SNDK (NAND storage for AI training data), and CRWV (pure-play AI GPU cloud). These are the three picks-and-shovels plays beyond NVDA.
📱 On-Device underweight (2.8% vs 10–15% target) — AAPL and GOOGL are HIGH conviction per IR 2026 — the device+OS layer owns the personal AI interface. At $70M portfolio, target is $7–10M in these two names; currently at $2M. Significant opportunity to add on any pullback.
🛡 Defense — LMT / RTX (0%) — MID conviction per IR 2026 — Global rearmament cycle is structural and bipartisan. NATO 2%+ GDP spending mandate adds $200B+ incremental annually. At MID conviction, a 3–5% allocation ($2–3.5M) is appropriate — provides geopolitical hedge with 2%+ dividend yield. Not urgent but entirely absent.

✅ Priority Actions to Align with IR 2026

Trim NU Sell $6–8M · Target 20–22% Sell > $18.50

RISK MANAGEMENT PRIORITY #1. Reduce from 29.3% → 20–22% over 3–6 months. Take profits at current levels. Proceeds fund the six missing IR 2026 themes below. Single-stock concentration at this level is not consistent with a well-constructed long-term portfolio regardless of conviction.

VRT $3M · ~4.2% Buy ≤ $260

HIGH PRIORITY — IR 2026 Theme 12 (HIGH conviction). Vertiv is the dominant power and thermal management provider for AI data centers. Liquid cooling TAM expanding from $14B → $35B by 2028 as GPU density scales with Blackwell/Rubin. Largest structural gap in this portfolio given HIGH conviction and 0% exposure.

NLR + CEG + CCJ NLR $2M + CEG $1.5M + CCJ $1M · ~6.4% NLR any / CEG ≤ $310 / CCJ ≤ $120

HIGH PRIORITY — IR 2026 Theme 04 (HIGH conviction). Nuclear is the only 24/7 carbon-free power for AI data centers. NLR ETF: diversified nuclear energy basket (IR 2026 recommended) — appropriate for portfolio scale. CEG: landmark hyperscaler PPA deal signed; URA ETF (uranium miners) also recommended as alternative to CCJ for diversification.

AAPL + GOOGL Add $3M each → target 4% each AAPL ≤ $270 / GOOGL ≤ $310

HIGH PRIORITY — IR 2026 Theme 03 (HIGH conviction). On-Device AI at 2.8% is severely underweight vs 10–15% IR 2026 target. At $70M portfolio, target 4–5% each = $3–3.5M per name. AAPL: Apple Intelligence iPhone supercycle, $100B+ Services annualized. GOOGL: Gemini, Search defense, Cloud 25%+ growth.

COPX + CPER COPX $1.5M + CPER $1M · ~3.5% COPX ≤ $42 / CPER any

MED PRIORITY — IR 2026 Theme 07 (MED conviction). COPX ETF (diversified copper miners basket) and CPER ETF (direct copper price exposure) are the two IR 2026 recommended ETFs for the copper supercycle. Appropriate for large portfolio — no single-mine execution risk. Copper structural deficit of 4–8M tonnes projected by 2030.

LNG + FCG ETF LNG $1.5M + FCG $1M · ~3.5% LNG ≤ $230 / FCG any

MED PRIORITY — IR 2026 Theme 08 (MED conviction). LNG (Cheniere Energy): $7–8B annual contracted EBITDA floor, largest US LNG exporter. FCG ETF (IR 2026 recommended): diversified natural gas exposure across producers and pipelines. Europe's structural LNG dependency from US is a multi-year demand driver.

COIN + IBIT COIN $1.5M + IBIT $1.5M · ~4.2% COIN ≤ $175 / IBIT any

MED PRIORITY — IR 2026 Theme 10 (MED conviction). COIN: dominant US-regulated crypto exchange; earns ETF custody fees + trading volume regardless of asset winner. IBIT (BlackRock Bitcoin ETF): cleanest institutional BTC exposure. US Strategic Bitcoin Reserve discussion provides potential sovereign demand layer. Bitcoin ETF inflows now $500M+ daily.

Trim OSCR Sell $2–3M · Target ~4% Sell > $15

REBALANCE. OSCR at 8.2% is heavy given healthcare regulatory headwinds. Trim to 4% ($2.8M) and redeploy into LLY (deepen the GLP-1 thesis properly) or into missing HIGH conviction themes. OSCR is MED conviction in the broader healthcare disruption space; LLY/NVO are the IR 2026 core GLP-1 picks.

📋 Thesis Breakdown

🔍 IR 2026 Framework — Theme Coverage

⚠️ Portfolio Gaps vs IR 2026 Framework

✅ Priority Actions to Align with IR 2026