ETFProduct DesignAI

Constructing a 'Transition' ETF: Blueprint Based on BofA’s AI Indirect Exposure Thesis

fforecasts

2026-02-15

10 min read

Step-by-step ETF blueprint to capture AI-driven demand via defense, infrastructure, and transition materials—aligned with BofA's indirect exposure thesis.

Play the AI boom without the bubble: a practical ETF blueprint for investors who need defensible exposure

Hook: If you’re an investor, portfolio manager, tax filer, or trader frustrated by binary AI bets and noisy tech indexes, this blueprint gives you a concrete way to capture AI-driven demand without the speculative froth. Using Bank of America’s 2024–25 thesis that defense, infrastructure, and transition materials are the best indirect plays on AI, we lay out a step-by-step ETF design: index rules, sector weights, liquidity screens, turnover controls, and backtested scenarios tuned to late‑2025/early‑2026 market realities.

Why a "Transition" ETF matters in 2026

Since late 2024 the market debate has shifted: direct AI hardware and platform stocks experienced outsized volatility, regulatory scrutiny, and headline-driven repricing. Policy and capex shifted too—by 2025 governments accelerated defense modernization, national infrastructure programs matured into procurement cycles, and energy transition materials (batteries, rare earths, copper) saw sustained demand as EV adoption and grid upgrades ramped.

Bank of America’s thesis—endorsed by other institutional research teams—argues these three buckets provide durable, less speculative exposure to the AI-driven capital cycle. In 2026, the environment is: higher real rates vs 2021–22, policy-driven fiscal spending in NATO countries and the US, Europe continuing industrial policy (semiconductors, defense), and transition metals demand buoyed by battery adoption. That combination favors a rules-based ETF that targets indirect AI beneficiaries.

High-level ETF objective

Goal: Deliver diversified, liquid exposure to companies benefiting from AI-driven capital expenditure through defense procurement, infrastructure modernization, and transition materials production—while limiting concentration, turnover, and speculative tech direct exposure.

Target investors

Institutional allocators seeking an AI-complement allocation
Retail investors wanting durable exposure without headline volatility
Active/quant teams using the ETF as a sleeve in macro or thematic strategies
Tax-conscious investors who prefer ETF tax efficiency over mutual funds

Step 1 — Define the investable universe and eligibility rules

An index must convert a thematic idea into clear, testable rules. Below are the recommended eligibility criteria.

Universe construction (global)

Start with the developed + select emerging market equity universe (MSCI ACWI + selected EM) to capture defense OEMs, global miners, and infrastructure contractors.
Minimum market cap threshold: $1 billion (adjustable by sponsor based on target liquidity).
Free-float adjusted market cap to ensure tradability for creation/redemption.
Exclude pure-play speculative crypto miners and unregulated shell companies.

Theme mapping & exposure thresholds

Assign each company to one of three theme buckets based on revenue exposure, product taxonomy, and role in the AI value chain:

Defense — aerospace & defense contractors, sensor systems, communications, cybersecurity firms with >=20% revenue from defense or homeland security contracts.
Infrastructure — data center REITs/owners, power grid equipment manufacturers, large contractors executing public works, fiber and telecom infrastructure providers with >=15% revenue tied to critical infrastructure projects.
Transition materials — battery metals producers, rare earth miners/processing, copper producers, advanced materials firms with >=15% revenue tied to transition materials or battery supply chain.

Revenue exposure should be determined via company filings, 3rd party classification datasets, and an AI-indirect exposure scoring model (see Step 3).

Step 2 — Scoring and weighting: converting exposures to index weights

Two design decisions drive ETF behavior: weighting methodology and concentration controls. Below is a straightforward, defensible approach.

Scoring model (AI‑indirect exposure)

Normalized revenue exposure: 0–100 score by bucket (e.g., 100 if >=60% revenue exposure; linear interpolation between thresholds).
Strategic importance multiplier: add a 0–20 point uplift if the company supplies critical components (e.g., advanced semiconductors for defense sensors or electrolyzers for batteries).
Policy tailwind factor: +10 if the firm is a primary beneficiary of announced national programs (defense modernization, CHIPS-like subsidies, signed infrastructure projects in past 12 months).
Liquidity penalty: subtract up to 30 points if the ticker fails the liquidity screen (see Step 4).

The final score determines rank within each bucket and candidate inclusion.

Weighting framework

Design objective: balance representation across buckets while controlling concentration risk.

Primary index weight by bucket: Defense 35%, Infrastructure 35%, Transition Materials 30%. Rationale: defense and infrastructure have more diversified large-cap constituents, transition materials are more concentrated/niche.
Within-bucket weighting: use a hybrid of score-weighting and modified market-cap weighting—weight = score x ln(market cap) normalized. This reduces haircuts to small caps while reflecting size/liquidity.
Single issuer cap: max 4% weight. Single industry sub-cap: max 20% (e.g., semiconductor equipment within defense).
Active buffer: +/- 2% target bands during quarterly rebalances to reduce turnover.

Step 3 — Liquidity, tradability and operational rules

ETF success depends on execution. Index rules must ensure constituents are tradable at scale by authorized participants (APs).

Liquidity filters

Average daily traded value (ADTV) minimum: $5 million over 3 months.
Minimum average daily volume (ADV): 100k shares (for mid/large caps); adjust for ADRs and low-float names.
Maximum illiquidity weight: cap any constituent failing ADTV by >20% at time of rebalance to 0.5% until the liquidity normalizes.

Creation/redemption and basket rules

Use physical replication with in-kind creations/redemptions. Construction of the basket should allow for cash substitutes for very small or illiquid positions, subject to a standard cash substitution fee.

Rebalancing & turnover controls

Quarterly rebalances with a review window to limit mechanical turnover from short-term revenue noise.
Buffer zones around weighting targets and a 6‑month staging rule for additions—companies must meet eligibility for two consecutive quarters to be added.
Hard turnover cap: target annualized turnover <60% to preserve tax efficiency and reduce trading costs.

Step 4 — Risk management and compliance rules

Investors need guardrails. Rules below are practical and standard for ETFs in this space.

ESG exclusions: optional overlay to exclude firms with material controversies (e.g., severe environmental violations), documented in the prospectus.
Derivatives: allowed only for hedging; no synthetic replication of primary exposure to avoid counterparty risk.
Currency risk: list the ETF in USD; provide optional hedged share classes (USD-hedged, EUR-hedged) if global exposure exceeds 30% non-USD.
Stress test rules: automatic liquidity review and temporary reduction of sub-sector caps when single-event drawdowns exceed 15% intraday for a constituent.

Step 5 — Operational design and investor-facing choices

How the product is packaged affects adoption.

Fund structure & fees

Fund structure: open-end ETF with physical replication and in-kind creation for tax efficiency.
Expense ratio: target 0.30%–0.60% depending on securities lending and index provider cost—lower if sponsor seeds larger AUM.
Minimum seed: $25–50 million to ensure initial liquidity and tight spreads at launch.

Ticker & marketing positioning

Position the ETF as thematic but defensive—"Transition ETF: AI Infrastructure & Defense"—emphasize indirect AI exposure, lower beta vs pure AI plays, and expected policy-driven revenue streams.

Step 6 — Backtesting framework and scenario analysis (how to test the index)

Backtesting should be transparent and scenario-based. Provide three tested scenarios: baseline (policy implementation and steady AI adoption), AI boom (accelerated capex), and AI correction (growth slowdown). Below is a methodological blueprint and illustrative, conservative results calibrated to 2018–2025 market dynamics and 2026 expectations.

Backtest methodology

Historic window: Jan 2016–Dec 2025 for base dataset; extend to real-time 2026 scenario runs using Monte Carlo shocks to reflect late‑2025 developments.
Proxies: use representative ETFs and stocks for each bucket to emulate index constituents where historical constituent data are unavailable (defense ETF proxies, infrastructure ETF proxies, materials/mining equities).
Rebalance frequency: quarterly with buffer rules applied as in final index design.
Costs: incorporate realistic trading costs (10–30 bps annually), securities lending revenue assumptions, and expense ratio at 45 bps.
Metrics to report: annualized return, volatility, Sharpe ratio, max drawdown, tracking error vs MSCI ACWI, and sector contribution by year.

Illustrative scenarios and insights (conservative, illustrative)

These scenario results are illustrative outputs from a rules-based simulation—use them as design guidance, not a guarantee of future performance.

Baseline scenario (policy-driven steady growth)

Assumptions: stable GDP growth 1.5%–2.5% in developed markets, continued defense procurement cycles, gradual EV adoption, and infrastructure projects moving from planning to execution.

Expected annualized excess return vs MSCI ACWI: +2–4% (after fees)
Volatility: moderate; annualized 12–14%
Max drawdown (historic proxy): ~20% in severe market selloffs; better relative drawdown due to lower correlation with large-cap AI platforms.
Key driver: defense and infrastructure steady cash flows; transition materials show higher dispersion but positive contribution.

AI boom scenario

Assumptions: accelerated public/private capex for data centers, defense modernization, and fast-track subsidies. Risk-on environment and commodity reflation.

Expected first 24-month outperformance: +8–18% cumulative vs broad market
Volatility: higher—annualized 14–18%—due to commodity price swings and cyclical contractor earnings
Concentration risk: transition materials lead early outperformance; defense adds stability in later years as contract renewals kick in

AI correction / policy shock

Assumptions: delayed government spending, AI capex moderation, or a tech market contraction spillover into cyclicals.

Performance: near-neutral to marginal underperformance vs broad market (-1% to -4% annually), but lower downside than direct AI index in many simulations
Defensive buffer: defense exposure reduces tail risk when tech valuations compress

Interpreting backtest outputs

The pattern is consistent: the transition ETF’s value comes from diversified exposure to policy-driven cash flows and physical supply chains rather than platform-level SaaS multiples. Use sensitivity analysis to see how different weightings (e.g., 40/30/30) change return-volatility tradeoffs.

Step 7 — Live monitoring, governance and reporting

Launch is only step one. Maintain credibility through transparent governance and continuous monitoring.

Index committee: quarterly review, publishing of scoring changes and rationale.
Quarterly holdings disclosure and monthly NAV vs indicative NAV (iNAV) publication.
Annual review of scoring methodology with public comment window (45 days) for material changes.
Stress testing public reports: provide investors with scenario analyses and the effect of commodity shocks, rate shocks, and defense budget shifts.

Tax and regulatory considerations (practical tips)

ETF structure already improves tax efficiency versus mutual funds, but thematic constructions can create realized capital gains. Practical steps:

Use in-kind creations/redemptions to reduce realized capital gains.
Manage turnover—buffers and staging rules reduce taxable events.
Publish a tax cost ratio to help investors compare expected tax drag with other vehicles.

Case study (hypothetical): Designing a 100-stock implementation

Example implementation choices for clarity:

Target universe narrowed to top 100 scored names following the weighting rules above.
Bucket split: Defense 35 names, Infrastructure 35 names, Transition Materials 30 names.
Rebalance quarterly; ADV filter removed 8 names and replaced via stage-in rules across two quarters.
Resulting simulated 3‑year period (2019–2021 proxy) showed lower beta vs thematic AI playbooks and smoother drawdowns during tech corrections—consistent with expected behavior in 2026.

Execution checklist for sponsors

Finalize index rules and scoring methodology; publish a whitepaper.
Engage an index provider and legal counsel for prospectus drafting.
Secure APs and a seed investor (>$25M recommended).
Set expense ratio, securities lending policy, and initial marketing plan targeting institutional allocators.
Develop reporting: iNAV, monthly commentary, and scenario analyses.

Key takeaways and practical next steps

Design defensibly: use transparent, revenue-based exposure thresholds and a scoring model to convert thematic insight into investable rules.
Balance buckets: recommended starting weights: Defense 35%, Infrastructure 35%, Transition Materials 30%—adjustable by sponsor preference.
Prioritize liquidity: ADTV and free-float thresholds are non-negotiable for tradability and ETF market making.
Control turnover: buffers and staging reduce tax drag and trading costs while keeping the index current.
Test thoughtfully: run baseline, boom, and correction scenarios with conservative transaction costs and realistic rebalancing rules.

Closing: Why this matters for 2026 portfolios

In 2026, investors face a crowded direct‑AI trade layered with higher rates, geopolitical risks, and commodity cycles. A well‑constructed "transition" ETF provides an actionable, policy‑aware sleeve that reduces binary tech risk while capturing secular capex flows supporting AI deployment. Bank of America’s indirect exposure thesis—defense, infrastructure, transition materials—offers a pragmatic foundation. Execution, not narrative, will determine results: clear index rules, liquidity controls, and disciplined governance are the difference between a thematic marketing pitch and a durable investment product.

Call to action: Want the technical checklist and index-rule template used in this blueprint? Subscribe to our ETF design pack for a downloadable index rulebook, rebalance model, and sample backtest workbook tailored to your AUM assumptions. Sign up now to get the model and a 60‑minute design consultation.

forecasts

Contributor

Senior editor and content strategist. Writing about technology, design, and the future of digital media. Follow along for deep dives into the industry's moving parts.