Advisors adopting model portfolios with buffered ETFs and direct indexing must understand liquidity risk from NAV divergence and the operational friction of fragmented platforms.
Alpha Score of 60 reflects moderate overall profile with weak momentum, strong value, strong quality, moderate sentiment.
The wealth management industry is quietly reshaping how portfolios are built. On Zephyr's Adjusted for Risk podcast, Modelist co-founder Alex Graf walked through the trends advisors are demanding: buffered ETFs, covered call strategies, real assets, and direct indexing. Below the surface of these product shifts lies a set of risks that the wire summaries skip.
Graf described Modelist's approach as blending OCIO/TAMP-style investment support with a focus on letting advisors maintain their own brand. The mechanisms matter more than the buzzwords, and the risks are specific to each layer of the model.
A buffered ETF uses options to deliver a defined outcome. The typical structure: the fund buys a call spread to capture upside up to a cap, and sells a put spread to fund a downside buffer for a specific range of losses over a one-year outcome period. An investor might receive 100% protection on the first 10% of index losses, then full exposure beyond that. In exchange, the upside is capped at, say, 10% to 15%.
The naive read is that this product offers cheap downside insurance on equities. The better market read starts with the NAV/market-price divergence Graf flagged.
Key insight: A buffered ETF's protection only applies to the specific outcome period. Buying at a premium near the end of the period means the protection is already partially priced in.
These ETFs trade on exchanges like ordinary ETFs. Options-based structures can cause the market price to deviate from net asset value, especially during high volatility or near the end of an outcome period. An advisor who buys a buffered ETF at a significant premium to NAV is effectively paying for protection that may not materialize at the full rate. The bid-ask spread can widen sharply during market stress, making exit expensive. Graf emphasized that manager pedigree and liquidity vary by issuer. Not all providers have the same options-trading infrastructure, and secondary market liquidity shifts by the specific outcome period.
The practical implication: an advisor using buffered ETFs as a fixed-income replacement needs to track not just the buffer level but also the ETF's daily trading pattern relative to NAV. A running discount or premium that persists signals structural pricing risk.
Graf described a concept he calls "two layers of the same model." For accounts below roughly $250,000, the advisor uses an ETF-based model portfolio. For larger accounts, the advisor replaces those ETFs with direct indexing – owning the individual securities in the index. This allows tax-loss harvesting at the security level and the ability to exclude stocks for ESG or personal reasons.
The efficiency gain is clear: one strategic asset allocation serves multiple client tiers. The risk, however, is platform fragmentation. Most wealth management platforms offer either model marketplace functionality or direct indexing capabilities, not both in a seamless workflow. Advisors who use separate systems for each layer introduce operational friction: rebalancing delays, tax-lot tracking errors, and increased oversight costs.
Graf argued that the "holy grail" is an integrated engine that combines model portfolio construction with direct indexing tools. That integrated system is not yet widely available. Until it is, the two-layer approach carries execution risk that can erode the tax and customization advantages.
Direct indexing lets an advisor harvest losses on individual stocks that have declined, offsetting realized gains in the same portfolio. The mechanism is powerful: instead of being limited to an ETF's aggregate performance, the advisor can sell Microsoft after a 5% drop while keeping the rest of the S&P 500 exposure.
The concealed risk is tracking error and wash-sale violations. Over-optimization – harvesting losses too aggressively – can cause the portfolio to drift from the benchmark. An advisor who sells one stock and buys a close substitute must navigate the 30-day wash-sale rule. A poorly integrated system can generate wash sales across accounts owned by the same client. The benefit of security-level tax harvesting shrinks if the advisor spends more time managing the exceptions than the allocation.
A change in tax law that reduces the value of tax-loss harvesting would weaken the case entirely. The two-layer model depends on a favorable tax regime.
Graf sees artificial intelligence as a tailwind for independent advisors. The automation of rebalancing, tax optimization, and client communication frees advisors to focus on client relationships and financial planning. He pointed to three specific areas where AI is already adding value: portfolio rebalancing, tax-loss harvesting across multiple accounts, and generating personalized market commentary.
The risk is not that AI disrupts advisors. The risk is that advisors who outsource the entire investment process – including portfolio construction – become indistinguishable from robo-advisors. The brand differentiation that Graf emphasizes comes from the client experience and behavioral coaching, not from a black-box algorithm.
Risk to watch: Advisors who automate the full investment decision chain lose the trust-building narrative that justifies their fee.
Two sets of factors will determine whether the two-layer model becomes standard practice or remains a niche tool.
What would confirm the adoption:
What would weaken the adoption:
The conversation between Nauman and Graf reflects a structural shift. Advisors are moving toward customizable, tax-aware, outcome-oriented portfolios. The technology to deliver that at scale is evolving. Until the integrated engine arrives, the two-layer model demands active oversight of the risks embedded in each layer – from buffered ETF pricing to direct indexing operational drag.
Prepared with AlphaScala research tooling and grounded in primary market data: live prices, fundamentals, SEC filings, hedge-fund holdings, and insider activity. Each story is checked against AlphaScala publishing rules before release. Educational coverage, not personalized advice.