Week 48: Structured Products — Buffer ETFs, Principal-Protected Notes, and Replicating Them DIY

---

1. Why This Is Important

Wall Street's most reliable business is repackaging things you could buy yourself, charging a fee for the wrapping, and selling the result as a "solution." Structured products are the cleanest 2026 example. Buffer ETFs and principal-protected notes have grown from a rounding error in 2018 to over $50 billion in defined-outcome ETF assets and several hundred billion in retail-distributed structured notes. The marketing is irresistible — "S&P 500 upside up to a 18% cap, with a 15% downside buffer, over a 12-month period" — and the mechanics, once you see them, are a strict subset of what Weeks 25 through 30 already taught you.

You need to understand these products for four reasons:

They are aggressively pushed at retail. Innovator (BUFR

series), First Trust Cboe Vest (the FT series), and Allianz are running television and brokerage-platform campaigns explicitly framed at investors who lived through 2008, 2020, or 2022 and now want "protection." Your advisor, your 401(k) target-date overlay, your in-laws' brokerage statement — you will encounter these. The fastest way to evaluate them is to know the option positions inside.

*The mechanics are exactly* the option structures from Weeks

25–30.** Every buffer ETF is a long zero-coupon bond plus a bull-call-spread plus a short put — three pieces, all of which we priced together in Week 29 with Black–Scholes. Once you see the decomposition, the "magic" of the buffer disappears. Alpha is rare, and a 79bps fee for arithmetic you can reproduce in your own brokerage account is not alpha — it is a convenience tax.

The fees are far higher than the prospectus number. A buffer

ETF's headline expense ratio is around 0.79%. The hidden costs — bid-ask on the underlying SPX options, the spread between the cap the issuer offers and the cap the structure could mathematically support, the timing of the quarterly reset — bring the all-in cost to roughly 1.5%. A retail principal-protected note from a bank can embed 2% to 4% of issuance fees inside the cap, term, and credit-spread haircut. The DIY equivalent, executed once a year on 1-year SPX options, costs around 5 basis points of slippage per leg, or ~20bps total. That is a 7-to-30× cost ratio.

The barbell and the tax wrapper still apply. A barbell sits

on the base of the portfolio (boring compounder) plus a small convex sleeve. A buffer ETF tries to sell you the whole portfolio: capped upside, cushioned downside, no convex sleeve. On the tax side, SPX options inside a DIY buffer are 1256 contracts — 60/40 long/short tax — while a buffer ETF is ordinary CG and the structured note is paid out as ordinary income at maturity. The post-tax difference compounds.

This lesson teaches the mechanics, the DIY replication, the realistic fee comparison, and the conditions under which the packaged product is actually worth its cost.

---

2. What You Need to Know

2.1 What a Buffer ETF Actually Owns

A 1-year buffer ETF on the S&P 500 with a 15% downside buffer and an 18% upside cap is, on the inception date, a portfolio of three positions on the underlying index (SPX or SPY):

Long a 1-year zero-coupon Treasury (or a synthetic

equivalent) that returns par at maturity. This guarantees the principal delivery before any option payoff is computed.

Long a 1-year bull-call-spread, struck at-the-money on the

bottom and 18% out-of-the-money on the top. This is the participation: you receive 1-for-1 upside up to the cap.

Short a 1-year put struck 15% OTM. This is what finances

the spread. The premium you collect on the short put is exactly what pays for the long call leg of the spread, which is why the cap exists in the first place.

The four legs net out to a payoff that, at expiry, looks like:

• Index up by more than 18%: you receive 18%, capped.
• Index up by 0% to 18%: you receive the index's return, 1-for-1.
• Index down by 0% to 15%: you receive 0% (the buffer).
• Index down by more than 15%: you receive (index_return + 15%) —

the buffer absorbs the first 15%, you take the rest.

The shape is identical to what you would build by hand from four SPX options. That is the central observation.

2.2 The DIY Replication on 1-Year SPX Options

With SPY at $500 in April 2026, a 12-month buffer with the same 15% buffer and 18% cap is replicated on SPX (10× the SPY notional, so strikes are 5000-anchored). Per $50,000 of notional:

Leg	Action	Strike	DTE	BSM Price
1	Long zero-coupon T-bill notional	—	365d	95.79% par
2	Long SPX call	5000 (ATM)	365d	$310
3	Short SPX call	5900 (+18% cap)	365d	$48
4	Short SPX put	4250 (-15% buffer)	365d	$112

Net out:

• T-bill cost: $47,895 of $50,000 notional (1y at 4.3%).
• Bull-call-spread debit: $310 - $48 = $262 per share-equivalent.
• Short-put credit: $112 per share-equivalent.
• Net spread cost: $262 - $112 = $150 per share-equivalent ≈ 3.0%

of notional.

• Add T-bill drag: total = ~$2,255 of explicit cost on $50,000 =

4.5% of notional consumed by the structure, which is exactly what makes the cap 18% rather than the 21–22% it would be at zero cost.

Total slippage executing all four legs at the mid-point ± typical retail spreads on 1y SPX options: roughly 5bps per leg × 4 legs = ~20bps round-trip, plus ~1–2 commissions per leg. Call it 25bps.

2.3 The Fee Comparison Over 5 Years

Buffer ETFs in the Innovator BUFR series and First Trust Cboe Vest series charge expense ratios in the 0.79%–0.85% range. Add typical secondary-market bid-ask of ~12bps and you are at roughly 91bps per year of explicit cost. Over a 5-year horizon, on a $100,000 starting balance, the wealth gap vs. the DIY equivalent is not small.

The gap is ~$4,900 on $100k — roughly 3.7% of the starting balance — purely from the fee differential. Scale to $500k and that is $24,500 paid to a fund company for option trades you could place yourself.

The honest counter-argument: the DIY version requires a Level 3 options account, four execution legs per year on roll dates, and discipline through a year-long expiry cycle. The fund company is selling the discipline plus the operational lift. Whether that is worth 80bps/yr is the real question — and for most retail investors who can run Week 30's spread builder, the answer is no.

2.4 Principal-Protected Notes and Bank Credit Risk

Principal-protected notes (PPNs) are bank-issued debt instruments, not ETFs. A typical 5-year structured note from a major US bank in 2026 might promise: "100% principal back at maturity; participation in the S&P 500 with a 30% cap." On the surface this looks like a buffer ETF with a longer term and no buffer-floor erosion.

What is buried in the prospectus:

• Issuer credit risk. When Lehman Brothers collapsed in 2008,

investors holding Lehman-issued PPNs received pennies on the dollar in bankruptcy — even on notes that explicitly said "principal protected." The protection is unsecured senior debt of the bank. Stick to US-listed claims you can verify. ETFs are bankruptcy-remote; bank notes are not.

• Embedded fees of 200–400 bps. The bank prices the note such

that the all-in expected cost to the investor is 2–4% of notional, amortised over the 5-year term. This shows up as a lower cap, a longer term, or a wider underlying participation discount.

• Illiquidity. PPNs trade rarely on the secondary market. If

you need cash mid-life, the dealer who issued the note will quote a price that bakes in another 1–3% of haircut.

• Tax treatment. Principal-protected notes are typically taxed

as ordinary income on the contingent payment at maturity (CPDI — contingent payment debt instrument rules), not capital gains. At a 32% federal bracket plus 5% state, this is a 19-percentage- point hit relative to qualified LTCG.

The DIY equivalent for the 5-year note above: a 5-year zero-coupon Treasury for principal protection, plus 5-year SPX call options for the upside. Both pieces are bankruptcy-remote (Treasury is the sovereign, options are CCP-cleared). All-in cost: a tighter cap, no issuer credit risk, no illiquidity, and 1256-contract tax treatment.

Horace's view — I do not buy structured notes. Bank credit risk is the unhedgeable risk in a US-equity-centric global portfolio where the tail-hedge sleeve is supposed to reduce exotic counterparty exposure, not add a new one — and the tax envelope on a CPDI-classified note is meaningfully worse than the DIY 1256 alternative. The packaged product solves a problem retail investors do not have. If you cannot place the four legs yourself, hold the index; if you can, build the structure on Treasuries plus listed options and keep the bank's profit margin in your own pocket.

The barbell-compatible DIY alternative is concrete and small. For a 5-year horizon: buy a 5-year zero-coupon Treasury sized to your principal-protection target — that is the safety end of the barbell expressing itself, with the issuer being the US Treasury rather than a single bank balance sheet — then buy a 5-year SPX call (or a small ladder of LEAPs rolled forward) sized to the participation you wanted. Both legs are bankruptcy-remote, both are publicly priced, and the call leg gets 1256 60/40 treatment instead of CPDI ordinary income. The shape is exactly what the structured note was trying to sell you, with the wrapper cost and the credit risk both removed.

2.5 The Three Conditions Under Which the Packaged Product Wins

There are situations where buying the wrapper is rational:

Account constraints prevent options. Many 401(k) plans and

employer brokerage accounts do not permit options trading at all. In a non-options account, the buffer ETF is the only way to get the buffered exposure and the 80bps fee is unavoidable.

Discipline failure mode. If you know yourself well enough to

suspect you would not actually execute the four-leg roll on the anniversary date — or worse, would close one leg early in a panic — the fund company is selling you the commitment to the structure. A buffer ETF cannot be "broken into" mid-period the way a self-managed structure can.

Notional below ~$50k. SPX options have a $100 multiplier.

Below ~$50k of notional you are forced into SPY options, which are more expensive, less tax-efficient (no 1256 treatment until they are SPX-cash-settled), and require more contracts. The fund company's pooling lets sub-$50k accounts access the SPX-priced structure indirectly.

Outside these three conditions, the DIY version is strictly better on cost and tax. Alpha is rare, and 80bps for an arithmetic identity is not alpha.

2.6 The Barbell Read on Buffered Products

The barbell view frames the portfolio as a base (boring, compounding) plus a small convex sleeve. Buffered ETFs collapse both legs into a single flat-payoff structure: capped upside, buffered downside, convex sleeve sold to finance the structure (the short put leg). At the portfolio level this is anti-barbell: you have given up the convex left tail (because the short put leg loses badly past the buffer) and you have given up the convex right tail (because of the cap). What remains is the middle of the distribution.

For most retail compounders this is the wrong trade. The middle of the distribution is what you already get from index funds; the tails are where the volatility-tail-wags-the-dog effect compounding shows up. Paying 80bps to delete both tails is not insurance — it is a tactical bet that the index will spend the year between -15% and +18%. The 1928–2024 base rate for that range is roughly 55%.

---

3. Common Misconceptions

"Buffer ETFs protect my principal." No — they buffer the

first 15% of a drawdown over the outcome period (typically 1 year). Below the buffer, you take the full loss minus the buffer amount. In a 2008 scenario (-37%), a 15%-buffered product loses -22%, not zero.

**"The 18% cap is generous because the market rarely returns

that much."** The 1928–2024 S&P 500 returned more than 18% in roughly 36% of calendar years. The cap binds more than 1 in 3 years.

"Principal-protected notes are safer than ETFs." They are

less safe in one critical dimension: bank credit risk. ETFs are bankruptcy-remote trusts; PPNs are unsecured debt of the issuer. Lehman PPNs returned ~9 cents on the dollar in 2008.

"The fee is just the expense ratio." The all-in cost of a

buffer ETF includes the spread between the cap offered and the theoretically achievable cap (typically 50–100bps), the bid-ask on the underlying SPX options the fund executes (10–30bps), plus the headline expense ratio (~80bps). Total is closer to 1.4%.

"Buffer ETFs reset annually so I always have protection."

The reset only protects against losses in the new outcome period. If the market drops 10% in month 11 and recovers 5% in month 12, your current buffer absorbs nothing because the buffer is measured from the period start, not from your peak.

"DIY replication is too complex for retail." Four legs of

SPX options, executed once per year on the anniversary date, with strikes computable from a Black–Scholes calculator. Week 30's spread builder already showed this is Level 3 retail homework.

"The buffer means I can sleep through a bear market." Until

the market drops more than 15%, after which every percent down is 1-for-1 yours. A -30% year on a 15-buffer product is -15%, which is not sleeping-well territory.

**"Banks would not sell something they expect to lose money for

the client."** Banks price PPNs with 2–4% of profit baked in. They expect to make that profit on average. The client pays it either as opportunity cost (a worse cap), liquidity cost (no secondary market), or tax cost (CPDI rules).

"Defined-outcome means defined-good-outcome." The outcome is

defined; whether it is good depends on what the index does. The structure is still a directional bet — narrower than a long-only equity bet, but a bet nonetheless.

"If I do not need the buffer, I can sell the ETF mid-period."

You can, but you receive the secondary-market price, which reflects the current mark-to-market of the underlying option structure — not the "buffer" or "cap" advertised. Mid-period holders have no buffer; they have an opaque option exposure.

---

4. Q&A Section

Q1: Why does the cap exist in the first place? Because the short-put leg is what finances the bull-call-spread leg. The premium implicitly received on the short put is fixed by implied volatility on the buffer strike. That premium is enough to pay for some of the upside spread, but not all of it — so the upper strike of the spread is set at the level where the math balances. Higher IV → wider achievable spread → higher cap.

Q2: Can I get a buffer larger than 15%? Yes — there are 20% and 30% buffer products in the Innovator series. The cost is a lower cap (typically 10–13% on a 30-buffer product). The deeper buffer requires selling a more-OTM put (less premium) plus a more-OTM call (less premium received from the sold call leg of the upper bound), so the achievable cap shrinks roughly linearly.

**Q3: How does this interact with the 1256 tax treatment from Week 39?** SPX options are Section 1256 contracts and receive 60/40 long-term/ short-term treatment regardless of holding period. A DIY buffer implemented on SPX options inherits this — the realized P&L at expiry or close is taxed at the blended 1256 rate (~21.8% at a 32% ordinary bracket). A buffer ETF holds the options inside the fund, so you get fund-level tax: ordinary CG on distributions and your own LTCG/STCG on the ETF shares depending on holding period. PPNs are taxed as ordinary income under CPDI rules. Ranking from best to worst at a 32% bracket: DIY SPX (~21.8%) > buffer ETF (LTCG 15–20% on ETF shares, but you pay annual ordinary CG distributions on internal P&L) > PPN (32%+).

Q4: What happens if I buy a buffer ETF mid-period? You inherit the remaining buffer and cap from the period-start strikes. If the market is already +10% into the period and you buy in, your effective upside is now the cap minus 10% (i.e. ~8% if cap was 18%) and your effective buffer is the original 15% buffer relative to the period start, which means you are only protected if the market drops more than 10% from where you bought. The labelled "15% buffer" is no longer your buffer.

**Q5: How do I size the four DIY legs to match $50k of buffered exposure?** For SPX-anchored: $50k / SPX_level / multiplier = number of SPX spreads. With SPX at 5000, that is $50k / 5000 / $100 = 0.10 contracts per share — so for cleanest sizing, use $100k notional = 2 contracts of each leg. For SPY: $50k / SPY_price / 100 = $50k / $500 / 100 = 1 contract of each leg, which works at a single 1-contract scale.

**Q6: Why not just hold SPY plus a long put for the same protection?** A long put has no cap on the upside but costs 4–5% of notional per year (for ATM 1y puts). The buffer structure trades the ~5%/yr put cost for the cap — capping upside at 18% in exchange for not paying that 5%/yr drag. It is a different tradeoff: barbell (SPY + put) keeps the convex tails at the cost of carry; buffer collapses the distribution at the cost of tails. If your portfolio already has a convex sleeve (Week 47's tail hedge), you do not need the buffer structure on top of it.

Q7: What is the worst-case outcome on a buffer ETF? The market drops to zero. You get max(index_return + buffer, -100%). On a 15% buffer, you lose 85% of capital. The buffer never guarantees principal — it only absorbs the first 15% of loss.

Q8: Do PPNs ever make sense? Niche cases: an investor with very specific tax circumstances (e.g., trust structures where the deferral of CPDI matches a planned liquidity event), or institutional accounts that need a CUSIP-based debt-instrument wrapper for compliance reasons. For a typical retail investor in a taxable account, the answer is essentially never — ETFs and DIY beat them on every dimension that matters.

**Q9: How do I evaluate a new structured product I have not seen before?** Decompose it into its option legs. Every structured product is a linear combination of zero-coupon bonds plus options on the underlying. Price the legs in a Black–Scholes calculator (or use the Week 48 interactive). Compute the all-in cost as (prospectus_value - sum_of_legs) / notional. If the answer is greater than 50bps/yr, you are paying a fee high enough that DIY is strictly better in any account that can trade options.

Q10: What is the right portfolio role for a buffer ETF? For most investors: zero. The middle-of-distribution exposure is already in the index portion of the portfolio at lower cost. If the investor is constrained to a non-options account and psychologically needs the floor to stay in the market through a downturn, allocating 5–15% to a buffer ETF as a behavioural bridge can be defensible. This is a behavioural product, not a return- enhancement product.