A Second Look at CAIE: Correcting What I Got Wrong About Calamos’s Strategy

When a consulting client first asked my opinion on the Calamos U.S. Equity Autocallable Income ETF (CAIE), I approached the product with healthy skepticism. I have more than 30 years of experience analyzing similar structured products, including offerings from J.P. Morgan, and over that period I have reviewed hundreds of these structures — finding only one minor exception that made real economic sense for investors.

After publishing the article, I spoke at length with the Calamos people who built the fund. It turns out they had addressed most of the problems I raised in the design of CAIE, and the conversation changed my mind. That led me to remove the article, and I promised to correct the errors

This is the reassessment I promised, and on the substance, it is largely a correction. I hold others to the standard of following the evidence, and the same applies here. The following is where I was wrong, and the two places where my reservations remain.

What CAIE Actually Is

CAIE holds a total return swap. It references a laddered portfolio of 52 or more autocallable notes, around 79 on average, with the collateral held in the fund itself in box spreads that are used to synthetically borrow or lend money at fixed, low rates and Treasury bills. (Box spreads are an advanced options strategy that combines a bull spread and a bear spread to create a virtually risk-free, delta-neutral position.) Each note references a single volatility-controlled version of the S&P 500. In plain terms, the investor is selling a long-dated, deep out-of-the-money put. I called the wrapper a way to obscure that. It is not. Calamos describes the strategy plainly as put-writing with equity-like risk.

Selling options for premium harvests the variance risk premium, the documented tendency—across stocks, bonds, commodities, and currencies— for implied volatility to exceed the volatility that actually shows up. The seller is paid for bearing the risk that volatility spikes. It is real income for real risk, not a return of the investor’s own capital.

Return of Capital Is Not Principal Coming Back

This was the point I most got wrong.

Return of capital is a tax classification. It describes distributions a fund makes in excess of its taxable income. It does not mean the fund lacks economic earnings, and it does not mean investors are getting their own money back. CAIE’s NAV has risen from $25 at inception to roughly $27.50, and the fund has paid about $3.60 in distributions, a total return near 26% since launch.

Calamos made the distribution tax efficient on purpose, and the mechanism is worth understanding because ordinary-income treatment on a 14% payout would erode after-tax returns. The reason the distributions are classified this way is structural—the option premium never reaches the fund as premium or coupon income. The reason is that the fund neither writes the options nor holds a note. Instead, CAIE holds a total return swap on the MerQube index (MQAUTOCL), and the autocallable coupons are synthetic. They accrete into the index level instead of being paid out. So what would arrive as a coupon in a note instead arrives at the fund as price appreciation on the swap, which is capital in character. Classified as return of capital, it defers the tax to sale, at capital-gains rates. The treatment has been reviewed by Calamos’s tax advisor (EY) and fund auditor (Deloitte).

In some cases the deferred tax is not merely postponed — it is eliminated entirely. Because return of capital lowers the cost basis over time, a long-term holder builds a large embedded gain. An investor who donates appreciated shares held more than a year to a qualified charity generally deducts their full market value and never pays the deferred capital-gains tax on the appreciation. An investor who holds to death passes the shares to heirs at a stepped-up basis, and the embedded gain is forgiven. In both cases the deferred tax is not merely postponed, but eliminated.

Counterparty Risk Is Materially Mitigated

I treated the J.P. Morgan exposure as if it were an unsecured note. It is not. CAIE reaches the strategy through a swap governed by a credit support annex, with the full mark-to-market value posted as collateral to a segregated account at State Street every night, as the rules for a 40-Act fund require. If J.P. Morgan failed, shareholders would be made whole from collateral the fund already controls. If the same exposure came as a structured note, that credit risk would be real. The collateralized ETF structure effectively eliminates it.

The swap is also a binding legal contract. Neither party can unwind it alone, and both are bound to its terms through maturity. So the scenario I worried about, a terminated swap crystallizing the loss, cannot simply be triggered. As the fund grows, Calamos can also spread the exposure across more than one counterparty, capacity it estimates at upwards of $30 billion. J.P. Morgan is the counterparty, but the fund does not depend on J.P. Morgan’s solvency.

The Ladder Is Time Diversification

In the earlier article I described a “worst-of” basket that references several indices and breaches if any one falls through the barrier. That structure exists, and it appears genuinely inferior. However, Calamos does not currently use it. CAIE references a single volatility-controlled S&P 500 index, a single volatility-controlled S&P 500 index, which is easier to hedge and passes a large majority of its premium through to investors.

The ladder is time diversification, not asset diversification. The fund holds 52 or more single-index notes, around 79 on average, issued on different dates, with different starting levels and maturities, none more than 5% of the portfolio. That addresses the timing risk a single note carries, the risk I had noted. Positions are marked daily and the methodology is public.

The Fees Are Not Stacked

Investors pay one management fee of 0.74%. There is no second layer of direct management fees. The swap financing cost, SOFR plus 10 basis points, is disclosed in the published holdings. The dealer’s economics on the underlying notes are already inside the index returns, so the roughly 14% coupon and the index history are net of them.

What I missed is the collateral. The cash behind the swap is invested in box spreads through the Calamos Tax-Aware Collateral ETF, whose fee is waived for this fund, and it has earned roughly SOFR plus 40 basis points. What makes the fee picture unusu al is that the collateral actively works against it. The cash behind the swap earns roughly SOFR plus 40 basis points — 30 basis points more than the 10 basis points paid to finance the swap — and that surplus offsets a meaningful portion of the 0.74% management fee. The math nets to about 44 basis points in total cost: 0.74% minus the roughly 0.30% net collateral benefit.

Volatility Targeting Solves the Problem I Raised

This is a mechanism I misunderstood, and it is the heart of the product.

The reference index is managed to a 35% volatility target. It levers up in calm markets, when volatility is lower, and de-levers in turbulent ones, when volatility spikes. When the S&P’s realized volatility runs near 17.5%, the index carries about twice the market’s exposure. When volatility spiked to about 70% in March 2020, it cut exposure to about half. That 35% is the index’s internal volatility target, not the volatility the investor actually experiences, and it governs how much premium the strategy is paid. The fund holds coupon-bearing notes on that index, and those notes deliver a return whose volatility is closer to the market’s own, around 18 to 19%. So the investor is paid as if selling options on a 35% index but lives with roughly S&P-like volatility. The coupon runs roughly SOFR plus 10 to 11% per note. With SOFR near 3.7%, that is the roughly 14% headline. It is driven by interest rates rather than by the day-to-day swings in implied volatility that buffet an ordinary note, because the volatility target holds the volatility input steady. There is no target distribution to defend or cut. The fund pays out what the index collects.

The volatility targeting also explains the crisis record, which is worth stating in both directions. The mechanism works best when volatility rises before the market reaches its lows, because rising volatility is its signal to cut exposure. The COVID shock was exactly that. Volatility spiked almost instantly, the fastest on record, with implied volatility breaching 80% on March 16, 2020. The MerQube reference index, which holds a 35% volatility target, cut its exposure to the S&P 500 by about half within days, before the market bottomed. So by the time the S&P was down 34%, the strategy was carrying far less exposure, and the laddered index as a result only fell around 12%. A highly defensive move. The Russell 2000, on the other hand, was down significantly more (a popular index tied to worst-of autocalls). The 2008 and 2009 experience was less positive. The decline was deep and drawn out, and the notes that were caught had been struck in the calm, low-volatility years before the crisis, when the index carried its highest exposure. A prolonged crash that begins from a calm starting point is the one setting where a single note can fall more than the index, and the strategy fell about 61% against the S&P’s 55%. Ultimately 29 notes breached their maturity barriers, crystallizing a 17% loss – still less than that of a single note’s maturity risk. Across the full ladder though, the strategy’s realized volatility has resembled the S&P’s own, around 18 to 19%, and gross coupons have averaged about 12.5%.

There is also the matter of dividend risk, and here the index does something innovative. It is built on S&P 500 futures rather than the cash index, and futures pricing carries no dividend assumption. That removes dividend risk from the bank’s pricing model. A bank that writes autocallables on the cash index (total return or price return) has to forecast dividends, and it loses if that forecast is wrong. Many European banks learned this in 2020, when companies slashed dividends in the COVID shock and the banks that expected to receive those dividends did not, ultimately suffering loss. CAIE’s framework removes that exposure.

In place of the real dividend, the index applies a fixed 6% synthetic dividend, called a decrement, and its link to the coupon is the part worth understanding. A forward price is the market’s best estimate of where the index will trade later, and subtracting a dividend pulls that estimate down. The lower the index is expected to trade, the more valuable the put the fund is selling, because the put is more likely to pay out. So subtracting a fixed 6% a year rather than the S&P’s real 1.5% pushes the forward down, raises the premium the fund collects, and lifts the coupon. The reference index grows about four and a half points a year slower than the real S&P as a result, but that drag is on the index’s price, not on the fund’s return, because the fund holds notes that pay the coupon rather than tracking the index. So the higher coupon flows through. The decrement adds about 6% a year to the coupon, the line in the table below, and historically, in the MQAUTOCL Index, it has lifted total return by about 3% a year against a comparable index without a decrement. In other words, a decrement is additive in this case.

NAV Erosion Is Not the Expected Outcome

In the earlier article I stated that the NAV must decline if distributions are mostly return of capital. The record says otherwise. CAIE’s NAV is higher than at inception. Across the index history to 2005, the only real principal impairment came when a cluster of about 29 notes (those struck in 2006 and 2007 that matured into the post-crisis trough around 2011) took a loss of about 17%. That cluster is the source of the roughly 2.8% of all maturities that have breached the barrier since 2005. The rest of the time the price-return series sits at par when the market is flat or positive, at a discount to par when the market is negative and returns to par as the market recovers. By design, the notes mature or are called at par. Unlike a covered-call strategy, they are not built to cap the rebound after a drawdown. This is much different from selling off upside as the market declines and eroding NAV over time. It is more akin to a basket of bonds that are called or mature at par as long as the equity market is not down too far.

Note Pricing

I had initial reservations about the note pricing behind CAIE. My discipline is to ask not whether you will make money, since a put writer usually does in a rising market, but whether you are paid the market’s fair price for the risk. Calamos puts fair value near 16% against the roughly 14% the fund receives, so the implied 2% spread is the dealer’s profit. That is well below the 4 to 5% overpricing the research found in retail notes, and the wrapper buys liquidity and transparency those notes never had. But the coupon still sits below the fair value of the risk, and investors should know it. We can see this borne out in the “coupon math” of a single autocallable note with the same parameters.

The Math of a Single Note

Take one note within the laddered index. It runs for five years against the MerQube US Large Cap Vol Advantage Index, with a barrier 40% below the starting level that is tested only at maturity (European style). Here is how J.P. Morgan builds the coupon on that note, data as of May 2026. The table below breaks the coupon into its components, showing exactly where the 14% comes from.

Breakdown of a Single Autocallable Note

Contribution

5-year risk-free rate

+3.7%

5-year 40% down-and-in put, vanilla

+1.8%

35% volatility target

+2.0%

6% annual decrement

+6.0%

Autocall provision (The ability to autocall the note before the 5-year maturity is beneficial and costs about 50bp)

−0.5%

Conditional coupon strip (the digital calls)

3.2%

Estimated dealer hedge costs

-2.0%

Gross indicative coupon

14.2%

Net coupon after fee and collateral offset

~13.8%

Source: J.P. Morgan, May 2026. For illustrative purposes only.

The coupon is easiest to read as two parts. Because the position is fully collateralized, it starts from the five-year risk-free rate, about 3.7%. Everything above that, another 10 to 11 points, is the premium earned for selling the downside risk. Add the two and you reach about 14%. So when the coupon is described as the risk-free rate plus a 10 to 11 point spread, or simply as about 14%, those are the same number the table reaches line by line.

The pieces in between are where that premium comes from. The vanilla put is the raw cost of the 40% downside protection, and a 40%-out-of-the-money put is cheap on its own, worth about 1.8%. The 35% volatility target adds roughly another 2%, because the index is sized to behave as if it were about twice as volatile as the market usually is, which doubles the premium that selling the put earns. The 6% synthetic dividend adds about 6% more, by the forward-price mechanism described earlier. The autocall provision subtracts about half a point, the price of letting the note redeem early at par when the index is flat or higher. And the conditional coupon strip adds about 3.2%. That strip is the set of digital options that pay the coupon on each observation date the index is above the barrier, and selling it is the core of the income. Those pieces sum to about 16.2% — fair value for the risk — and about 2% comes back out as the dealer’s hedge costs, with the implied dealer’s profit embedded in them, leaving the roughly 14.2% gross indicative coupon. After the fund’s fee, offset by the collateral, the net coupon is close to 13.8%.

The payoff is then easy to state. If the S&P is higher, flat, or down by less than 40% at maturity, the investor collects the coupon and gets par back, earning about 14% a year after cost. If the S&P finishes more than 40% below where it started, the investor takes the index’s loss from that starting level, cushioned by the coupons already collected. In exchange for the roughly 14% coupon, the investor gives up any annual return above it. In a year the S&P returns 25%, CAIE still pays about 14%. The capped upside, the left-tail risk below the barrier, and the risk-free base are together what fund the high, steady income. The volatility target holds the volatility input steady, so each new coupon moves mainly with interest rates, not with market volatility or dividends, and laddering 52 or more of them smooths the result. As new money comes in, the fund sizes up the swap tied to the laddered index, MQAUTOCL.

The second reservation I have is the data. The live index history goes back only to 2005, because the volatility products this strategy relies on did not exist before the early 1980s, and the volatility-surface data becomes questionable before then as well. The backtest shows 96% of coupons paid, with 2.8% of maturities breaching the barrier, going back to 2005. That window covers the financial crisis, the fastest rate hikes in decades, and the Liberation Day drawdown, but the worst equity tails on record, 1929 to 1932, 1973 to 1974, 1987, and 2000 to 2002, are not in it. Equity markets have a long habit of underpricing fat-tail risk. A position in CAIE should be sized knowing that the deepest tail risk is the part the data can show us least well.

Who Should Consider It

Investors who want high, potentially stable and tax-efficient monthly income, and who accept that they are taking equity left-tail risk to get it. CAIE has a volatility profile similar to the S&P 500 and belongs inside an equity allocation, not as a bond substitute or a low-risk holding. It suits investors who can tolerate equity-size drawdowns, who will hold through them rather than sell at the bottom, and who value daily liquidity over the opacity of a held-to-maturity note.

It has been my experience that no institutional investors bought the structured notes I reviewed, and neither did the employees of the issuers, who knew better. In my discussions I learned that Calamos employees own this fund themselves, and some hold it for their families’ income. That is generally a good sign.

Who Should Avoid It

· Anyone who thinks it is low-risk or fixed-income-like.

· Anyone unwilling to accept a coupon that sits modestly below fair value.

· Anyone who cannot stomach an equity-size drawdown, or the chance of principal impairment if the index finishes more than 40% below its start at maturity.

The Bottom Line

The specific charges in my original piece do not hold up. Investors are not simply getting their principal back. The fees are not hidden in layers. The counterparty risk is not uncompensated. NAV erosion is not the expected outcome. What remains is a fair debate about whether the risk premium is priced at fair value, and how much weight to give the tails the data cannot show.

Hamilton Reiner, who runs JPMorgan’s Equity Premium Income ETF (JEPI), has called this approach “an intellectually honest way to invest.” Having done the work, I understand why. The lesson is the one I preach to others. Do the work on the specific product, not the category. CAIE is not what I described the first time. It is a high-income equity strategy with real and clearly disclosed risk, and an investor who understands what they own can decide whether the trade is worth making.

Larry Swedroe is the author or co-author of 18 books on investing, including his latest, Enrich Your Future. He is a consultant to RIAs as an educator on investment strategies. This article is for informational and educational purposes only and is not specific investment, accounting, legal, or tax advice. Figures are as represented by Calamos and drawn from fund filings, and should be independently verified.