Theo Casey is CEO of EquityQuant.dev, co-founder of Marketcolor, and a veteran FT Alphaville contributor.
At some point, the “it’s only newsworthy because Cathie Wood is doing it” bandwagon will run out of road. Today, however, is not that day.
One of Wood’s recent ventures involves the favourite downside-protection derivatives strategy among ETF hipsters, the so-called buffer. ARK Invest applied to the SEC in July to create four defined outcome buffered ETFs referencing its Innovation ETF, one for each quarter, but has somehow spurned the opportunity to code them BARKK.
Say hello, potentially, to ARK Defined Innovation ETF family: ARKI, ARKJ, ARKL and ARKM. Dated January-January, April-April, July-July and October-October, buffer-ARKKs will list in sequence and roll concurrently in one-year series. Each ETF will have a rolling 12-month outcome period and, while there’s no fund-of-funds structure, weighting the four series in a portfolio means one can be simulated. Each series expires and resets on its first birthday.
The structure includes a 50 per cent downside cap provided by an at-the-money put, financed by an at-the-money call. For upside participation, a cheaper 5 per cent out-of-the-money call is bought with the remainder of the call premium. As with other buffers, b-ARKK will use FLEX options, non-standard contracts with customised terms sold over the counter, to avoid the headaches associated with assignment.
This is a high-beta variant of the buffers we’ve seen to date. It relies on a greater than 5 per cent gain in any 12-month period, which for ARKK would have been a safe bet in recent years.
Why now for a buffer-ARKK launch? It probably helps that defined outcome funds have been warmly received, with around $70bn net assets invested in downside-protecting exposures to the S&P, Russell, Nasdaq and others. Cboe this week announced a slated Q1 2026 launch of FLEX options in Europe, presumably paving the way for DAX, CAC and AEX buffers in the future. There is some anecdotal evidence to suggest investors buy and hold their buffers longer than they do funds without downside protection.
What’s curious is the buffer-ARKK’s gap between downside protection and upside participation. What happens in a 12-month period where the portfolio bounces between 0 per cent and 5 per cent? Very little. The investor will have no losses, but also no returns.
In our dataset starting January 2016, ARKK spent 175 of 1,745 trading up-days in the range between 0 and 5 per cent. On those days, an investor would be better served holding ARKK than buffer-ARKK.
That said, we’ve modelled the hypothetical performance of buffer-ARKK against ARKK. And it’s not bad.
Below you will see a Sigbox visualisation that simulates realised historical return distributions. The passage of time is rising, from the start of January 2020 at the bottom of the box to the end of June 2025 at the top. The parameters are arithmetic average return and realised volatility. (We call it a Sigbox because it looks like a signature in a box.)
When the performance line is blue, ARKK should outperform buffer-ARKK. When the performance line is red, buffer-ARKK should outperform ARKK. The ghostly grey wall (black in static version) represents the equivalence curve where the funds perform about as well as each other on a risk-adjusted basis.
Now follows a note to fellow quants. Normies are welcome to skip this bit.
We took some liberties to visualise ARKK vs. buffer-ARKK, and what we might understand of its potential future. Sigbox is designed to give insightful analysis on ETFs, potential future ETFs and their derivative holdings. In this example, we use a dynamic model to capture the realised distribution of the underlying return from historical data. Then we calculate an equivalence curve to distinguish where ARKK and its comparator (ie. a composite of hypothetical Defined Innovation ARKK ETF series) have the same expected return. One can observe relative outperformance from a risk-return perspective over time. The simulation is *NOT* a comparative backtest using historical data. And of course, any start date at any time is itself a bias. Our bespoke model filters out some extreme features, such as Hurst statistics, which do not persist and could skew the long-term plot.
Guessing the intended customer is what makes Ark’s run-of-the-mill N-1A filing so intriguing. Who asked for a high beta fund in a drawdown-protecting wrapper? To reiterate, buffer-ARKK limits losses to 50 per cent. If ARKK falls 72 per cent (as happened between 2021-2022, per the next chart), a hypothetical buffer-ARKK falls around 36 per cent. That’s a neat trick, but — with reference to AQR’s argument on buffers — why not just buy a low beta fund?
Perhaps it’s another manifestation of Ark’s indefatigable brand equity. Maybe buffer-ARKK is the Diet Coke to ARKK’s full-fat original. Customers who should probably have a glass of water are happier when they put faith in a brand. They want Ark’s storied upside potential, but they might prefer a less gap-downy, less circuit-breakery, less left tail-risky version of that upside potential. And maybe investors will be willing to give up 5 per cent every 12 months to achieve such a defined outcome.
When New York magazine published The Rise and Fall of Cathie Wood in 2022 (“to hear her talk was to feel your mind liquefy in a clickbait-like flood of dopamine-inducing buzzwords”) it may have been hard for many people to take her technobabble seriously. But Wood always had the aura and presentation more of a shepherd than a fund manager, and her flock cannot get enough of ARKK. With eight active ETFs, four crypto ETFs, one venture fund and two private funds, Ark has been happy to provide.
ARKK led the active ETF revolution and today has puts and calls trading with ARKK as the reference asset. We might soon be adding buffer funds made up of FLEX-option equivalents of those American-style listed options on ARKK to the list of ARKK funds.
What comes next? Levered, inverse and income ARKKs? No — because they already exist, courtesy of other ETF issuers. Tradr ETFs offers 2x long and inverse ARKK ETFs, and Leverage Shares has 3x long and inverse ARKK ETPs. YieldMax has a premium-selling income ARKK ETF.
How many times can the industry envelop the same custom basket of stocks before notional presence is so huge it creates its own gravitational pull … or push. Even before buffers are added to the ranks, some may believe we are already there.