Tail-Risk Hedging: Puts, CAOS, HIDE, and the Managed-Futures Alternative

A tail hedge can do exactly what it promises and still make you poorer. That is the central lesson of the most transparent hedging experiment on record: a Cboe index that always owns the S&P 500 and always buys put protection. Over the 39 years from July 1986 through May 2025, the protection softened every crash it met, cut volatility, and improved the shape of the return distribution. It also gave up about three percentage points of compound return per year. The useful question about any tail hedge starts there: what does the protection cost over decades, and could a simpler portfolio have delivered the same sleep at a lower price?

This matters now because tail-risk products have become easy to buy. CAOS packages option-based crash protection into an ETF, HIDE packages a defensive trend model into an ETF, and every brokerage account can roll its own puts. Each of these hedges a different tail, at a different cost, with a different failure mode. For most DIY investors, we think the right default remains a stock/bond/cash allocation whose ordinary bear markets you can survive without any of them. But the case for that default is stronger, and more interesting, than “hedging is a scam,” and there are specific situations where explicit insurance earns its keep.

What counts as tail risk

An ordinary bear market, a 20% to 30% decline that unfolds over months and recovers over a few years, is already priced into sensible planning. Tail risk refers to something worse: an outcome far into the left tail of the return distribution, severe enough, fast enough, or badly enough timed that ordinary diversification stops being sufficient. The 33.8% S&P 500 collapse in 23 trading days in February and March 2020 was a tail event. So, arguably, was 2022, when stocks and long-term bonds fell together and the classic 60/40 portfolio had one of its worst years in decades.

Those two examples already expose the core problem with the phrase “tail hedging”: different tails call for different defenses, and a hedge built for one can be useless against another.

The most common error in tail-hedging debates is judging every hedge against the same crash. A put position that pays spectacularly when the market falls 25% in four weeks can contribute almost nothing to an 18-month grind lower, because each monthly put expires before the cumulative decline reaches its strike. A trend-following strategy has the opposite profile: it may need weeks to cut exposure or go short, so a gap-down crash can hit it at full equity exposure, while a slow bear market gives it all the time it needs. Keep that asymmetry in mind through everything below. It decides which product fits which fear.

Someone has to sell you the insurance

A put option transfers your crash losses to a counterparty, and the counterparty knows exactly when that bill comes due: at the worst possible moment, when everyone else also wants protection. Sellers demand compensation for standing on that side of the trade, so index option prices embed premia for volatility risk, jump risk, and crash risk on top of actuarially expected losses. Coval and Shumway showed in 2001 that under mild assumptions the expected return on a put option sits below the risk-free rate, and that S&P 500 index option data matched the predicted sign, with puts even more negative than their market exposure alone would explain.¹ Bollerslev, Tauchen, and Zhou later showed the variance risk premium, the gap between option-implied and realized variance, is itself an economically significant priced quantity.² And Chambers, Foy, Liebner, and Lu, examining index option returns from 1987 through 2012, concluded that the average cost of buying out-of-the-money puts for tail protection may include a significant premium over model-implied fair value.³

Resist the tempting next step, which is to declare puts irrationally overpriced. Broadie, Chernov, and Johannes showed that put returns are so fat-tailed and so dominated by rare events that even decades of data struggle to distinguish “overpriced” from “fairly priced compensation for jump risk.” Apparently enormous put-selling profits are less statistically anomalous than they look.⁴ The defensible summary: equity tail protection has historically been expensive, and that expense may be exactly the price a functioning insurance market charges. Insurance can carry a negative expected excess return and still be rational for a sufficiently risk-averse buyer, the same way homeowners insurance is.

One recent wrinkle complicates the long-run averages. Dew-Becker and Giglio document a structural break around 2012, after which S&P 500 option returns stopped showing statistically significant negative premia; the VIX, which once sat about three points above subsequently realized volatility, has nearly converged with it.⁵ Our guide to option-income and buffer products covers what that decline means for option sellers. For option buyers, it cuts the other way: protection today may cost less than the 1986-2025 record suggests. A smaller markup still leaves the buyer holding an asset whose expected return sits at or below cash, so the finding narrows the debate without settling it.

What 39 years of rolling put protection returned

The cleanest evidence on permanent put buying comes from an index with no manager discretion at all. The Cboe S&P 500 5% Put Protection Index (PPUT) holds the S&P 500, reinvests dividends, and buys a one-month SPX put roughly 5% out of the money every month, with back-tested history starting June 30, 1986.⁶ In 2025, Asness, Cao, Ilmanen, and Villalon published the long-run accounting in the Journal of Portfolio Management:⁷

Read the table as an insurance receipt. The put program did part of its job: volatility fell by three points and the monthly skewness improved from -0.70 to -0.28, meaning the worst months got meaningfully less bad. In exchange, the strategy gave up 3.1 percentage points of compound return per year for 39 years and ended with a worse Sharpe ratio than the unhedged index. Deeper protection told the same story: the Cboe S&P 500 Tail Risk Index (PPUT3M), which buys 10% out-of-the-money quarterly puts, compounded at 8.0% versus 10.3% for the S&P 500 from April 2004 through May 2025.⁷

The protection was also far from absolute. A Wilshire study for Cboe covering June 1986 through December 2018 put PPUT’s worst peak-to-trough drawdown at -38.9%, against -51.0% for the S&P 500, with PPUT compounding at 6.6% versus 9.8% over that window.⁸ Monthly 5% out-of-the-money puts turn a catastrophic drawdown into a severe one. They do not turn it into a mild one, because in a grinding bear market each put expires before the next leg down and the investor keeps paying to re-arm.

This is the exhibit to remember whenever a tail-protection pitch shows you only the crash months. Over full cycles, always-on put protection behaved exactly like insurance: real payouts, and premiums that exceeded them by a wide margin.

Waiting for cheap volatility does not fix it

The natural response is to buy protection only when it looks cheap, meaning when the VIX is low. Israelov and Nielsen tested that idea directly. Sorting March 1996 through June 2014 into VIX deciles, they found that delta-hedged 5% out-of-the-money S&P 500 puts, the cleanest measure of what protection costs once you strip out the equity exposure, lost money on average in every single decile, including the calmest ones. Across ten global equity index markets, the same delta-hedged puts had negative average returns in all ten.⁹

The mechanism is worth internalizing: a low option price and a cheap option are different things. The relevant comparison is implied volatility against the volatility that subsequently arrives, and that spread has historically stayed positive even in quiet markets. In their sample, the VIX exceeded subsequent realized volatility by 3.4 points on average and was higher 88% of the time; even in the lowest VIX decile the gap averaged 2.5 points. A $2 insurance policy is still expensive if the actuarially fair price is $1. Buying puts only when premiums are low, in other words, still meant overpaying; buying them only right before crashes would work wonderfully, but that is a market-timing skill, and anyone who has it does not need puts.

The option-strategy menu

The long VIX ETP row deserves its own warning. VIX futures spend most of their life in contango, with longer-dated contracts priced above near-dated ones; Whaley found the curve upward-sloping on roughly 80% of trading days for 30-day-maturity futures. A fund that permanently rolls long VIX futures therefore sells low and buys high month after month, and Whaley estimated that ETPs tracking short-term VIX futures indexes had already destroyed nearly $4 billion of investor capital by 2012, concluding they were “virtually guaranteed to lose money through time.”¹⁰ The spike protection is real; the carrying cost historically consumed it and more. Our option-pricing guide walks through why premiums above fair value are the norm for crash-linked assets rather than a market error.

CAOS tries to make the insurance pay for itself

The Alpha Architect Tail Risk ETF (CAOS) is worth separating from everything above, because its design attacks the exact weakness the PPUT record exposes: the chronic cost of staying protected. Per the February 2026 prospectus, the sub-adviser (Arin Risk Advisors) runs combinations of long and short SPX puts and calls to gain a varying amount of index exposure and generate cash flows, allocating roughly 20% of capital to that exposure and 1% to 10% to separate protective options meant to appreciate in a market dislocation. The rest sits in a collateral portfolio of Treasuries, money-market instruments, and option box spreads, which may be invested entirely in the affiliated BOXX ETF. Excluding the protective options, index exposure generally ranges from 120% down to -40%. The fee is 0.63%, with the adviser contractually absorbing the acquired-fund fees from the BOXX position.¹¹

That structure gives the fund three engines: collateral that earns cash-like returns, an option book that harvests premium and market exposure rather than only burning it, and a standing inventory of protective puts. The prospectus defines the tail event those puts target precisely: a decline generally greater than 25% within a few months, accompanied by implied volatility sustained above 50, with 2008-09 and the 2020 COVID crash given as the examples.¹¹ The sponsor’s marketing materials use a looser empirical threshold, describing the protective puts as triggering when the S&P 500 falls 10% or more within 60 days.¹⁴ Either way, the design target is speed. Slow declines are explicitly outside the mandate.

The 2020 evidence, with its asterisk

The headline evidence for the design comes with a history lesson attached. CAOS carries an August 2013 performance inception, but until March 6, 2023 it was the Arin Large Cap Theta Fund, a mutual fund whose record the ETF legally adopted (followed by a 1-for-8 reverse share split that same month). The SEC reorganization filing describes the two vehicles as having identical investment objectives and similar principal investment strategies, the main difference being the ETF’s ability to hold BOXX as collateral.¹² With that caveat stated: the predecessor returned +28.99% in the quarter ended March 31, 2020, its best quarter on record, and +22.42% for calendar 2020, both figures from the fund’s own SEC filings.¹¹ A fund that gains 29% while the market loses a third of its value in five weeks is demonstrating live convexity, in real dollars, at the exact moment convexity is hardest to buy.

The 2022 evidence matters just as much

Alpha Architect’s own fund materials publish the other half of the record, and the contrast is the entire lesson:¹³

In 2022 the S&P 500 took 282 days to fall 24.5%, implied volatility never sustained crisis levels, and CAOS lost 18.5%, capturing most of the decline it was nominally bought to hedge. Anyone holding CAOS as a bond substitute learned that year what they owned. The sponsor, to its credit, says the same thing: its materials pair CAOS against fast crashes and point to managed futures for slow ones, and they classify the April 2025 tariff selloff (an 18.1% drop in 35 trading days) as a fast crash in which CAOS outperformed.¹⁴

What the full record looks like

Over the whole adopted history, CAOS has behaved like a cash-plus diversifier with a lottery ticket attached: a 3.38% annualized NAV return since August 2013 as of July 31, 2026,¹⁵ roughly 11.6% annualized monthly volatility with strongly positive monthly skew,¹³ 0.19 correlation to the S&P 500 through the end of 2024,¹⁶ and a maximum daily-NAV drawdown of -26.4% over 2013 to mid-2025.¹⁴ Those numbers describe a genuinely uncorrelated asset that has paid for its own insurance, a real engineering achievement relative to PPUT-style bleed. They also describe an asset returning less than T-bills over stretches of the sample, with equity-scale volatility and a deeper max drawdown than an aggregate bond fund. The strategy reduces the cost of convexity; the economics of insurance are still in there.

Our verdict: a small, low-single-digit allocation is defensible for an investor who specifically wants fast-crash convexity next to an equity-heavy portfolio and understands what 2022 demonstrated. Sizing it like a bond allocation, 10% to 30% of a portfolio, treats it as something its return distribution has never been. For the mechanics of the fund family, including BOXX itself, see our Alpha Architect fund review.

HIDE hedges a different tail, and does it long/flat

The Alpha Architect High Inflation and Deflation ETF (HIDE) often gets shelved next to CAOS as the firm’s other “tail fund,” and the pairing misleads. HIDE holds no options and owns no crash protection. Per its prospectus, it runs a monthly-or-faster absolute-momentum and trend model across three sleeves, with target weights of 50% intermediate-term US Treasuries, 25% real estate, and 25% commodities when every sleeve reads positive. Each sleeve scales to half exposure or fully to cash as its two trend signals turn negative, and the fund can sit at 100% cash and equivalents. It builds the exposures from other ETFs, and it charges 0.34% gross, waived to 0.29% net through February 1, 2027.¹⁷

That makes HIDE defensive tactical asset allocation aimed at slow macro regimes, especially the inflation regime where nominal Treasuries and stocks fall together. The sponsor’s own tagline calls it “a cheaper, simpler, lower volatility alternative to a managed futures strategy,”¹⁸ and Alpha Architect’s model diversifier portfolio uses the two funds as complements, 50% HIDE and 25% CAOS alongside 25% intermediate Treasuries, one for slow regimes and one for fast crashes.¹⁹

Two limits keep it in the “managed-futures-lite” category. First, it is long/flat: when bonds are collapsing, a diversified trend fund can short bond futures and profit, while HIDE can only step aside into T-bills, which avoids losses without converting the decline into gains. Second, the live record is short and lands entirely after the event that best motivates the strategy. HIDE launched November 16, 2022, one month after the 2022 inflation-driven bear market bottomed. Since inception it has returned 3.73% annualized at NAV as of July 31, 2026,¹⁸ and through the end of 2025 it trailed its own prospectus benchmark, the Solactive US Aggregate Bond Index, 2.06% to 4.13% annualized.¹⁷ A defensive trend model that lags a plain bond index during a calm, rising market is behaving as designed; that is what the insurance costs. But nothing in the live history yet shows the payoff side, so an allocation to HIDE today is a bet on the century of trend evidence behind the design rather than on anything the fund has demonstrated.

Managed futures, the hedge with a positive historical return

If the case against permanent put buying is its cost, the case for trend-following managed futures is a long-run record of crisis protection without that cost. The academic record here is unusually deep. Moskowitz, Ooi, and Pedersen documented time-series momentum across all 58 liquid futures and forward markets they examined, spanning equity indexes, bonds, currencies, and commodities, with the diversified strategy performing best in extreme markets, both up and down.²⁰ Hurst, Ooi, and Pedersen pushed the evidence back to 1880 across 67 markets and found positive average returns in every decade since, plus positive performance in 8 of the 10 largest 60/40 drawdowns in the sample.²¹ Work on actual CTA funds shows the mechanism operating live: Asif, Frömmel, and Mende found managed futures earned their crisis alpha through broad diversification plus rapid de-risking, cutting exposure to crisis-hit markets within roughly 15 days.²² And on the tail that broke 60/40 in 2022, Neville, Draaisma, Funnell, Harvey, and Van Hemert examined nearly a century of US, UK, and Japanese inflation episodes and found trend following the most reliable dynamic strategy during major inflation shocks.²³

The weakness is the mirror image of the put’s strength: trend needs time. A market that gaps down 20% and rebounds before the model flips can whipsaw the strategy into losses on both legs, and a one-day 1987 offers it nothing. AQR’s direct comparison of the two approaches lands where the evidence points: put strategies provide the more dependable protection over days and weeks but have carried a reliably negative long-run premium, while trend following provides less certain immediate protection with positive long-run returns, making it the better strategic, long-horizon tail hedge for investors who can hold it.²⁴ We agree, with the standing caveat that the diversification arrives with years-long stretches of dull or negative performance while stocks run. Our full managed-futures guide covers the evidence, the 2022 lesson, the fund menu, and the tax mechanics.

The case for plain stocks and bonds

Before adding any of the above, ask the prior question. An investor who cannot tolerate the drawdowns of a 90% equity portfolio has a simpler lever than derivatives: hold less equity. Shifting 10 or 20 points from stocks to cash and short Treasuries cuts drawdown risk immediately, costs nothing to maintain, adds no manager risk, and requires no beliefs about option pricing or trend persistence. AQR made this the punchline of its tail-hedging critique back in 2011: after pricing direct insurance, the alternatives that survived analysis were portfolio-construction moves, including simply de-risking.²⁵ The buffer-fund literature makes the same point empirically: 86% of roughly 100 options-based downside-protection funds trailed a beta-matched blend of stocks and cash from January 2020 through April 2025,⁷ and Mill Creek found that a 62.5/37.5 stock/cash blend matched the Cboe buffer-index volatility from mid-2006 through September 2025 with a higher return and a smaller maximum drawdown.²⁶ Match the risk level first with cheap assets, then ask what an option structure adds. Usually the answer is fees.

The strongest argument against stopping at stocks and bonds is 2022 itself. Stock/bond correlation is regime-dependent: Brixton and co-authors show stocks and bonds respond with the same sign to inflation surprises and opposite signs to growth surprises, so inflationary regimes push the correlation positive, exactly when 60/40 is most exposed.²⁷ Treasuries hedge recessions, and hedged them well in 2000-02, 2008, and 2020; they amplified the damage in 2022. The conclusion we draw is measured: stocks plus high-quality bonds remain an excellent core, and they leave one genuine gap, the inflationary decline, which TIPS, cash, and trend following address more directly than crash insurance does. Our guide on what bonds actually hedge works through that regime dependence.

Try it: hedge, or hold less equity?

The calculator below makes the de-risking comparison concrete. It builds a portfolio with a tail-hedge sleeve, then constructs the stock/cash portfolio that would lose the same amount in the fast crash the hedge is designed for, and compounds both over 20 years. The assumptions you set for the hedge’s ordinary-year bleed and its crash payoff decide the winner, which is precisely the point: those two numbers are the entire economics of tail insurance, and the crash payoff is unknowable in advance. Flip the crash to a slow bear to see the 2022 problem.

Who a dedicated tail hedge is for

Because explicit insurance has a negative expected cost, the question for each investor is whether the protection buys something their situation specifically needs. Utility, in the economist’s sense, varies a lot more across investors than expected returns do.

The behavioral rows deserve emphasis. A hedge costing 1% per year that stops one panicked liquidation at the bottom of a 35% decline is cheap, and that benefit never shows up in a Sharpe ratio. The condition is discipline in the other direction: the insurance must be in place before the crash and held through the boring years. Buying protection after volatility explodes means paying the highest premiums of the cycle at the moment the psychological pull is strongest. If protective puts appeal to you for a concentrated stock position rather than a portfolio, that narrower use case has better economics, covered in our options guide.

Sequence risk is a different problem

Retirees hear tail-hedging pitches most often, because sequence-of-returns risk is real: a severe crash in the first years of withdrawals does disproportionate, sometimes unrecoverable damage. But look at what the retiree needs, which is to avoid selling depressed assets to fund near-term spending. A ladder of cash, short Treasuries, and TIPS covering several years of withdrawals accomplishes that directly, with positive expected returns and no dependence on crash speed. Flexible spending rules do the rest. A portfolio that prints +25% during the next fast crash is a luxury for this purpose; spending capacity that survives the crash is the requirement, and it is much cheaper to buy. Explicit tail insurance starts to earn a place only for the retiree who insists on holding an unusually equity-heavy portfolio through the danger window. Our sequence-risk guide covers the standard defenses in detail.

CAOS, managed futures, and HIDE side by side

What we recommend

For the typical DIY investor building around broad global equity index funds, our default remains the simple one: hold globally diversified stocks plus enough high-quality fixed income, cash, and TIPS that the ordinary bear markets are survivable, and skip dedicated tail-risk products entirely. The order of operations if you want to go further:

1. Fix the allocation first. If a 35% equity drawdown would break your plan or your resolve, the cheapest hedge is holding less equity. Run the calculator above before pricing anything more exotic.
2. Consider diversified managed futures as the strategic diversifier. The long-run evidence supports crisis protection with positive historical returns, against both deflationary and inflationary tails. Common educational ranges run 5% to 20% of a portfolio, and the allocation only works if you can watch it lag stocks for five or more years without capitulating. Tracking-error tolerance is part of risk tolerance.
3. Consider a small CAOS-like sleeve only for a specific job. Fast-crash convexity next to a deliberately equity-heavy or leveraged portfolio is a coherent use. Low single digits. It complements trend following rather than substituting for it, since the two cover opposite crash speeds.
4. Skip DIY perpetual puts and long VIX products. The 39-year PPUT record and the VIX ETP roll-cost evidence are the base rates. Overriding them requires an explicit insurance objective, a sizing rule, a monetization rule for when the puts pay off, and a willingness to fund the premium indefinitely. Few DIY programs survive all four requirements.

HIDE sits outside the hierarchy because it answers a different question. As a cheap, simple, long/flat diversifier against slow regimes it is a reasonable holding; as tail insurance it is miscategorized, and its live record has yet to show the payoff side of its design.

Key takeaways

• Insurance that works can still be a poor investment. From July 1986 through May 2025, the always-hedged PPUT index cut volatility and improved skewness, and compounded at 7.7% versus 10.8% for the S&P 500, with a lower Sharpe ratio.
• Cheap-looking options are usually still expensive. Delta-hedged 5% out-of-the-money S&P puts lost money on average in every VIX decile from 1996 to 2014. Low premiums reflect calm markets, and calm markets usually deliver even less volatility than the premiums imply.
• Match the hedge to the tail. Option convexity covers fast crashes; trend following covers slow declines and inflation shocks; Treasuries cover deflationary recessions. No single product covers all three.
• CAOS is engineered for one tail. Its predecessor gained 29% in Q1 2020 during a 34% five-week crash, then lost 18.5% in the slow 2022 decline. Both results are the design working as specified.
• HIDE is trend-lite rather than tail insurance. Long/flat across Treasuries, commodities, and REITs at 0.29% net, launched November 2022, after the inflation shock that best motivates it.
• Managed futures carry the strongest long-run case among the alternatives. Positive average returns in every decade since 1880 and gains in 8 of the 10 largest 60/40 drawdowns in that sample, paid for with long stretches of tracking error rather than an insurance premium.
• De-risking is the benchmark every hedge must beat. Holding less equity provides drawdown protection at zero ongoing cost. 86% of roughly 100 options-based protection funds failed to beat that benchmark from 2020 through April 2025.

How Summitward helps

The decision underneath this whole guide is whether your plan survives bad markets without exotic protection. Summitward’s retirement tools are built to answer exactly that: Monte Carlo simulation with fat-tailed return assumptions, stress tests that replay severe historical sequences against your actual allocation, and spending-policy comparisons that show how much protection flexible withdrawals buy on their own. If the plan holds at your current allocation, you have your answer about tail hedges. If it breaks, the fix is usually visible in the allocation and spending levers before it is visible in any fund catalog.

Frequently asked questions

Is CAOS a good bond substitute?

No. Its long-run return has been cash-like (3.38% annualized since August 2013 as of July 31, 2026), its volatility has run near 12%, and its maximum drawdown of roughly -26% is deeper than an aggregate bond index fund’s. What it offers that bonds cannot is positive convexity in a fast crash. That is a different job, and it deserves a different, smaller line in a portfolio than a bond allocation.

Did tail-risk hedging work in 2022?

It depended entirely on the hedge. Short-dated put strategies and CAOS suffered alongside stocks because the 282-day decline never triggered crash-level payoffs; CAOS lost 18.5% against the S&P 500’s 24.5% drawdown. Trend-following managed futures had one of their best years on record, because a slow, persistent, inflation-driven decline across stocks and bonds is the exact environment trend models are built for. See the managed-futures guide for why that year is also easy to over-extrapolate.

Are put options overpriced?

The long historical record shows put buyers paying more than actuarially fair prices, but the academic literature is careful about calling that mispricing, since sellers bear crash risk at the worst possible times and demand compensation for it. Recent research also finds the premium has weakened substantially since around 2012. Expensive insurance can be rational to buy; the PPUT record simply shows what the expense added up to for a buyer who never stopped paying.

Is HIDE a tail-risk fund?

Its sponsor markets it as an inflation-and-deflation diversifier and an alternative to managed futures, and that is the right shelf. It holds no options, cannot profit from a crash, and can only step out of falling assets into cash on a monthly-or-faster signal. Expect it to help in slow regimes like 2022 and to do little in a fast crash.

Should I buy VIX calls or a VIX ETF before a crash?

Before a crash, yes; the problem is knowing when that is. Held continuously, long VIX futures products have been among the most expensive hedges available, because the futures curve sits in contango on roughly 80% of days and rolling a long position eats the difference. Whaley estimated investors had lost nearly $4 billion in short-term VIX ETPs by 2012. As standing insurance, they are the clearest no of any strategy discussed here.

How big should a tail-hedge allocation be?

If a hedge needs to be large to make your portfolio survivable, the portfolio has too much equity, and reducing it is cheaper than insuring it. Dedicated convexity sleeves that make sense are small, low single digits of the portfolio, sized so the ordinary-year drag is tolerable and the crash payoff is meaningful but not load-bearing. Managed-futures sleeves run larger, commonly 5% to 20%, because they are diversifiers with positive expected returns rather than insurance.

Related guides

• Managed Futures and Trend Following: the full evidence, fund menu, leverage debate, and tax treatment for the diversifier this guide recommends first
• Alpha Architect ETFs: the fund-family review covering CAOS and HIDE mechanics, BOXX collateral, and the rest of the lineup
• Boomer Candy: buffer funds and option-income ETFs, the sell side of the same volatility market, and the evidence that stock/cash blends beat them
• Sequence of Returns Risk: why early-retirement crashes hurt most and the standard defenses that need no derivatives
• Bonds: Diversifier or Hedge?: the regime dependence of stock/bond correlation that creates the gap trend following fills
• The Black-Scholes Equation: how option premiums are set and why protection prices exceed model fair value
• The Hungry Caterpillar Portfolio: an all-weather framework that deliberately dropped its long-volatility sleeve, and the reasoning behind that choice
• Options: Risk and Reward: the narrower cases, concentrated positions and lockups, where protective puts and collars earn their cost

Author disclosure

The author holds a small managed-futures position through RSST (Return Stacked U.S. Stocks & Managed Futures ETF) in a retirement account, and no position in CAOS, HIDE, or dedicated put or VIX strategies. This guide is descriptive, not promotional. Nothing here is a recommendation to buy, sell, or hold any specific security.

Sources

1. Coval, J.D. & Shumway, T. (2001). “Expected Option Returns”. The Journal of Finance, 56(3), 983-1009. Source for the theoretical prediction that put expected returns sit below the risk-free rate and the S&P 500 index option evidence.
2. Bollerslev, T., Tauchen, G. & Zhou, H. (2009). “Expected Stock Returns and Variance Risk Premia”. The Review of Financial Studies, 22(11), 4463-4492. Source for the variance risk premium as an economically significant priced quantity.
3. Chambers, D.R., Foy, M., Liebner, J. & Lu, Q. (2014). “Index Option Returns: Still Puzzling”. The Review of Financial Studies, 27(6), 1915-1928. Source for the finding that out-of-the-money index put buyers paid a significant premium over model-implied fair value, 1987-2012.
4. Broadie, M., Chernov, M. & Johannes, M. (2009). “Understanding Index Option Returns”. The Review of Financial Studies, 22(11), 4493-4529. Source for the sampling-uncertainty and jump-risk counterpoint to calling puts mispriced.
5. Dew-Becker, I. & Giglio, S. (June 2026). “The decline of the S&P 500 variance risk premium”. Earlier version issued as Federal Reserve Bank of Chicago Working Paper 2025-17. Source for the post-2012 structural break in option premia and the VIX-realized convergence.
6. Cboe Global Indices (August 2025). “Cboe S&P 500 Put Protection Indices Methodology”, v1.1. Source for PPUT’s monthly 5% out-of-the-money construction, the June 30, 1986 base date, and PPUT3M’s quarterly 10% out-of-the-money construction.
7. Asness, C., Cao, J., Ilmanen, A. & Villalon, D. (September 2025). “Rebuffed: An Empirical Review of Buffer Funds”. The Journal of Portfolio Management, 51(10). Source for the PPUT statistics (July 1986 to May 2025), the PPUT3M comparison (April 2004 to May 2025), and the finding that 86% of roughly 100 options-based funds trailed beta-matched stock/cash blends from January 2020 through April 2025 (the companion “Buffer Madness” analysis).
8. Wilshire Analytics (March 2019). “Options-Based Benchmark Indexes: Performance, Risk and Premium Capture (June 1986-Dec. 2018)”, prepared for Cboe. Source for the -38.9% versus -51.0% maximum drawdown comparison and PPUT’s 6.6% versus 9.8% annualized return over that window.
9. Israelov, R. & Nielsen, L.N. (2015). “Still Not Cheap: Portfolio Protection in Calm Markets”. The Journal of Portfolio Management, 41(4), 108-120. Source for delta-hedged 5% OTM puts losing money in every VIX decile (March 1996 to June 2014), the ten-market extension, and the implied-versus-realized volatility gap statistics.
10. Whaley, R.E. (2013). “Trading Volatility: At What Cost?”. The Journal of Portfolio Management, 40(1), 95-108. Source for the roughly 80% contango frequency in VIX futures and the near-$4 billion cumulative loss in short-term VIX ETPs through 2012.
11. Alpha Architect Tail Risk ETF (February 1, 2026). Summary Prospectus and the accompanying 485BPOS filing. Source for the strategy description, the 20% / 1-10% / collateral allocation, the 120% to -40% exposure range, the 0.63% fee, the greater-than-25%-with-volatility-above-50 tail definition, the +28.99% best quarter (Q1 2020), the +22.42% calendar 2020 return, and the -14.14% calendar 2022 return.
12. EA Series Trust (November 23, 2022). Form N-14 registration statement for the reorganization of the Arin Large Cap Theta Fund. Source for the “identical investment objectives and similar principal investment strategies” characterization and the reorganization terms.
13. Alpha Architect (October 2024). CAOS fact sheet, data through September 30, 2024, archived copy. Source for the 2018/2020/2022 drawdown comparison table and the fund’s volatility and skewness statistics through September 2024.
14. Alpha Architect (2026). “Managing Fast & Slow Crashes” investment-case deck, standardized returns as of March 31, 2026. Source for the -10%-in-60-days fast-crash threshold, the CAOS-for-fast / managed-futures-for-slow framing, the April 2025 episode characterization, and the -26.4% maximum daily-NAV drawdown figure.
15. Alpha Architect, official CAOS fund page, performance as of July 31, 2026. Source for the 3.38% annualized since-inception NAV return.
16. Alpha Architect (early 2025). “CAOS | Investment case” deck, data through December 31, 2024. Source for the 0.19 correlation to the S&P 500.
17. Alpha Architect High Inflation and Deflation ETF (February 1, 2026). Summary Prospectus. Source for the 50/25/25 target weights, the long/flat signal mechanics, the ability to hold 100% cash, the 0.34% gross / 0.29% net fee with waiver through February 1, 2027, and the since-inception benchmark comparison through December 31, 2025.
18. Alpha Architect, official HIDE fund page, performance as of July 31, 2026. Source for the 3.73% annualized since-inception NAV return, the November 16, 2022 inception, and the “cheaper, simpler, lower volatility alternative to a managed futures strategy” characterization.
19. Alpha Architect (February 2026). Model portfolio investment case. Source for the diversifier model allocating 50% HIDE, 25% CAOS, and 25% intermediate Treasuries.
20. Moskowitz, T., Ooi, Y.H. & Pedersen, L.H. (2012). “Time Series Momentum”. Journal of Financial Economics, 104(2), 228-250. Source for time-series momentum across 58 futures and forward markets and the strong performance in extreme markets.
21. Hurst, B., Ooi, Y.H. & Pedersen, L.H. (2017). “A Century of Evidence on Trend-Following Investing”. The Journal of Portfolio Management, 44(1), 15-29. Source for positive returns in every decade since 1880 and gains in 8 of the 10 largest 60/40 drawdowns, 1880-2016.
22. Asif, R., Frömmel, M. & Mende, A. (2022). “The crisis alpha of managed futures: Myth or reality?”. International Review of Financial Analysis, 80. Source for crisis alpha arising from diversification plus de-risking within roughly 15 days.
23. Neville, H., Draaisma, T., Funnell, B., Harvey, C.R. & Van Hemert, O. (2021). “The Best Strategies for Inflationary Times”. The Journal of Portfolio Management, 47(8), 8-37. Source for trend following as the most reliable dynamic strategy across nearly a century of US, UK, and Japanese inflation episodes.
24. Ilmanen, A., Thapar, A., Tummala, H. & Villalon, D. (July 2020). “Tail Risk Hedging: Contrasting Put and Trend Strategies”. AQR white paper, republished in the Journal of Systematic Investing (2021). Source for the direct put-versus-trend comparison and its strategic-horizon conclusion.
25. Nielsen, L.N., Villalon, D. & Berger, A. (2011). “Chasing Your Own Tail (Risk)”. AQR white paper. Source for the analysis that portfolio-construction alternatives, including simple de-risking, beat the cost of direct tail insurance.
26. Crook, M. (October 2025). “Boomer Candy is Bad for Your (Portfolio’s) Health”. Mill Creek Capital Advisors. Source for the 62.5/37.5 stock/cash blend matching the Cboe buffer index’s volatility with a higher return and smaller drawdown, June 2006 to September 2025.
27. Brixton, A., Brooks, J., Hecht, P., Ilmanen, A., Maloney, T. & McQuinn, N. (2023). “A Changing Stock-Bond Correlation: Drivers and Implications”. The Journal of Portfolio Management, 49(4), 64-80. Source for the regime dependence of stock/bond correlation and the inflation channel behind 2022.