Too Long Didn’t Read Section
FIRST, THE PART THE BLOOD-SUCKING LAWYERS MADE ME SAY 🙄😉😂
“I brought you here to defend me, and the only ones on my side are the blood-sucking lawyers 😂.” John Hammond, Jurassic Park
So here's what they want you to know, and what I'd tell you anyway: this report is about what ZEUS — the family charity hedge fund I run — is doing with its hedging bucket this week, and EXACTLY why. It is not a recommendation for your account. GNG publishes research. GNG does not manage your money, doesn't know your tax lot, your risk tolerance, or whether you think this war ends at Christmas or in 2029. Every number in here is sourced so you can check my work and decide for yourself... which was always the point of this place 😉
One more thing the lawyers did NOT make me say, and it's the part you actually need to read: the ZEUS tracker has been off. Here's why, what happened, and why it's back on today 🤗
WHY THE ZEUS TRACKER HAS BEEN OFF, AND WHY IT'S BACK ON TODAY
I'm going to tell you exactly what happened, because you deserve it 🙏
In May and June, CTA fell 22% in a matter of weeks. Our anchor — the largest position in the hedging bucket, the fund that was supposed to go up when everything else went down — lost a fifth of its value while the S&P rose 4%. At the same time, the hyperscalers and NVDA sold off, even though the AI boom was getting stronger every week and every earnings report proved it. The hedge broke, and the growth names dropped at the same time. That's the one scenario a hedged portfolio is built to avoid, and ZEUS lived it.
I ate every bite. ZEUS owned CTA at the full 30% of the bucket that Part 1 ratified — the same position, the same weight, the same loss you had if you followed the research. I don't run a model portfolio. I run the real one, with my family's money and my charity's money in it. Whatever CTA cost you, it cost me first 🫂
And we didn't know why. CTA was supposed to be a quant trend fund with a rates book and a short book; it should have been fine in May and June. We now know, from Simplify's own SEC filings — filed weeks after the damage — that the fund had already been quietly rebuilt into a levered long-oil bet. The hedge didn't fail. The hedge had been replaced, and nobody told the people holding it. That's the rest of this report.
Now the timing. The hedge is broken for reasons we can't see yet. The growth names are down for reasons that make no sense. And right in the middle of that, GNG — a business with payroll, servers, lawyers, and a family behind it — had bills come due. ZEUS is the balance sheet that funds it. I had to raise cash, roughly half the portfolio, near the bottom, a week after publishing a report telling you to shut up and buy something smart. Picture the tracker that week: "Adam is pounding the table and screaming “shut up and buy these stocks and you’ll thank me in a year” WHILE selling half the portfolio. That's not transparency. That's a screenshot with the wrong caption 🙄
So I turned it off — not to hide what I was doing, but to make sure none of you did it too. A tracker showing "Adam selling half" the week after "Adam pounded the table" is a signal, but it was wrong. The sale wasn't a view on the market. It was a cash call, forced by a hedge that had stopped hedging. If it had led one member to dump a position they should have kept, that's a cost I wasn't willing to put on you. Buffett doesn't post Berkshire's trades in real time either, and he has a better excuse than I do. Some of you will disagree with that call. I'd make it again. And thank you — genuinely — for the patience while the lights were off 🙏
In a crisis, you act to protect your family and company, and if you don’t have the time or bandwidth to explain it to others, then just pull the trigger.
What did not change: the thesis. Not one percent. When ZEUS shrank, it shrank proportionally — every position was cut to the same weight it had before. One of you said it best in chat: "I don't have that much money in my brokerage anyway; I just follow the percentages." That's exactly right, and it's why the weights are what I publish, and the dollars are not. The dollars are my business. The weights are the research.
And here's what I learned again, the hard way: sequence-of-return risk is real. I was right about the hyperscalers, right about the direction, right about the AI boom — and I got hit anyway, because being right and being solvent are two different things. Keynes had it exactly: the market can stay irrational longer than you can stay liquid. That is the entire reason a hedging bucket exists. And it's why, when I found out the hedge had been gutted from the inside, I spent the whole week rebuilding it 🛡️
CTA spent 4 years as the best hedging asset you could own… and then Management lost its damn mind and decided it knew better than the best algos.
What's happening today? Before this article goes live, ZEUS rebalances to the new bucket. Then Connor flips the tracker back on. When you check it, it will match this report: 16% DBMF, 10% IALT, 6% SDCI, 4% KMLM, 4% IAUM, CTA at zero, equities at 60%. Not "soon." Now.
What's coming. Super ZEUS — the same tickers, run by the live optimizer instead of by me on a spreadsheet at midnight — is being tracked in a real-money account right now. By year-end, we'll show you both side by side, with a full article on how it works. That's what we've been building toward, and it's the thing that eventually runs without anyone having to sell at 2 AM 😉
Here's what the new ZEUS looks like, straight from Morningstar, and what the math says it can do:
OK, NOW THE ACTUAL REPORT. THREE DECISIONS, THEN THE RECEIPTS
This was part 1.
Is CTA Broken? Three Wars, Nine Funds, and Exactly What To Do About It 😉
BUT then the facts changed…in a big way
Three decisions ZEUS is making. Then the receipts.
One. CTA: ZEUS sells it all this week.
Super ZEUS is a real-time optimized version of ZEUS (same tickers) that is going live (hopefully) at the end of the year. Running on real money right now.
Connor will be optimizing the new Super ZEUS to adapt to the new hedging bucket in the coming weeks; that’s why the live tracker isn’t working at the moment.
Two. What replaces it in ZEUS: Hedging Bucket Will Look Like This (DBMF 40% · IALT 25% · SDCI 15% · KMLM 10% · gold (IAUM) 10%)
Three. How big: 40% of the ZEUS portfolio, up from 35% — because the loudest instrument just left the orchestra, and a quieter orchestra has to be bigger to fill the hall.
So translated to the actual portfolio (which is being rebalanced on Thursday or Friday)
16% DBMF
10% IALT
6% SDCI
4% KMLM
4% IAUM (Gold)
How the finding translates beyond ZEUS. The finding is about the FUND, not about our portfolio, so here's how the same analysis reads in three other structures — as analysis, not as instructions. Where CTA is the entire hedge, the numbers say the five-fund bucket at the same dollar size does the old job, and that DBMF alone is the single closest thing to what CTA used to be. Where CTA is one sleeve of a larger hedge, the numbers say the proceeds belong in trend funds, not in a long-only commodity fund — Part 4 shows what that substitution costs in a deflationary crash. And where CTA is the "alternative" in a 60/40, understand that since July it has been a levered long-oil fund, which means it now ADDS to equity risk in a recession instead of subtracting from it. What any of that means for your account depends on things I don't know about you.
Now the receipts 😉
What this buys ZEUS — April 2022 through August 2026, downside capture first, as always
Three columns matter. The first is the bucket we built in August, with CTA's historical returns — the fund that no longer exists. The second is the SAME bucket with CTA modeled as what it actually holds today, a 1.27x levered long-only basket that's three-quarters petroleum. The third is where ZEUS is moving.
Hedge downside capture: −57.2% → −31.2% → −41.7%. Hedge max drawdown: −7.7% → −9.7% → −4.8%. The 2022 bear market: +8.4% → +2.2% → +6.1%. The two COVID months, when the S&P lost 19.6%: CTA didn't exist → −11.0% → −4.5%. Hedge Sortino: 1.27 → 1.60 → 2.61. Whole portfolio, with the S&P 500 standing in for the equity sleeve as in Part 1: max drawdown −7.04% → −9.79% → −7.12%, CAGR 13.77% → 13.79% → 13.83%. On ZEUS's actual seven stocks, the same move takes the portfolio's max drawdown from −13.6% to −12.7% and its CAGR from 22.2% to 23.2% — and ZEUS funds the five points from WTRG, not from the growth names, for reasons the tracker update will spell out.
Read the middle column twice. That column is the bucket ZEUS owns RIGHT NOW, if CTA keeps doing what it has done since July. The bucket's downside capture is nearly halved. The bucket's own worst drawdown gets 25% deeper. The 2022-style bear goes from +8.4% to +2.2% — the hedge stops hedging — and the reference portfolio's max drawdown goes from −7.0% to −9.8%. The only thing it buys is the war column: +13.8% in six months, because a levered basket of oil and softs is a wonderful thing to own while oil goes up.
That is the trade Simplify quietly made on every CTA holder's behalf this summer. They swapped a crisis hedge for a bet that the war goes on forever 😬
And the right column, which is the point. Move ZEUS to the new bucket at 40% and the August protection comes back: reference-portfolio downside capture 51.7% versus 54.4%, max drawdown −7.12% versus −7.04%, CAGR 13.83% versus 13.77%. Worst month better. 2022 bear better. We give up five points of equity exposure and get the August portfolio back, minus the fund that left. The bucket's own Sortino DOUBLES. Its max drawdown falls 38%. And it makes more money 🎯
The one thing ZEUS gives up is the war column — +6.0% instead of +13.8%. That is not a bug. A hedging bucket that makes 14% in six months when oil spikes is a bucket that loses 14% when oil un-spikes, which is exactly what CTA did in May and June: −19% in two months while the S&P went UP 4%. We keep a deliberate, unlevered, diversified slice of that bet through SDCI at 15%, and Part 6 explains why 15 and not 30.
On speed — why ZEUS is doing this in one session. Part 1 argued that gliding out of a concentrated position is the risky choice, not the safe one, because every month of the glide is spent holding the thing you decided to reduce. That argument was about volatility. This one is about identity. ZEUS is not reducing a position in a fund; we understand. We are exiting a fund whose model, managers, sub-adviser, market universe, sizing engine and position count ALL changed between July 7 and August 31 — and whose September 3 filing adds swaps to the toolkit, which means more change is coming. There is no version of "wait and see" where the thing we'd be waiting to see is the fund we bought. And the session happens to come with Brent back above $100 for the first time since July, which means we're selling the oil bet at a good print, not a bad one 😉
CTA's September 3 swaps filing signals the strategy is still changing. Simplify has not said what.
CTA is now a black box and has earned zero trust from investors since its actual track record began in August 2026.
Morningstar's 4-star rating is invalid. The 3-year window for a valid star rating on the NEW fund doesn't close until August 2029.
INTRODUCTION: SAME TICKER, DIFFERENT FUND — HOW A $1.5 BILLION HEDGE QUIETLY BECAME AN OIL BET 🤯
Carl Sagan liked to say that extraordinary claims require extraordinary evidence.
Here is an extraordinary claim: the largest position in the ZEUS hedging bucket — a fund with $1.5 billion of assets and 53 months of exemplary crisis-alpha history — is, as of this month, a fund with no track record at all. Not a fund in a drawdown. Not a fund whose strategy is out of favor. A DIFFERENT fund, wearing the old one's ticker 🤯
The evidence, it turns out, is extraordinary too, and almost all of it comes from Simplify itself. Three prospectus supplements were filed with the SEC between August 11 and September 3. The fund's own daily "Portfolio Risk Profile" pages. The CEO's statements to Bloomberg. The departing portfolio manager's own farewell post. And one GNG member who did the thing nobody at a fund company expects a retail investor to do: he downloaded the fund's risk profile every single trading day for twelve weeks and kept the files. What Michal Szafranski found is the spine of Part 2, and I want to say his name at the top of this report, not the bottom 🙏
I want to be precise about what this report is NOT saying. It is not saying CTA will lose money. A 1.27x levered basket of Brent, diesel, gasoil, copper, cotton, sugar, and cocoa may well make a fortune if the Strait of Hormuz stays shut for three years. That is a perfectly coherent bet, and Simplify's CEO may turn out to be right to make it. What this report IS saying is that the reason ZEUS owned CTA — the reason it had the ANCHOR slot in a hedging bucket — was a set of statistical properties that the fund no longer possesses and, on the evidence of its current holdings, is not trying to possess.
You do not keep a fire extinguisher because it might make a good flamethrower 😉🤔😂
Part 1 of this series argued that "Is CTA broken?" was the wrong question, because CTA was doing exactly what it was built to do. That is still true. What changed is what it was built to do. So the question for Part 2 is simpler, and harder: now what? 🤔
The candidates are the ones members raised — SDCI, which several of you have watched climb the Morningstar rankings; WTMF, the oldest managed-futures ETF on the market; and COM, because "trend on commodities" kept coming up and COM is the fund that actually does that. Each gets the treatment every asset got in Part 1: downside capture first, worst-quintile equity months second, the 2022 bear third, and only THEN returns. The most important number in any hedging bucket is what it does when stocks go down. Everything else is a bonus collected on the side.
PART 1: THE WAR, THIRTY DAYS ON
Part 1 dated the war on day 164. Today is day 194. In the month between, it worsened in every dimension that matters to an energy hedge, and the market's own estimate of when it would end moved further out. Again.
The short version: drones hit two ADNOC tankers in the Strait in mid-August. A Greek bulk carrier's chief engineer was killed on the 18th. On September 6 and 7 the IRGC launched ballistic missiles at US destroyers two days running, and on the 8th CENTCOM destroyed five Iranian tankers — ten for the week — after Iran fired twenty missiles at a US base in Jordan. On September 9, Brent traded through $100 for the first time since late July. Qatar is warning of an "industrial catastrophe" if the crisis continues, which is a remarkable thing for Qatar to say out loud. GlobalSecurity's daily OPREP and the ABC and Euronews live blogs carry the details.
Is the Strait open? Same answer as August, slightly worse. About 7 million barrels a day are transiting against 20 million before February 28 — roughly a third. A senior US official described the Strait as "fully open and under US Navy control," a statement that can be both true and meaningless at the same time. Open to a destroyer is not open to a VLCC whose insurer has withdrawn 😐
The odds. Part 1 reported Polymarket's probability that Hormuz traffic would return to normal by December 31 at 48%. On September 8, it was 25% of $10.6 million in volume. That is the sixth consecutive month in which prediction markets have revised the reopening date to a later date. I said in August that this pattern should change how you think. It hasn't stopped 📉
And here is the oil path in the BNO fund's own monthly returns for 2026: +16.4%, +5.6%, +49.4%, +12.5%, −13.6%, −19.6%, +23.8%, +4.9% — and +9.3% in the first week of September.
That series is the whole reason this report exists. Whatever is long, that series with leverage is not a hedge. It is that series, with leverage. Hold the thought. Part 3 shows it happening to CTA month by month 🎲
PART 2: WHAT HAPPENED AT SIMPLIFY — THE RECORD
There is a version of this story told in headlines, and there is the version told in filings and holdings files. The second one is better, because nobody at Simplify has yet told the first one.
The paper trail. On August 7, Michael Green and Paisley Nardini ceased to be CTA's portfolio managers. On August 9, Green announced on his own Substack that he was leaving Simplify to found Tier1 Alpha — he'd joined in April 2021 when the firm managed $200 million; it manages $14 billion today. On August 11, an SEC 497 supplement removed all references to Green and Nardini from the prospectus. On August 31, a second 497 stated that "Altis Partners has concluded its services to the Funds as the Futures Adviser" and that Simplify "will continue to manage each Fund in accordance with its current strategy using its proprietary models." On September 3, a third 497 added that "the Fund may use swaps as a supplement to its futures-based strategy." On September 4, Bloomberg reported that Simplify "began intervening in the product sometime around March," that the oil build started in April, that positions fell from "usually north of 100 to under 50" from July, and that 42% of the fund's risk sat in Brent with a further 33% in diesel. On September 8, Seeking Alpha's Jack Bowman — a longtime CTA bull — downgraded to Hold and said he was divesting.
Three things in that sequence deserve underlining ✍️
First, the ORDER. The intervention began in March. The portfolio managers left in August. The sub-adviser's termination was filed at the end of August. The swaps language arrived in September. The strategy changed first, and the paperwork followed — which is legal, because CTA's prospectus has always allowed discretion, and is also EXACTLY the opposite of how an investor would want to learn about it.
Second, the phrase "in accordance with its current strategy using its proprietary models." Read literally, that sentence claims continuity. Read against the holdings — next paragraph — it cannot mean continuity of the Altis model, because the Altis model was terminated. It can only mean Simplify's own model, which has existed inside this fund since, at the earliest, March 🤔
Third, the missing document. The only public commentary on any of this is a July 23 Simplify webinar in which the portfolio manager described "recent tweaks" — "less vol constrained," "removed that barrier," "more concentrated, intentional by design." That is a description of an intention, not a model. There is no announcement, no fact sheet, no methodology note describing what the new model is, what it trades, how it sizes, or why it holds no rates and no shorts. Everything anyone knows about the new CTA, they know from watching it .
The holdings — 42 days of Simplify's own data. Michal downloaded the fund's Portfolio Risk Profile from simplify.us every trading day from June 12 to September 3 and reconstructed what the fund actually held. The report is his; the interpretation is ours; his numbers are quoted as published.
The headline: in twelve weeks, the fund went from 47 long/short positions across nine sectors — 226% gross exposure, 120% of it in interest-rate futures — to eight long-only commodity positions, 127% gross, ZERO in rates. It happened in three jumps, not a glide. July 7 to 9: the entire long leg in US and Canadian bonds closed, the Gilt cut by two-thirds, net rates from +27% to −65% in two sessions. July 17 to 21: every interest-rate contract and 26 of 40 positions liquidated, gross exposure from 174% to 81%. August 5: the last short closed. August 10 to September 2: leverage rebuilt with long commodities only, from 74% to 141%.
Simplify's own page on September 4 shows the same book: Brent 42.4%, gasoil 16.4%, ULSD 16.3%, copper 16.1%, cotton 12.9%, sugar 11.2%, cocoa 10.1% — seven futures totaling 125% of NAV, sitting on 71% in Simplify's own money-market ETF as collateral.
Five things a trend model does not do. This is the part that settles the "is it still the Altis model, just run by someone else?" question, and it comes straight from the snapshots.
One: position sizes are quantized. Since August, new positions have entered in blocks of about 5% of NAV and then doubled — cotton 5.0 → 15.5, sugar 5.0 → 7.5 → 10.6, copper 8.5 → 15.2. In June, coffee grew 0.4 → 0.9 → 1.6 → 2.2 → 2.8 → 3.0 → 3.5 → 4.0 → 4.3 → 4.6 → 5.1 → 6.3%, about half a point a session. Two sizing behaviors mean two models.
Two: Brent was PINNED at 49.5–50.0% of NAV for five consecutive snapshots, August 12 through 24, while oil moved. A constant notional weight during a moving market means daily buying and selling back to a target. That is target-weight sizing. It is not volatility-scaled trend following.
Three: the whipsaw. Cocoa entered August 5, doubled August 10, exited August 12, re-entered August 27, doubled August 31. Coffee entered August 24, exited August 27 — two sessions. Grains built to 20% of NAV between August 20 and 31 and were fully exited on September 3, seventeen points in a DAY. Between June 12 and July 6, by contrast, ONE position out of 47 was closed. A medium-term trend model does not behave like this. A swing trader does.
Four: zero rates since July 21 — seven weeks — in a fund whose prospectus describes "commodity and interest rate futures." Zero shorts since August 5 — five weeks — in a "long/short" fund.
Five: Brent has been the largest single contributor to volatility in all 42 snapshots. Eighteen percent of the fund's risk at the June low, sixty percent at the July 30 peak. Energy as a sector went from 49% of the fund's volatility to 68%. At no point in this period was the fund a diversifier independent of oil — including in June, under the old model. Part 1 measured that dependence at 65% of variance and called it the fund's central weakness. It is now the fund's DESIGN.
Michael's one-sentence conclusion, which I cannot improve on: "Nothing in the data suggests 'the Altis model without Altis': sizing method, turnover, and market universe all changed at the same time in July. The termination of Altis on 31 August formalized something visible in the book since 21 July."
If peers are short things, including rates and some commodities, then CTA has become a long-only (leveraged) commodity fund. SDCI has an 8-year track record of doing something similar. So if we have a 99th-percentile ETF with an 8-year track record of beating CTA at its new game, why take the risk of owning CTA? If Peter Lynch announced he was retiring from the Magellan Fund and Fidelity replaced him with Cathie Wood, who proclaimed that "Tesla is GARP because the bear case is $10 trillion market cap in 10 years...because of Robot Taxis!"🤔 😂 Sorry, but we invested to have our money managed by the king of GARP (growth at a reasonable price), not the queen of hyperbolic models 🤣.
A quant fund is its model. The model that produced CTA's April–December 2022 — +15% while the S&P fell 14% — and its 2024 −115% downside capture over 47 months was Altis Partners' model, run by two portfolio managers who are gone, under a sub-advisory agreement that is terminated. What runs the fund now is a proprietary Simplify model that nobody outside Simplify has seen. It may be brilliant. It has no track record. And CTA's track record is not its track record.
PART 3: CTA IS A BRAND-NEW FUND — THE NUMBERS
Holdings tell you what a fund owns. Returns tell you what it IS. Here is CTA measured as two funds, because that is what the data says it has been.
The fingerprint changed. Over the Altis era, April 2022 through February 2026, CTA's downside capture against the S&P 500 was −114.7% — it went UP more than the market went down. Its correlation to stocks was −0.37. Its volatility was 15.5%. And here's the number nobody looks at: its R² to the broad commodity index — the share of CTA's monthly variation explained by "commodities went up or down" — was 2%. It was a multi-asset trend fund that happened to trade commodities.
Over the intervention era, March through August 2026, downside capture is +79.9%. Correlation to stocks +0.24. Volatility 27.1%. And R² to the commodity index: 65%. Correlation to Brent went from 0.30 to 0.62. It has become a commodity fund that happens to have a managed-futures name 🦎
Six months is a short window, and those numbers carry wide error bars. But they point in the same direction as every holdings snapshot, and the month-by-month tells the story without statistics. In March, the war month, Brent rose 49%, and CTA made 0.4% — because the intervention was just beginning and the big oil position hadn't been built. Then look at May and June: CTA lost 19% over two months, while the S&P rose 4%. That is the signature of a fund that is long oil, and nothing else, with leverage. A hedge is supposed to be inversely correlated to the thing it's hedging. CTA in May and June was inversely correlated to the HEDGE 🤔🤬
The crash test. To put the new CTA into a bucket, you need a return series longer than six months, so we built one from what it holds: 1.05 times the broad commodity index, plus 0.11 times Brent, plus 0.11 times the petroleum-products basket, less the collateral yield that leverage doesn't earn — 1.27x gross, three-quarters petroleum. Over May through August, that proxy's monthly correlation with actual CTA is 0.96. It is not a model of Simplify's engine, which nobody has seen. It is a faithful model of exposure, and exposure determines what happens in a crash.
Run that book back through every crash we have data for. Q4 2018: S&P −13.6%, CTA-today −25.7%. COVID, February–March 2020: S&P −19.6%, CTA-today −31.5%. 2022 bear, April–December: S&P −14.3%, CTA-today −6.3%, and the ACTUAL Altis-era CTA +15.0%. Tariff shock, February–April 2025: S&P −7.5%, CTA-today −9.4%, actual CTA −1.1%. War shock, February–March 2026: S&P −5.7%, CTA-today +29.4%. Oil bust, May–June 2026: S&P +4.2%, CTA-today −19.6%, and actual CTA −19.0%. The proxy and the real fund agree on a point in common in the one window where both exist. That's the validation.
Over the full window since June 2018 — 99 months — the current book's downside capture is +61.8%. The fund ZEUS owned captured −115% of equity declines. The fund ZEUS owns now captures +62% of them: it goes down about two-thirds as much as the market, then adds a −31% COVID and a −26% Q4 2018 on top. The war shock is the only window it wins, and it wins there by being what it is.
Positive downside capture is the definition of "not a hedge." That's the whole sentence. Everything else in this report is a footnote to it.
"Is this the new KMLM?" Part 1 asked whether CTA was becoming KMLM, the trend index that "just sucked" for four years. It's a fair worry, and the answer is instructive. KMLM never changed species. Through its bad years, it remained a rules-based, multi-asset, long/short trend index with a rates book; it underperformed because the environment punished its rigid look-back, and it snapped back +44.8% in the 2022 bear market precisely BECAUSE it never stopped being what it was. It’s September 1; holdings are the same animal they've always been. CTA didn't underperform its category. It LEFT its category. A fund in a drawdown can recover. A fund that has changed what it is cannot return to what it no longer is.
PART 4: SDCI — WHAT IT IS, WHO BUILT IT, WHAT IT ISN'T, AND "TOP 1% OF WHAT?"
Several members have watched SDCI climb the Morningstar rankings this year — top 1% of its category over three and five years — and asked the obvious question: is THIS the fund that replaces CTA?
Short answer: it is the fund that replaces what CTA has BECOME, at a fraction of the risk and cost. It is not the fund that replaces what CTA WAS. Here's why, in the fund's own documents and then in its numbers. And because most of you are meeting this fund for the first time, let's start with who made it.
Who built it, and why is the design worth taking seriously? SDCI is the USCF SummerHaven Dynamic Commodity Strategy No K-1 Fund. USCF is United States Commodity Funds — the people behind USO, the world's largest oil ETF — so the plumbing is run by a firm that has been trading commodity futures for two decades. The INDEX is the interesting part. It's designed by SummerHaven Investment Management, whose co-founder is Geert Rouwenhorst, a Yale finance professor, and the strategy rests on two papers he co-wrote that changed how the industry thinks about commodities. The first, with Gary Gorton in 2006, showed that commodity futures, as an asset class, had delivered equity-like returns over decades, with negative correlations to stocks and bonds and positive correlations to inflation. The second, with Gorton and Fumio Hayashi in 2013, showed WHY some commodities beat others: inventory levels. When inventories are low, the futures curve goes into backwardation — near-dated contracts trade above later ones — and those are the commodities that go on to outperform. Low inventory means scarcity; scarcity means the market pays you to hold the front of the curve. SDCI is that finding turned into a rulebook 🤓
The rules. Launched May 2018. Expense ratio 0.60% against CTA's 0.75%. About $650 million in assets. From a universe of 27 commodity futures, each month the index picks 14 — the seven with the greatest backwardation, then seven more by momentum — equal-weights them at about 7% each, chooses the most backwardated contract on each curve, and rebalances monthly. Long only. No short book, no rates book, no currency book, no equity book, no leverage. And the "No K-1" in the name is a practical gift: most commodity ETFs are structured as partnerships and send you a Schedule K-1 every spring, which is a tax-season headache. SDCI holds its futures through a subsidiary, so it sends a plain 1099 like a stock fund. If you've ever cursed at a K-1 in March, you already understand why that's in the name 😂
Equal-weighted value + momentum for commodities?
Growth + value?
That's what we call elegant, evidence-based simplicity🤔
The rationale is elegant and well-supported: it's a systematic way to own the tightest markets in the world. In a world where the tightest market is oil because a strait is closed, it will own a lot of oil.
What it owns this month — and the overlap that should give every CTA holder pause. SDCI's fourteen for September: five energy contracts at 36% of the fund, four metals at 29%, three livestock at 21%, two ags at 14%. Now put CTA's September 4 book next to it. Four of CTA's seven positions — Brent, ULSD, gasoil and copper — are in SDCI's fourteen. SDCI holds those four at about 7% each, 29% of the fund. CTA holds them at 42%, 16%, 16%, and 16%: 91% of NAV in the same four contracts, three times SDCI's weight, PLUS leverage, MINUS the other ten markets 🤯
CTA today is SDCI's four favorite trades, tripled, with the diversification removed. That's the cleanest way I've found to say it. And it answers the question a couple of members asked — "how realistic is it that Simplify can pull off SDCI's returns?" — with a different question: why would anyone pay 0.75% for a discretionary, 1.27x-levered, seven-market, 75%-petroleum version of a bet that a 0.60%, rules-based, unlevered, fourteen-market fund already makes with a published methodology, a Yale professor's name on the index, and eight years of history? For ZEUS, if we want the bet, SDCI is the better instrument. The only thing Simplify offers that SDCI doesn't is leverage and a new model. We're not paying for either 😉
What it has done — and why the window matters. This is where "top 1%" needs its denominator. Start SDCI's history in April 2022 — a commodity bull market with an inflation-driven bear in stocks — and its downside capture is −20.9%. It went up when stocks went down. It looks like a hedge. Start in June 2018, so the sample contains Q4 2018 and COVID, and its downside capture is +22.0%. It went DOWN when stocks went down, with a −40.6% peak-to-trough. Same fund. Same rules. The only thing that changed is whether the sample contains a deflationary crash.
—
In the ten worst S&P months since 2018, SDCI fell with the market in seven. March 2020: S&P −12.4%, SDCI −16.9%. The three that didn't fall in — April 2022, December 2022, March 2025 — were inflation scares and a tariff scare, not growth scares. DBMF, over the same months, was positive in five of the seven it was alive for and never lost more than 2.1%. That is the difference between a trend fund and a commodity fund, and it is the WHOLE reason DBMF takes the anchor in ZEUS and SDCI takes a sleeve.
Regress SDCI's monthly returns on the broad commodity index and you get a beta of 0.82 with an R² of 0.82 — four-fifths of everything SDCI does is "commodities went up or down." Its correlation with DBMF is 0.19. SDCI is a commodity beta with a clever tilt. The tilt is real: it has beaten the index by 2.8 points per year since 2018, and that’s worth paying for. But you cannot tilt your way out of a deflationary bust when everything you own is long.
So what is SDCI FOR? For exactly the scenario I've been worried about since March: the war goes on for years, the Strait stays effectively shut, and energy prices grind to new highs. In that world, a bucket made only of trend funds will make SOME money — DBMF and KMLM made +3.7% and +8.6% in the war shock — but a long commodity sleeve will make a lot more, and it will make it without the 27% volatility, the 5% blocks, and the discretionary cocoa trades. SDCI is the honest way for ZEUS to hold that view. It is also the honest SIZE for it, because the same instrument that makes +25% in a six-month oil spike makes −22% in a two-month deflationary bust. Part 6 has the dial.
Part 4 verdict: SDCI is a superb long-only commodity fund and a poor bear-market hedge, and both are true at once. In ZEUS, it replaces the war bet, simplifying smuggling into CTA at lower cost, lower risk, wider diversification, and with a methodology anyone can read. It does not replace the crisis hedge Altis was running — nothing long-only can. Fifteen percent of the ZEUS bucket. Not thirty. Labeled "war sleeve," so nobody forgets what it is.
PART 5: WTMF AND COM — ONE REJECTED, ONE RESPECTED
WTMF has the same disease as CTA — it changed species in 2021. The WisdomTree Managed Futures Strategy Fund launched in January 2011, making it the longest-lived ETF in the category and the reason it keeps coming up: fifteen years of history looks like fifteen years of evidence. It isn't. On June 4, 2021, WisdomTree rebuilt the fund: equity index futures added at up to 40% of the fund, currencies cut, and a VIX-and-credit-spread "macro indicator" that rotates the whole fund between long-only and long/short. WisdomTree's own explanation: "The exclusion of equity futures in WTMF going back to its inception has left performance on the table." As of September 8, the fund holds 32% in equity futures.
So the ten years before June 2021 describe a different fund — sound familiar? — and the five years since describe this one: downside capture +14.8%, correlation to the S&P 0.49, and −7.9% in the April–December 2022 bear, the one year in a decade when managed futures were supposed to shine, and every other trend fund in the ZEUS bucket made 10 to 15%. A fund that carries a third of its book in stock index futures has a 0.5 correlation with the stock market, and a 0.5 correlation is not a hedge. It's the rare candidate that loses on BOTH axes. Rejected — and Part 6 shows the one test where it LOOKS like it wins, and why that's the trap.
COM is what "trend on commodities" actually looks like. Because that phrase kept coming up in chat as a description of SDCI, we tested the fund that literally does it. The Direxion Auspice Broad Commodity Strategy ETF: twelve commodities, each held long when trending up and moved to cash when not trending up. Long/flat, never short. Compare COVID — SDCI −21.6%, the commodity index −22.8%, COM −5.9%. That is what the "flat" in long/flat buys: a parachute. And 2022: COM +7.5% while stocks fell 24%. COM is commodity exposure with an EXIT, and for anyone who wants commodity trend rather than commodity beta, this is the fund to use, not SDCI.
Why it isn't in the ZEUS bucket: it's dominated. Put COM in SDCI's 15% slot, and the bucket's downside capture goes from −41.7% to −39.0% AND its CAGR from 12.2% to 10.5% — less protection and less return, because COM's cash-heavy years drag and its trend rules react a month or two late to the whipsaws that have defined this war. A fund that makes 8% a year with a +3% downside capture is a fine thing to own and a mediocre thing to hedge with. Respected, not selected.
PART 6: THE RE-OPTIMIZED BUCKET — DOWNSIDE CAPTURE FIRST
Same method as Part 1, same window — April 2022 through August 2026, the only one in which every sleeve, including gold, is alive — same ordering. The CTA sleeve is modeled in two ways: as the history it had, which no longer applies, and as the book it holds today.
Every candidate, alone. The top three hedges on downside capture are the three trend funds ZEUS already owns — Altis-era CTA at −114.7%, KMLM at −78.9%, DBMF at −50.2% — and two of them are still at the same levels. CTA-as-it-is drops from first place to NINTH, at +4.6%, below Brent, below SDCI, below a long/flat commodity index, with a 2022 bear LOSS. And here's the thing that jumps out of the table: the war column is a mirror image of the capture column. Everything that made money in the war has poor capture; everything with great capture made little in the war. That is not a coincidence. It is the whole problem, and the bucket is how you hold both.
Sixteen buckets, sorted by what matters. We ran every allocation across SDCI, DBMF, KMLM, IALT, gold and WTMF in 5% steps — 53,130 portfolios — and reported the most negative downside capture at each CAGR floor. The purist bucket, D — DBMF 50 / IALT 25 / KMLM 15 / gold 10, no SDCI at all — gets −47.1% downside capture and 10.4% CAGR. The recommended bucket, A — DBMF 40 / IALT 25 / SDCI 15 / KMLM 10 / gold 10 — gets −41.7% and 12.2%. Move from D to A, and you give up 5.3 points of capture for +1.7 points of CAGR, a 35% smaller hedge drawdown, a Sortino of 2.61 versus 1.44, and twice the war return. Part 1 found that every point of hedge CAGR cost about eleven points of capture. With CTA gone, it's about three. That's the SDCI-and-DBMF trade, and it's the one ZEUS is making.
Why A and not more SDCI? Because the war column isn't free, and the April-2022 window hides the bill. Starting from the CTA-ejected bucket and adding SDCI-funded pro-rata: every 10% of SDCI costs about two points of capture, buys about half a point of CAGR and two points of war return — cheap on this window. But every 10% ALSO costs about 2.2 points in a deflationary crash, which this window doesn't include.
So here are the same finalists over the longest window in which the anchor exists — June 2019 through August 2026 (87 months), the one that contains COVID. Bucket D loses 1.3% in the two COVID months. Bucket A loses 4.5%. SDCI at 30% — in CTA's old slot — loses 7.7%. The August bucket with CTA-as-it-is loses 11.0%. The S&P lost 19.6%. Where does A's 4.5% come from? DBMF −0.3 points, IALT proxy −0.8, SDCI −3.2, KMLM flat, gold −0.1. The war sleeve is three-quarters of the bill, exactly as the dial predicted, and the rest of the bucket does what it's there for.
At 15%, the bucket can carry the war sleeve through a COVID at −4.5%. At 30%, it's the worst month the bucket has ever had, by a factor of 2.5, in the EXACT scenario a hedging bucket exists for. Fifteen is not a preference. Fifteen is the largest war sleeve the bucket can carry and still be a hedge.
The sleeve members will doubt why the "broken" fund keeps its 10%. I know what several of you are thinking, because I thought it too. KMLM by calendar year: 2021 +7.0%, 2022 +30.6%, then 2023 −5.6%, 2024 −1.7%, 2025 −3.0%. Three losing years in a row, a −22% peak-to-trough, a 2.9% CAGR over the window. Why is a fund with THAT record still 10% of the bucket and 4% of ZEUS, when WTMF — a fund with a smoother line — is out?
Because a hedge isn't judged by its calendar years. It's judged by what it does in the months when stocks fall. In the eleven worst-quintile S&P months of the window, KMLM was positive in ten of eleven — the one exception was −1.4% in June 2022 — averaging +3.73%, the best of anything the bucket owns. It made +44.8% in January–September 2022, while the S&P fell 24%. It made +10.5% in this year's war. Its correlation to the S&P is −0.51, the most negative in the bucket, and its correlation to DBMF is only 0.60, so it's not a duplicate. It is the sleeve that shows up in the exact months a hedge exists for, and it pays for that with dead years in between. That is what deep-crisis insurance looks like on a statement. Three losing years are the premium. +44.8% was the payout.
12.8% CAGR historical returns from the entire hedging bucket, SO don’t assume that everything you own has to do well for the bucket to serve its purpose AND make market-like returns (but with a -41% downside capture ratio).
Now WTMF, and here is the trap. Put WTMF in KMLM's slot, and the BUCKET looks better: Sortino rises from 2.62 to 3.17, the bucket's max drawdown improves from −4.7% to −3.5%. If you were optimizing the bucket in isolation, WTMF wins. But WTMF is a third stock-index futures with a 0.49 correlation to the S&P. It makes the bucket smoother by making it MORE LIKE THE THING THE BUCKET IS HEDGING. That's the Sharpe trap from Part 1 in a different costume. In the same eleven worst equity months, WTMF averaged −1.00% and was negative in seven of them.
The only test that counts is the whole portfolio, because that's what pays the bills. On ZEUS's actual seven stocks at 60/40: with KMLM at 10, max drawdown −12.7%, worst month −10.4%, 2022 bear −16.7%. Swap in WTMF: max drawdown −13.2%, worst month −10.9%, 2022 bear −17.4%, downside capture worse, for no gain in return. WTMF makes every protection number WORSE the moment you attach the bucket to the equities it's supposed to protect. Rejected on the only test that counts. And KMLM at 10 versus zero is a wash on every portfolio column — CAGR 23.2% versus 23.4%, drawdown −12.7% versus −12.8% — so the tie goes to the sleeve that was positive in ten of eleven bad months 🛡️
KMLM is a bad investment and a good hedge. WTMF is a decent investment and a bad hedge. A hedging bucket is built out of the first kind. That's the whole lesson in one line, and it's why you optimize the portfolio, not the bucket, and why recent returns are the LEAST useful thing to know about a hedge.
Why gold stays at 10%. With CTA gone, gold's empirical case on this window gets weaker, not stronger: each 5% costs about 2.6 points of capture and 1.3 points of war return, buying half a point of CAGR. It stays for the reason it was there in Part 1 — the scenarios this window does not contain, a Taiwan blockade, a dollar-debasement response to an AI-credit unwind, are the ones in which bullion is the only thing in the bucket that works, and 10% is the price of that insurance. Why does IALT stay at 25%? It's at its cap — 25% of a 40% bucket is exactly 10% of the portfolio, the limit Part 1 set because the fund has eight months of live history. Why does DBMF take the anchor? It's the only trend fund in the bucket with a COVID record, −0.8% in the two months. Its downside capture is −50%. And a replicator that averages 20 CTAs is the closest thing to "the old CTA" that still exists.
How big — the sizing decision. Removing CTA removes the bucket's most volatile sleeve: bucket volatility falls from 8.3% to 7.1%, and a quieter hedge protects less per dollar. So the question isn't "is the new bucket better than the old one" — at the same 35% size it's slightly worse on reference-portfolio drawdown, −7.80% versus −7.04% — but "how big does the new bucket have to be to give the portfolio back what it had?" At 60/40: reference-portfolio downside capture 51.7%, max drawdown −7.12%, worst month −4.99%, 2022 bear −5.76%, CAGR 13.83%. Every row at least as good as August. Forty is the point where every protection statistic is back, five points of equity is a clean instruction, and the CAGR doesn't move. Had we stayed at 35%, we'd have kept most of the improvement and accepted −7.8% instead of −7.0%. Either is defensible. ZEUS chose 40.
What the recommended bucket has actually done. 2022 (April–December) +6.1%. 2023 +0.7%. 2024 +13.4%. 2025 +17.4%. 2026 through August +16.7%. It’s 2026 by month: +4.8, +5.1, +0.4, +2.4, −0.3, −2.6, +3.1, +2.9. Worst month in 53: June 2026, −2.6%. In the eleven worst-quintile equity months of the window, it was positive in nine, with the worst reading at −0.9%. That is what a hedge is supposed to look like: boring in the good months, present in the bad ones.
PART 7: WHAT ZEUS IS DOING WITH CTA RIGHT NOW, AND WHAT WOULD BRING IT BACK
ZEUS sells it. All of it. This week.
Part 1 tested every start date and every glide path and found that moving out of a concentrated, high-volatility position is the one case where dollar-cost averaging works AGAINST you. That argument stands, and this case is stronger on every count. There is nothing to average into — a glide is a bet that the thing you're leaving is roughly the same thing next month, and CTA's book, model, managers and sub-adviser all changed in eight weeks. The risk being reduced is not just volatility, it's SIGN — the correct weight for an asset with positive downside capture in a hedging bucket is zero, and there is no glide path to zero that beats zero. We're selling at a good print — Brent through $100 on September 9, CTA up about 12% since July 1; selling a long-oil fund into an oil spike is the opposite of panic selling. And liquidity is a non-issue: $1.5 billion of assets, a tight spread, one trade.
The one argument for waiting is that Simplify might publish the new model tomorrow, and it might be brilliant. If so, it will still be brilliant in a month, and ZEUS can buy it back with a track record that's one month longer than zero 😉
On keeping a token position. Some members will want to keep a slice to watch the new model from the inside. The numbers say it's expensive: every 5% of the bucket left in CTA-as-it-is costs about two points of downside capture for no gain in CAGR. If someone does it anyway, the arithmetic says cap it at 5% of the bucket, fund it from SDCI's slot rather than DBMF's — they're the same bet, so it's only trading some of the unlevered version for the levered one — and call it tuition, not a hedge. ZEUS is holding zero.
Would ZEUS ever own CTA again? Let me answer that the way Khan answered Kirk: "From hell's heart I stab at thee" 🖖😂
Here's the calm version. CTA's 53-month track record belongs to a fund that no longer exists. The fund that wears its ticker now has a track record that started in August 2026 — zero years, zero crashes, zero bear markets, run by a model nobody has seen. Meanwhile, SDCI does exactly what the new CTA is trying to do — long commodities, curve-first, systematic — with a published rulebook, a Yale professor's research behind it, an eight-year record, and a top-1% ranking over three and five years. Why would anyone pay 0.75% for the unproven, levered, discretionary version of a bet that a 0.60%, rules-based fund already makes better?
So the honest answer isn't "here's the checklist for bringing it back." It's "CTA starts from zero, and zero doesn't get a slot." If Simplify ever publishes the model, restores the rates book and the short book, sizes positions like a quant instead of a swing trader, and then runs it for YEARS through a real bear market — then it becomes a new fund with a record, and it can compete for a slot on that record like anything else. Not six months. Years. Michal's watch-list tells you what to look for if you want to watch. I'll be watching SDCI instead 😉
Until then, the 53-month history stays where it belongs: in the archive, next to Altis Partners' name
EVERY QUESTION YOU'RE GOING TO ASK
"Isn't this an overreaction to six months of data?" The six months are the least of it. The case rests on the fund's own filings — sub-adviser terminated, managers gone, "proprietary models" — on 42 days of its own published holdings, and on the CEO's statement to Bloomberg that the firm has been intervening since March. The returns merely agree with the documents. If CTA had kept its book and changed its name, ZEUS would have kept it. It kept its name and changed its book.
"CTA made +4.7% in August and +2.4% so far in September. Why is ZEUS selling a fund that's working?" Because Brent made +4.9% and +9.3% in the same two windows. The fund is "working" like a long oil position when oil rises. ZEUS isn't selling a fund that's working. ZEUS is selling a fund that is currently being PAID for the risk we don't want.
"SDCI is in the top 1% over three and five years. Doesn't that settle it?" It settles that SDCI is the best way to go long commodities. Being the best fund in a category that falls with the stock market doesn't make you a hedge. It makes you the best passenger on the same bus.
"KMLM has lost money three years running. Why is it still 10% of the bucket?" Because in the eleven worst equity months of the window it was positive in ten, averaging +3.7% — the best in the bucket — and made +44.8% the last time the market fell for eight straight months. A hedge is judged by the bad months, not the calendar years. At the whole-portfolio level, keeping it versus dropping it is a wash, so it stays for the months it exists. A member who can't stand it can put its 10 into DBMF at almost no cost. That's the honest answer, and it's also why it shouldn't go anywhere else.
"What about the purist bucket D — no SDCI at all?" It's a legitimate answer for anyone who weighs a deflationary bust more heavily than a forever war. Five points more capture, no commodity beta, −1.3% in COVID, where A lost 4.5%, and it costs 1.7 points of return a year and half the war upside. If the Strait reopens and oil goes to $65, D looks smarter. If the war runs for three years, A does. ZEUS chose A because the base rate — Part 1's Tanker War and Red Sea analogs, six straight months of the prediction markets pushing the date out — says the war is the more likely world. But D is not wrong.
"What if Simplify announces the new model next week and it's good?" Then it's a brand-new fund with a published model and zero years of history, and it can come back in three to five years with a record that includes a real bear market — the same bar every other fund in the bucket had to clear. Six months proves nothing. A good year proves nothing. KMLM's three losing years are the whole reason it's in the bucket. That's what a hedge is judged on, and CTA hasn't been through one yet.
"Isn't DBMF just a replicator? Can it really be the anchor?" That's exactly why it can. DBMF replicates the average positioning of the twenty largest managed-futures programs. Its downside capture is −50% on the Part 1 window and −36% on a window that includes COVID, in which it lost 0.8%. It is the closest thing to "the old CTA" that still exists, because it's the average of twenty funds that still do what the old CTA did.
"You keep saying CTA's history is 'void.' Isn't Simplify saying it's the same strategy?" The August 31 filing says the fund is managed "in accordance with its current strategy using its proprietary models." The prospectus language about "commodity and interest rate futures" and "long/short" is unchanged. The book has held no interest-rate futures since July 21 and no shorts since August 5. Both statements can be legally true at once. Only one of them describes a hedge.
If a Michelin-star chef quits and takes the recipes, the owner will swear the food is unchanged — the sous chef "was in the kitchen the whole time." Then you order the tasting menu, and it's a burger (and a NON-VEGAN burger at that)!. 😂🤬
Same sign on the door. Different restaurant 🤔😉
"Is this a recommendation that I sell CTA?" No. It's a report on why ZEUS is, with every number and every source so you can check the work. Your account, your tax situation, your risk tolerance, and your read of the war are yours. GNG publishes research; it doesn't manage your money. Read it, audit it, decide.
CONCLUSION: THE HEDGE LEFT WITH THE QUANTS. HERE'S WHAT STAYS
In August, the question was whether a fund could simultaneously be the best crisis hedge ZEUS owned AND a leveraged bet on Brent crude, and the answer was yes, at 30%, because the hedge was still in place. In September, the hedge is not there. It left with the quants.
What's left is a $1.5 billion vehicle running a model nobody outside Simplify has seen, holding seven long positions that are three-quarters petroleum with a quarter-turn of leverage, sized in blocks, pinned to targets, and traded like a swing account. It may be a fine way to bet on a forever war. It is not a hedge; its history is not its own, and the ZEUS hedging bucket has no slot for it.
The replacement is not exotic. The two trend funds ZEUS already owned are still trend funds, and they take the anchor. The market-neutral engine is at its cap. Gold keeps its scenario insurance. KMLM keeps the slot it earns in the bad months and pays for in the good ones. And the one new idea — SDCI — turns out to be the honest, unlevered, fourteen-market version of the bet Simplify smuggled into CTA, built on a Yale professor's inventory research, which is why it gets a labeled 15% war sleeve and not the 30% anchor slot. Downside capture −41.7%. Hedge drawdown under 5%. Twelve percent a year realized, seven planned. And a portfolio whose protection is back where it was — at 40% instead of 35%, because when you take the loudest instrument out of the orchestra, the orchestra has to be a little bigger 🎻
Sagan said we are a way for the cosmos to know itself. Sometimes the way is a filing. Sometimes it's a fund page. And sometimes it's a member who downloaded a risk-profile page forty-two days in a row because he wanted to know what he actually owned... and found out that so did we 🫂
Check the holdings, not the ticker.
Wonder, audited.
Sign in to leave a comment and join the discussion.
Sign Up Free