1987. 2008. 2020. Different crises, same ending: everything fell together. On paper, unrelated assets keep a portfolio safe; one falling is not supposed to drag the others. Then a real crisis arrives, and the correlations start to snap toward one. October 1987: markets in every major country crashed together in a single month. 2008: stocks, corporate bonds, real estate, and commodities fell together. March 2020: for a few days even gold and Treasuries fell with everything else. The reason is simple. In a crisis, assets stop being connected through their fundamentals and start being connected through their owners. Selling in calm periods is a choice; selling in stressed periods is often forced: margin calls, risk limits, redemptions. A forced seller sells whatever someone will still buy, so the pristine asset goes to cover the broken one, and a holding stays uncorrelated only until it shares an owner with something that is collapsing. Correlation, under enough stress, is a property of balance sheets: everyone's hedge turns into someone else's margin call. None of this makes diversification worthless. It is real, and it is worth having. But it is a statistical defense: it describes how assets have behaved, not how they will behave in the future, and in the moments when the whole market becomes one trade, the statistics loosen their grip. That is why protection written into the instrument itself matters. It is what RiskOFF is: a claim on BTC or ETH with a floor, paid for by giving up upside past a cap to RiskON. Its protection does not depend on correlation. It does not need bonds to zig when stocks zag. The floor comes from its structure alone, and structure does not get a margin call in a crisis.

The oldest option trade on record was placed over 2,500 years ago, by a philosopher, with pocket change, over olive oil. Aristotle tells it in the Politics. Thales of Miletus, tired of hearing that philosophy makes no money, read the winter sky and expected an unusually large olive harvest. He did not buy olives. He paid small deposits to reserve every olive press in Miletus and Chios for the coming season. If the harvest failed, he would lose only the small deposits. The harvest came in enormous, everyone needed presses at once, and Thales rented them out on his own terms. The story gets remembered as a forecasting win. The real invention was the structure. Thales capped his downside at the deposit and left his upside open. He was not predicting with more confidence than anyone else; he was shaping the payoff. That asymmetry, the right without the obligation, is the seed of every option product built in the two and a half millennia since. Twenty-five centuries later, we are building the risk layer of crypto, still the same business: turning risk from something you fear or avoid into something you exploit for alpha.

First a win at Paris Blockchain Week, now at @Arbitrum Founder House London. It is becoming clear: The Risk Layer is coming—a new primitive that turns risk from something to fear or avoid into something to harness for returns. And there is no better way to learn a new primitive than to trade it directly. RiskON and RiskOFF are live on Testnet: a BTC or ETH deposit split into two tokens with opposite jobs. RiskON is designed to deliver ~2X the asset's move, with no funding, no margin calls, no liquidations. RiskOFF gives up its upside beyond a certain point in exchange for a floor in the event of losses. Swap between them as the market turns: lean into the upside, step into calm, and learn the mechanics by trading them. Round 4 of our Trading Competition is now live. Finish in the Top 100 and earn protocol points.

BTC has spent the last 30 days or so inside one narrow band: $58,500 to $64,900. ETH has drifted the same way. Volatility on both is running far below its usual pace. Periods like this are where directional bets go to die quietly. Long gets chopped, short gets chopped, and waiting pays nothing. TradFi solved this decades ago. When markets go quiet, the desks do not go home—they trade the quiet itself. Covered calls, condors, carry: there is an instrument for every market environment, including the one where nothing happens. That is the range we are building into SMART Tokens: one split mechanism minting many shapes of risk—leverage, protection, yield strategies that earn while markets drift, and volatility itself, long or short—with no liquidations, no margin calls, no funding rate to manage. The future of DeFi is not more ways to bet on up. It is RiskFi—a way to earn in every environment.

Finance did not begin with chasing returns. It began with trading risk. London, 1686. Merchants in a coffee house paid underwriters to carry the risk of losing a ship at sea. One side paid to shed a risk it could not afford. The other side was paid to hold it. That trade predates the Bank of England by eight years. Chicago, 1848. Grain merchants opened an exchange, and within a few years a farmer could sell a crop he had not yet harvested and lock in the price today. Chicago again, 1973. An options exchange opened, and Black and Scholes published the formula to price optionality itself. Risk now had a market price of its own. The 1990s. Credit risk was cut away from the bonds that carried it and traded on its own. Three centuries, one direction: risk went from something you carried to something you could trade. As our manifesto puts it, every financial system that scales eventually becomes a system for managing risk rather than chasing returns. Crypto compressed most of that history into less than two decades. Spot markets, lending, derivatives—each rebuilt on-chain at astonishing speed. But the final layer, the one TradFi spent the longest building, got skipped. And even TradFi never finished the job: risk there is explicit in language but implicit in implementation, locked inside bilateral contracts, fund wrappers, and gatekeepers. Risk, in TradFi, is a prisoner of its own packaging. That is the opening. Crypto doesn't need to evolve over three centuries; we can go straight to the end state: risk as a first-class, on-chain primitive that can be isolated, priced, transferred, and composed. If tokenization freed the asset from its wrapper, RiskFi frees the risk from the asset. This is what we are building. Not the next product category. The next layer of finance. The RiskFi Manifesto:

Puts bleed premium. Perps bleed funding. There is a third door. Suppose there are two HODLers of BTC, but with opposite goals. One wants it with less risk, so they are willing to give up some upside in return for a floor on losses. The other wants amplified exposure and is willing to bear the extra risk to get it. Today, each pays a middleman. The first buys puts from an options desk and bleeds premium. The second opens a leveraged perp and feeds a funding meter that never stops, with a liquidation price waiting underneath. Notice that they both want exposure to BTC, but with different risk tolerances. So there is nothing to match, no trade to broker—the asset's own risk just needs to be divided. That is what a deposit in @TheRiskProtocol does. BTC or ETH goes in, and two SMART Tokens come out. RiskOFF is the calmer piece: for each 30-day epoch, it keeps a floor 5% below the starting price, and gives up the upside past a cap. RiskON is the leveraged part: the upside past that cap and the downside past that floor both land on it—which is exactly what makes it move at ~2X the underlying. Hold the piece that fits you. The cap is not picked; it is solved—set fresh each epoch at the level where the upside RiskOFF gives up is worth precisely what its floor costs. The pieces exactly offset, so no premium changes hands. In options language: a costless collar. And because both tokens are claims on the same pot, they always add back up to the underlying. The floor is not anyone's promise—it is enforced by how the deposit is divided. No funding meter, no margin calls, no liquidations, and no counterparty to trust. RiskON and RiskOFF are the first pair of SMART Tokens from The Risk Protocol. More are coming—each one a different way to hold exactly the risk you choose.

RiskON is built as a costless collar with RiskOFF. No premium changes hands; the two sides fund each other. RiskOFF gives up some of its upside in exchange for downside protection. RiskON takes that upside and, in return, offers downside protection to RiskOFF. That risk transfer is where RiskON's leverage comes from. Built this way, RiskON is meant to deliver ~2X the move of BTC or ETH. So I got curious: how would that actually compare to a 2X Perp? We knew RiskON would come out ahead, since it pays no funding. But the study revealed the gap is far bigger than we thought. A 2X Perp pays funding on the full notional, for as long as you hold it. Across the BTC and ETH bull markets in our study, funding on a long averaged roughly 17%–22% a year, and in the 2021 mania it briefly ran past 70% annualized. It climbs fastest exactly when you want the leverage most. We tested every bull market since 2020: 13 windows across BTC and ETH, long through each, cash in between. RiskON won all 13. $1 in a 2X Perp became $18 on BTC, $22 on ETH. $1 in RiskON became $53 on BTC and $97 on ETH. On ETH, the 2X Perp even finished below just holding ETH itself ($27)—all that leverage, all that funding paid, only to lose to buy-and-hold. Here is the full research:

RiskON is built to deliver ~2X on BTC or ETH. So we asked the obvious question: how does it compare against the way most traders actually hold 2X—a leveraged perpetual? As per our backtests, in every BTC and ETH bull market since 2020, 13 of 13, the perpetual finished behind RiskON. Here is the full study 🧵

Vitalik Buterin recently published this post: Building index-tracking assets on top of options instead of debt ( It lands almost exactly on what we are building. It helps to start with what @VitalikButerin is actually trying to achieve. His post is really about an old, hard problem: how do you build price-stable and index-tracking assets on-chain without trusting a central issuer? You want to give people exposure to something like the dollar using only a trustless asset, such as ETH, as backing. Every design like this has two sides. For everyone who is long, someone has to be short, and if the price moves far enough against the short side, the shorts go broke. The usual fix is liquidation: the system force-closes that position before the loss eats through the collateral. That fix is the real problem. To liquidate the moment a price is crossed, you need a feed that is fast and always right. As Vitalik puts it, "Real-time oracles are very hard to make safe." A fast feed cannot pause to be double-checked, and it is the easiest part of the system to attack. It is where a long line of collateral-backed stablecoins and lending markets have broken. His move is to change the building block from debt to options. He splits one unit of collateral into two complementary claims, which he calls P and N. P is the protected side, the one that behaves like the thing you actually want to hold, such as the dollar. N is the risk side that takes the other end of the trade. The two are built so that they always add back to the whole unit of collateral. Because P and N always sum to the collateral, no position can end up underwater, so there is nothing to force-close. Take away the liquidation, and the fragile real-time oracle goes with it. Settlement can be slow, and slow is safe. This incidentally is the same design architecture as our SMART Tokens. We do the same split. Deposit one asset, starting with BTC and ETH, and it becomes two tokens. RiskOFF is the calmer half: it gives up some upside in exchange for a floor under its losses. RiskON is the other half: it takes that downside and, in return, earns the extra upside. Put the two back together at any time, and you have your asset back. No loan, no margin, nothing to rescue. But here is where we differ. Vitalik favors a slow oracle, and to minimize increased exposure to the underlying as price ticks down, suggests that users independently rebalance prior to maturity. He acknowledges that such a design choice likely imposes potentially significant rebalancing costs that could potentially make the mechanism unworkable. We took a different tack. We favored abstracting the mechanics so that the product is something an ordinary user can simply hold and trade. We run it in repeating periods, called epochs, that reset on their own, with a safety barrier that ends a period early in a sharp crash, before the risk-taking side can fall below zero. That barrier is the trade-off: because it has to watch the price as it moves, our version leans on the oracle more than Vitalik's read-it-once design. We give up a little of that purity on purpose to achieve that abstraction. And in the process we solve for the rollover cost that Vitalik left unresolved. At the end of an epoch, users get automatically rolled over into a new epoch that resets the options at zero cost (the options are designed to be a costless collar). Our bounded, fully collateralized options turn risk into something you can safely hold, trade, and price. That is the piece DeFi has been missing. When the most credible architect in crypto independently arrives at the same thesis we have been working on, it tells us we are on the right track. RiskON and RiskOFF are only the beginning of what this design can do. We call the mechanism SMART, a Split Mechanism for Asset Risk-Tokenization, and the same split can produce a whole family of risk tokens: long-volatility and short-volatility tokens for trading swings directly, 10X Bull and -10X Bear tokens that give amplified exposure with no margin calls and no liquidations, yield-generating tokens, and tokens built for tail risk, across more assets over time. One mechanism, many shapes of risk…