Potion — Kelly Criterion Bonding Curve for Risk Instruments

Potion Labs (now Ntropika Labs)
5 min readApr 8, 2022


Published in defiprime.com

Nick Sawinyh on 25 Jan 2022

Hello! What’s your background, and what are you working on?

Potion is a web3 app that helps create automatic insurance markets where buyers and sellers come together and trade at mathematically sustainable prices.

For example, you can buy Potion insurance on the price of Ethereum, locking a $3000 price. If the price of Eth went below $3000, for example, to $2500, the insurance would kick in with a $500 payout.

You can also take the other side in Potion, and be the one selling risk insurance (be an LP). As an LP with Potion you can take advantage of the Kelly Criterion to trade at optimal combinations of premium and utilization at all times, leading to long-term sustainable markets.

What’s Potion backstory?

We wanted to build an automated tail risk hedger. A smart contract that would constantly offset a certain portion of the left tail risk of your position. We realized there wasn’t any sustainable market maker that could act as a counter-party for such a robo-hedger, so that became our v1 plan: building a sustainable automatic market maker for risk insurance. We’d love for V2 to be the robo-hedger buying from it, but the community will be the one deciding that.

Beyond Potion and this first particular insurance application, more generally, we believe it is worthwhile spending energy imagining and building trustless financial applications. For the people, by the people. We have a humble aspiration to contribute towards the betterment of society in our own small way.

It feels like an incredible privilege to be working in this industry at this time.

What went into building the Potion?

Building Potion was an incredibly tough effort. First, we had to reach submit for Hackmoney, which was a small miracle in itself. Then we had to create a coherent vision to raise capital and retain key members after the hackathon. We were incredibly fortunate with the seed round we were able to close. With the backing of investors, we then had to attract new talent and build an internal culture and project management discipline to ship robust and high-quality software while being 100% fully remote.

Our biggest challenge was first to develop an alternative to Black Scholes. We wanted a model we could genuinely believe would help LPs have an expectation of capital growth. The months of research were some of the most exciting and nerve-wracking of our lives. We were terrified we wouldn’t find an optimal bonding curve. But, in the end, when we did and saw the first Kelly optimal curves finally emerge, it felt like a dream come true.

Besides the tech itself, our biggest challenge by far was legal. It took a lot of effort to find a way to release the results of our research without engaging in unlicensed regulated activity across many jurisdictions. This almost killed the project and the team. We enduredand ultimately we believe we found a way out of the problem. This was our second breakthrough after coming up with the Kelly Bonding Curves. We hope it can be a step towards safer regulatory positions for other dev teams in DeFi too. It’s in everyone’s interest to see web 3 code flourish in a legally sustainable way.

What’s your business model?

We think of ourselves as digital artisans building public goods for the trustless metaverse. We have spent nearly two years researching and building, and now want to share our work with the world for anyone to build on and extend it.

Our business model is a combination of many experimental models being trialed in web3. We plan on generating revenue as a team by selling NFTs that commemorate the launches of our research.Importantly, NFTs sales can be run in a fully compliant way, allowing us to generate revenues that can help sustain our operational costs, pay back our debts, and keep us doing more exciting research in DeFi.

We don’t see other derivatives platforms as competitors but as co-creators. We are building towards the same common good: cool, reliable risk management markets in DeFi. In some sense, it doesn’t matter who gets there ultimately. The point is the same: for society to have a robust layer for trustless risk management.

Our differentiation angle with respect to the other tech out there is that we are maniacally focused on capital sustainability for the LPs. This is why we invested so heavily in implementing the Kelly Criterion on-chain, which we believe nobody else is doing at the moment.

What’s your position on the regulatory landscape today?

Derivatives trading is a highly regulated environment. We have consulted with two different European regulators, and both confirmed that the sale of NFTs is not to be considered a securities issuance. They also confirmed the team will not retain liability for open-sourcing code, provided we won’t engage in active engagement parametrization or de-facto management of the platform after release.

In this sense, we believe that we are working in complete compliance with existing regulations.

What are your goals for the future?

Our first goal is to execute a flawless NFT minting experience and to have fun hosting the Potion Unlock game. It is a novel release strategy and we are keen to make it the best experience possible for everyone who supports the project.

On the product side, we are excited about expirationless pay-as-you-go insurance and automated curve management, which are two big rocks for which we have exciting ideas for, and that are a natural next step for this first release. Long term, we dream about an automated tail risk hedger for asset portfolios, where one can store long term their hodlings in a black-swan protected manner.

What are your future thoughts for the DeFi market?

We believe Crypto and DeFi are a discontinuity moment in the history of civilization and that they unlock entirely new possibilities for us all.

Trustless coordination between actors of society can unlock huge value. Many important services have already been built in governance and financial services, but it feels like they are still just the beginning. The long term impact potential for reliable public good services remains enormous.

Where can we go to learn more?