This piece is only for educational purposes and does not constitute investment advice. Disclosure: I’m working with the Opyn team as an ambassador to help spread the word about their project. I’d also like to thank Shiliang Tang for providing some thoughts on how to think about delta-hedging on-chain options.
While we’ve seen incredible growth in centralized option markets over the past several months, developers across the world are working incredibly hard to bring options to DeFi. Opyn is one of these teams which recently launched their V2. In this update, I was happy to see users can now place option spread trades with improved capital efficiency. Furthermore, the market-making quoting system through the 0x API allows for competitive option prices relative to centralized venues such as Deribit.
Many DeFi users utilize options to make directional bets on the underlying price of an asset. For example, if we think ETH will go up in value we could purchase a call option. Conversely, if we wish to hedge our ETH holdings from a crash in the price we could purchase puts. However, things get interesting and slightly more complicated when we move beyond simply trading the directional price moves in ETH.
Unlike futures, with options, we can make bets on the underlying volatility of an asset. This style of trading is commonly referred to as “volatility trading” which is an advanced strategy commonly used by hedge-funds, market-makers, and sophisticated retail traders. As the name would suggest, our goal with this strategy is to focus on trading the volatility of the asset rather than the price direction.
Below are some key terms we need to understand as volatility traders:
- Realized Volatility (RV): This is the actual historical volatility of the underlying asset throughout the life of the option. For example, if we’re analyzing a 30-day option for ETH, we’d want to look at ETH’s volatility over a 30 day period. There is no definitive way to measure RV as this value must be estimated. Many people use the standard deviation of log returns to quickly approximate the historical volatility. We also see market participants use fancier models such as Yang-Zhang or Garman-Klass to measure volatility. Regardless of whichever approach we use, RV is just an approximation and there is no single correct value.
- Implied Volatility (IV): This is the market’s future estimate of the underlying’s actual volatility when the option matures. For example, if we have a 30-day option with an IV of 90%, then the market is implying the underlying asset will have an RV of 90% across 30 days. Options are priced in terms of IV which reflects the market’s consensus for future volatility. Our job as volatility traders is to assess whether IV is fairly priced relative to where we think actual RV will settle at expiry. If we believe there’s an upcoming market event that could drastically increase RV (ie: regulatory action) much higher than the current IV we could “go long volatility”. In this case, we’re betting on RV being greater than the current IV at expiry.
This value represents how much the option value will change for a small change in the underlying asset. Calls have positive delta and puts have a negative delta (ie: calls increase in value as the price rises whereas puts decrease in value as the price rises). For example, if a call has a delta of +0.40, this means for a $1 increase in the underlying, the value of the call should increase by around $0.40. Conversely, a $1 decrease in the underlying will result in a -$0.40 change in the option value. If we want to trade an option without worrying about the direction of the underlying price, we need to eliminate our delta exposure so we’re only exposed to the volatility of the option.
Let’s review an example using Opyn’s February 29, 2021, ETH option chain. For this example, we’ll look at the $1,280 strike call option with 39 days left to expiry. At the time of writing, we have the following values:
A: Spot Price of ETH: $1,239.16
Call Option IV: 170.53%
Call Option Delta: +0.586
The delta of this call option is +0.586 which means for a $1 increase in ETH we can expect a $0.586 increase in the option price — in other words, we are long 0.586 units of ETH. Furthermore, this call option is currently priced with an IV of 170.53% which implies the market is expecting RV to be 170.53% in about 40 days when the option matures.
By using a volatility cone we can visualize the historical volatilities across different time regions and compare them to the option’s current IV. Given our option has 39 days left to maturity, we’ll focus on the historical volatility near the 30-day region.
According to Genesis Volatility’s analytics data, ETH 30-day RV is 158.17%. We can see the current 30-day RV is close to its all-time highs which may suggest that volatility is richly priced at these levels (note: the volatility cone is based on historical data and assumes the past changes in volatility will be similar to the future — clearly this doesn’t have to be true going forward). In this case, we can see the spread between the current 30-day RV and option IV is around 12 volatility points. Historically we can see throughout time that 30-day volatility had a maximum RV of around 160%. The 39-day option we are trading should be lower than 160% as can be seen by the downward slope of the volatility cone (if we linearly interpolate the 39 day RV we’ll find that the volatility spread is even higher).
If we believe that ETH will stay relatively quiet for the next 39 days with minimal price movements, it’s probably worthwhile selling the IV for this call option. In other words, we should sell this option if we believe ETH’s RV will be less than 170.53% over the next 39 days. Conversely, if we believe there will be a major upcoming event that will cause RV to spike well past 170.53% over the next 39 days, then we’d be better off buying the option.
Given the points laid out above, I’d argue we can maximize our chances of making money by shorting this call option. When we sell an Opyn call we need to lock up 1 ETH for every option as collateral which may not be the best use of capital. As a result, we can resort to DeFi protocols such as Aave or Compound to use borrowed ETH and post it as collateral for selling these options. This can partially offset capital inefficiencies when trading on DeFi protocols while magnifying our returns.
Given that a call option has a positive delta if we short a call we are short delta. In this case, the call has +0.586 delta exposure and by selling it we now have -0.586 delta exposure. To purely trade the volatility we need to eliminate our delta exposure, therefore, if we’re exposed to -0.586 units of ETH on the option then we need to go long +0.586 units of ETH to neutralize our delta. This can be done by simply purchasing 0.586 ETH from Uniswap or perpetuals on dYdX. Regardless of the method, the goal is to get exposure to +0.586 ETH and reduce our short call option delta to zero.
These are the following transactions at the beginning of the trade:
SHORT: Selling $1280 FEB-29–2021 Call Option: -0.586 ETH Delta
LONG: Buying ETH from Uniswap or dYdX: +0.586 ETH Delta
— — — — — — — — — — — — — — — — — — — — — — — —
Net Portfolio Delta Exposure At Inception = 0 ETH
By placing these trades our portfolio is delta-neutral. However, the delta of an option is not constant and changes over time based on a variety of different factors. This requires us to constantly rebalance our net delta (dynamically hedge) so we are not exposing ourselves too much to moves in the underlying price. Suppose there was a large move in the market causing the call option delta to increase by +0.20. By not hedging our delta exposure we have ended up inherently shorting 0.2 units of ETH as shown below:
SHORT: Selling $1280 FEB-29–2021 Call Option: -0.786 ETH Delta
LONG: Original Delta-Hedge of ETH from Uniswap or dYdX: +0.586 ETH Delta
— — — — — — — — — — — — — — — — — — — — — — — —
Net Portfolio Delta Exposure = -0.20 ETH
It may be tempting to hedge every small change in price so we’re never exposed to the delta. Theoretically, this should work, however, in the real world traders must be mindful of transaction costs and slippage associated with the delta-hedging process. Given the abnormally high gas fees we’ve seen over the past few weeks, it’s unlikely a DeFi trader can rebalance their delta exposures frequently and still make money by capturing the volatility spread. Nevertheless, for now, we can stick with a simple approach of hedging every time the net delta exceeds a threshold of +/- 0.10 ETH. It’s important to stress that delta-hedging is very context-specific and there are a variety of different ways to hedge which depend highly on individual risk tolerance. For those interested in learning about different delta-hedging techniques I’d suggest exploring Euan Sinclair’s book, “Volatility Trading”.
At the end of this trade, if RV is less than the option’s IV of 170.53% then we should end up making money provided we hedged our deltas appropriately. Lastly, after the option’s expiry, we also need to close our hedge position in ETH so we’re not exposed to any price changes.
This was a simple overview of how we can think about trading volatility on-chain. DeFi investors can also explore strategies such as covered calls which can allow them to capture expensive upside volatility. Overall, I am optimistic that new scaling solutions which offer lower gas costs will one day make decentralized volatility trading a vibrant market for both retail and institutional participants!