Although there has been immense progress within Defi the last couple of years, options protocols have never really been able to catch any headwinds. If we’re looking at it from a degen/gambler perspective, perps are much more fun to use and can directly hedge against spot holdings. Conversely, if we look at it from a utility perspective, there does not seem to be much use-case for actually using options in defi just yet.
I am no expert in any type of trading or investing, and certainly not in options so I will not attempt to demonstrate any genius trading strategies or try to convince anyone that whatever written here is actionable in a live setting, but I do want to help people explore on the uses of options within defi itself using very simple logical scenarios. We will be using approximations throughout the article for ease.
If you are familiar with Impermanent Loss and how Liquidity Pools work, then feel free to skip to the section “2. Options as a Hedge”.
Denomination is an important aspect of trading and investing in crypto. Even if most people around the world denominate in USD, there are some that denominate in ETH. Furthermore, there is a time when one should denominate in USD and another when it is more efficient to denominate in ETH. I will spare the reasoning behind these as it depends heavily on unique investor profiles and their respective portfolio management, but this will be important to remember when we get to finding an actual use of options in Defi below.
If the collateral required to mint a stablecoin or deposit to receive yield in a farm behaves like an LP, in this example ETH/USD, then the following dynamics will occur:
If ETH price is decreasing, then the LP has more ETH and less USD because traders that are trading against the LP will be depositing ETH and receiving USD, thus making ETH price go down.
If ETH price is increasing, then the LP has less ETH and more USD because traders that are trading against the LP will be depositing USD and withdrawing ETH from it, thus making the price of ETH go up.
Thus, no matter the price action of ETH, the individual supplying the LP is always at the mercy of impermanent loss.
When ETH increases, they will have incurred an IL of ETH
When ETH decreases, they will have incurred an IL of USD
This type of collateral is somewhat hard to hedge against because we need to negate the IL of ETH when price increases, but conversely negate the IL of USD on the downside at the same time.
Disclaimer: An impermanent loss becomes absolutely permanent if either the token/price ratio never returns to the deposited ratio OR the tokens are withdrawn from the LP at a worse ratio.
One might that think perps could be a potential solution, but we soon realize that there is no 100% non-risk hedge with perps. Capital efficiency (amount deposited as collateral to open perp positions) and Risk (liquidation risk) are inversely correlated.
Supply less collateral = More risk + More capital efficient. The liquidation price of the perp position will be closer to spot price, thus more risky. If liquidated then the hedge is rendered useless. Another thing to note is that short side perps technically have infinite risk on the upside thus theoretically a short side perp can possibly require infinite collateral (theoretically under extreme circumstances).
Supply more collateral = Less risk + Less capital efficient. The liquidation price of the perp position will be further from spot price, thus less risky. The downside of this is that it is highly capital inefficient because it will take much more funds that are essentially stuck in this hedge.
We will use a very simplistic but practical example to understand these effects. For this experiment we will have the following parameters with all else equal (in a vacuum of some sorts):
ETH price is currently 1880 USD (as of 6 June 2023) .
You can deposit ETH/USD as collateral in FarmX to receive 30% APR paid in USD.
LP fees are 0 and the only LP that exists is the one used in each example.
Farmer A has 100 ETH and denominates in ETH (he is bullish on ETH long term and would rather accumulate ETH).
Farmer B has 188,000 USD and denominates in USD (he does not care for the price/thesis concerning ETH and would rather accumulate USD)
Both farmers have an equivalent investment as 100 ETH = 188,000 USD. They are both looking for a way to obtain the 30% yield by depositing in FarmX.
Farmer A will start by exchanging his ETH position into a ETH/USD LP position. This means his new assets from this point forward are now 50 ETH + 94,000 USD.
As random traders buy ETH, the amount of USD in the LP will be increasing and the amount of ETH will be decreasing. If the pool ends up with 100,000 USD then here are the dynamics that occur in the pool:
Thus when ETH price increases, Farmer A will have an LP internal position of smaller than 100 ETH since there is impermanent loss. The total position is now only ~94 ETH.
As random traders sell ETH, the amount of USD in the LP will be decreasing and the amount of ETH will be increasing. If the pool ends up with 88,000 USD then here are the dynamics that occur in the pool:
Thus when ETH price decreases, Farmer A will have an internal LP position of greater than 100 ETH since there is impermanent gain. The total position is now ~106.8180 ETH.
So to conclude:
When ETH price increases in USD, Farmer A has a loss in his denomination.
When ETH price decreases in USD, Farmer A has a gain in his denomination.
Farmer B will start by exchanging his USD position into a ETH/USD LP position. This means his new assets from this point forward are now 50 ETH + 94,000 USD.
As random traders buy ETH, the amount of USD in the LP will be increasing and the amount of ETH will be decreasing. If the pool ends up with 100,000 USD then here are the dynamics that occur in the pool:
Thus when ETH price increases, Farmer B will have an internal LP position of greater than 180,000 USD since there is a gain. The total position is now ~200,000 USD.
As random traders sell ETH, the amount of USD in the LP will be decreasing and the amount of ETH will be increasing. If the pool ends up with only 88,000 USD then here are the dynamics that occur in the pool:
Thus when ETH price decreases, Farmer B will have an internal LP position of smaller than 180,000 USD since there is a loss. The total position is now ~176,000 USD.
So to conclude:
When ETH price increases in USD, Farmer B has a gain in his denomination.
When ETH price decreases in USD, Farmer B has a loss in his denomination.
So given these circumstances and dynamics, how would Farmer A hedge his ETH denomination downside risk? When ETH increases in price, he actually ends up having less ETH as collateral. Common logic tells us that he could long ETH perps but we get the following dilemmas:
How many ETH does he long? It is not a linear hedge at all.
How much collateral is he willing to post on a perp exchange to hedge this way? This means his farming will actually need much more funds tied up than he initially thought.
What if ETH drops in price? He will have liquidation risk as well as an even bigger hole to fill. This is starting to not look like any sort of hedge at all…
Conversely, how would Farmer B hedge his USD denomination downside risk? When ETH decreases in price he ends up having less USD as collateral. Common logic tells us that he could short ETH perps but we get the following dilemmas:
How many ETH does he short? Again this is not a linear hedge.
How much collateral is he willing to post on a perp exchange to hedge this way? This strategy will finally tie up much more funds than he initially was willing to use to execute.
What if ETH increases in price? He will have liquidation risk as well as an even bigger hole to fill. This is not even taking into account the unlimited risk of ETH going up in price during a zhupercycle.
As we take all these issues together, we realize that we need to find a mechanism that gives Farmer A a gain when the price of ETH increases and limited/fixed losses when the price decreases. And reciprocally, this same mechanism has to give Farmer B a gain when ETH price decreases with limited/fixed losses when the price increases.
Using options would theoretically serve this purpose well:
Call options will give Farmer A a gain when the price of ETH increase and also limits his losses when the price decreases (option expires worthless and he will only lose the initial sunk-cost paid for the contracts).
Put options will give Farmer B a gain when the price of ETH decreases and also limits his losses when price decreases (puts expire worthless and he will only lose the initial sunk-cost paid for the contracts).
Thus, it is possible to use options to hedge against IL from LP positions (whether collateralized or not). Both farmers will have hedges that do not have any liquidation risk and are as close to linear as possible (discounting volatility). They will only have sunk costs which are known and paid for.
We will once again run a super simplified example built on top of the previous experiment with the same Farmers A and B under constant parameters with all else equal in a sort of vacuum to see if this is cost effective theoretically with the following parameters added:
Volatility is constant. We will use a snapshot of today’s options pricing (6th June 2023) to see if this is even worth pursuing.
There is no massive slippage for options pricing, so the bid/ask have an infinite bid on the listed prices.
Disclaimer: Options can vary in price greatly depending with volatility. As more liquidity and popularity grows for options, their pricing will be tighter and much more smoothed out. For the sake of keeping it uncomplicated I will be approximating a couple of things and simplifying others so it might grind their gears of professional options traders and I am terribly sorry for that.
Given these added parameters, Farmer A wants to hedge against a decreasing ETH denominated position as the price increases.
In case of a 30% increase in ETH price (from 1880 USD to 2444 USD), we can calculate his new LP asset ratios as well as his new internal ETH total with the following:
At 2444 USD per ETH, the IL of 12.2942 ETH converted into USD would be 30,047.02 USD so this is the amount that our hedge should be able to cover. If it fails to cover this amount upon a price increase of 30% for ETH then we are not able to find a delta-neutral strategy with options.
A straightforward method would be using ITM Call options (in the money: contracts with a strike price that is below the current price) in order hedge against Farmer A’s downside denominated in ETH, which I shall demonstrate below.
Disclaimer: Since Farmer A denominates in ETH, all his pricing will be denominated in ETH and not USD.
For this example, we will look at the Deribit orderbook in the image right above and we clearly see that a ETH 1400 call option expiring on 29 September 2023 has a cost basis of 575.44 USD = ~0.3060 ETH (with ETH price at 1880 USD).
If we purchase these contracts then we have the right to buy ETH at the price of 1400 for the whole duration of the option, which is 115 days from 6 June 2023 (today) to 29 September 2023.
If ETH ever reaches 2444 USD within this timeframe, then we have an implied profit of 1044 USD = ~0.4271 ETH per contract (with ETH price of now 2444 USD). We will also have an impermanent loss of 12.2942 ETH or 30,070.02 USD (which is actually permanent if we decide to withdraw it).
So in order to completely negate the 12.2942 ETH loss incurred if ETH does reach 2444 USD and with a profit of 0.4271 ETH per contract, Farmer A will have to purchase 12.2942 ETH/0.4271 ETH = ~28.78 contracts yielding 0.4271 ETH each. This is a total of 8.8066 ETH at current price (0.3060 ETH per contract cost basis).
This means that in our example: In order to hedge 100 ETH worth of total position in a ETH/USD LP in case of a 30% increase in ETH price, it is around 8.8066 ETH per 115 days in initial sunk costs. Now let’s model different scenarios where it might be possible we get ETH back from the exercising of the options if they are still in the money:
If ETH price at expiry (29 September 2023) ends up being 30% higher as predicted at 2444 USD:
We have a staggering cost that can only be offset by a farm giving us ~28% APR in yield for the whole duration of the hedge.
If ETH price at expiry ends up being the same price as when we bought the options at 1880 USD:
Can be offset by a farm giving ~4.63% APR in yield.
If ETH price at expiry is 20% lower at 1504 USD:
We will still end up with a net positive of 4.9853 ETH.
If ETH price at expiry is 50% lower at 940 USD:
We still end up with a net positive of 35.7232 ETH.
Thus, we can conclude that if we are hedging to protect our ETH denomination from a 30% increase in ETH price, Farmer A would need to receive ~28% in ETH to entirely offset IL/liquidation risk. If ETH price ends up lower than 1880, then the cost of the hedge will be eating into ETH profits from impermanent gain. As you can see, we will need to be prepared to pay a staggering price, or find very profitable yield, or else we will be losing ETH. This high price can also be attributed to the high IV (volatility, indicated to the left of the prices in the Deribit snapshot) of the call options, so it would be much cheaper if the IV was lower. Conversely, if markets are in fact efficient then a lower IV would mean less reason to hedge…
Logically speaking, if ETH is going up in USD price, then it is a very strong asset, so if someone wanted to find a way to supply his ETH into an LP to earn yield while also keeping his full exposure to ETH, well… you know what they say about having your cake and also eating it… One could even argue that it is ill advised to deposit your ETH into an ETH/USD LP under any circumstance if you are expecting a bull run…
Using the previously mentioned 30% APR yield attainable by depositing in FarmX and rolling over all USD profits into ETH, Farmer A would be able to receive a delta-neutral yield denominated in ETH, just barely.
Taking a look on the other side, Farmer B wants to hedge against a decreasing USD denominated position as the price decreases.
In case of a 30% decrease in ETH price (from 1880 USD to 1316 USD), we can calculate his new LP asset ratios as well as his new internal ETH total with the following:
At 1316 USD per ETH, the IL in USD would be 22,708 USD so this is the amount that our hedge should be able to cover. If it fails to cover this amount upon a price decrease of 30% for ETH then we are not able to find a delta-neutral strategy with options.
We will mirror a similar method Farmer A used and use ATM Put options in order hedge against Farmer B’s downside denominated in USD, which I shall demonstrate below (at the money = strike price is equal to current price).
As for Farmer B, since 1880 is awful close to 1900, we will be looking at the ETH 1900 put options expiring on 29 September 2023 in the very same image attached above with a cost basis of 163.23 USD for each contract.
If we purchase these contracts, then we have the right to sell ETH at the price of 1900 USD for the whole duration of the contract, which is the same 115 days.
If ETH ever reaches 1316 USD, then we have an implied profit of 584 USD for each contract. We will also have an impermanent loss of 22,708 USD on our position that needs to be mitigated.
To mitigate the 22,708 USD in losses if ETH does reach 1316 USD, Farmer B will have to purchase ~38.8835 contracts yielding 584 USD each. This is a total of 6348.01 USD at current price.
So in order to hedge 180,000 USD worth of USD in a ETH/USD LP position, it is around 6348.38 USD in initial sunk costs. Now let’s model different scenarios and see how this hedge performs in a couple different scenarios:
If ETH price at expiry ends up being 30% lower as anticipated at 1316 USD:
We have a cost of hedging that needs to be offset by a yield of 11.19% APR.
If ETH price at expiry ends up being the same price as when we bought the options at 1880 USD:
Can be offset by a farm giving ~9.82% APR in yield.
If ETH price at expiry is 20% higher at 2256 USD:
We will still end up with a net positive of 19,595.06 USD
It seems that mitigating the downside in USD denomination is much less costly than hedging while denominated in ETH. Farmer B would need to receive >11.19% in USD to entirely offset IL/liquidation risk. If ETH price ends up higher than 1880, then the cost of the hedge will be eating into USD profits from impermanent gain. As long as we are able to find a yield greater than 11.19% APR, then these costs are entirely offset whether ETH goes up or down.
Using the previously mentioned 30% APR yield attainable by depositing in FarmX, Farmer B would be able to receive a very generous delta-neutral yield denominated in USD.
Additionally, we can calculate the cost in % terms for Farmer B if he were to hedge for only 20% and 10% decreases in ETH price with the same Deribit snapshot image + same calculations and we get:
To hedge against a 20% price decrease in ETH, it will cost ~8.52% per year.
To hedge against a 10% price decrease in ETH, it will cost ~2.3% per year.
It is much less costly to hedge against a less volatile ETH, but you are also at risk of suffering more IL if the downside volatility is higher than what you have hedged for.
I hope that this experiment will not only help people with their various models and frameworks, but also allow them to do a further deep dive into options and the various possible use cases they might have.
Defi is still young and evolving at an extremely rapid pace. As crvUSD scales up, it is very likely we will see much more composability for LP positions or tokens, which will make things much more interesting. One might even start thinking about what Uniswap has in store for us with their NFT marketplace, which should fit very well with their LP positions stored in NFTs.
Couple this with the fact that many options protocols are experimenting with adding options for tokens other than BTC and ETH (Lyra is a notable example with a terrific AMM design and Aevo is an order-book based AMM similar to Deribit but on-chain) which should open even more doors for options to be used to hedge against LP positions of 2 volatile tokens (think ETH/altcoin LPs).
Great things are coming up ahead in Defi and I will try to make time to use my blog to share various thoughts and strategies that could be of use to anyone that finds them interesting.
~ 0xShrekt in gwei @0xshrekt