Bitcoin's volatility is one of the obstacles to become mainstream as a currency. Above all the fear from crashes is hindering adoption by merchants and consumers. Here I'll describe an idea I got to be able to hedge against crashes using Atomic Swaps. This may look strange, but it could make sense if we consider that atomic swaps are, in reality, a kind of option contract. See more below ...

Hedging with options and the collar strategy

Options contracts are a common way to hedge against volatility. We can buy a "put option" if we own a good or asset - e.g. Bitcoins or stocks - and want to avoid being surprised by a crash. The seller of a put option gives the buyer the right to sell a good to a certain price, the so-called strike price. The seller charges a premium for that, because he needs to deal with the risk of the contract. (We will use American-style options here).

Example: We buy a put option for 1 BTC with a strike price of 9500 USD, with an expiration date of 30 days. This means that we can sell 1 BTC for 9500 USD in the next 30 days, even if it at the spot market the BTC falls to zero.

This way, we are safe from any crash that may happen. But: The premium for a put option for a value of 1 BTC, with strike prices near the spot price, can cost hundreds of dollars (Put options are cheaper if the strike price is lower.). See current prices e.g. here (Deribit). How do we get that money back?

There is a second step we can do at the same time: selling a call option. This means that we sell another person the right to buy BTC to a certain value. Who would buy that? Basically, traders who are speculating with a price increase but do not want to take the risk to buy a full BTC.

Again an example: We sell a call option for 9500 USD with expiration date of 30 days. We give our counterparty the right to buy 1 BTC for 9500 USD, even if it goes to 1 million, and charge a premium for it because we are taking the risk in the contract.

So we can finance a hedge against a crash, if we at the same time limit our potential to profit from a price increase. This strategy is called "collar".

Obviously, if you want to speculate on a Bitcoin price increase, then a collar is not for you. But there are definitively use cases - above all in the business world. For example, merchants who accept Bitcoin without a centralized payment gateway (like Bitpay or Coingate) need a strategy to manage volatility risk, and a collar is one of the possible options.

Atomic swaps as American-style options

Now most people would trade options on centralized exchange platforms. But there is an interesting alternative which is (almost) decentralized: Atomic swaps (See here how they work.)

Why? Aren't atomic swaps meant to exchange currency and not to speculate and hedge? Yes. But they have one characteristic which some find "unfair": one of the two parties can decide if it does not accept the trade. This party can be compared to the buyer of an (American-style) option. (See more on this interesting paper)

Normally, atomic swaps have short expiration times (like some hours, or a day). But we can perfectly imagine atomic swaps where the "buyer" can wait 30 days to decide if he concludes the trade or not. This would give the "buyer" the right to buy the BTC or the other coin for a fixed price until the expiration.

Basically, the idea is now to use this strategy to implement the "collar" described above in a descentralized way, without a centralized exchange platform. We can use a collar against a stablecoin to mimic the USD or EUR price.

Advantages and disadvantages

Advantages:

• We don't need any centralized options platform. Nobody can run away with our money.
• We always own the private keys of the Bitcoins we're using.
• No KYC needed (see here why KYC can be a risk), privacy is on the same level than with normal Bitcoin usage.

Disadvantages:

• We can't use fiat currencies directly to hedge. We must use blockchain-based stablecoins, which have their own set of problems.
• Traditional atomic swaps need on-chain transactions which normally cost transaction fees. This makes this method unsuitable for small Bitcoin amounts.
• At this moment, there is low liquidity on atomic swap markets.
• A maybe a bit more technical problem is that we cannot easily use the same BTC we're using for the "call" option swap than the one we use for the "put" option swap. Maybe there is a solution for that though.

Discussion

What do you think? Is this idea factible or totally crazy? Is there anyone who is already using this kind of  hedging? Are there platforms supporting it? Post your opinions!

(This is part of my current research about a possible "tool set" for the unbanked and other people wanting to live without a bank account (see also this poll/thread), and would allow people who couldn't afford to be caught by a crash live with a Bitcoin wallet.)

Exactly, they can be coded in pure Bitcoin Script, although with Ethereum-style contracts you have some more possibilities regarding the management of the "premium" the option buyer pays to the option seller. With Bitcoin Script instead the premium payment would have to be done afaik with an additional transaction. In reality, like the authors of the linked paper wrote, it's exploiting a "flaw" in Tier Nolan's Atomic Swaps to use them for something which make sense.

However, there are some stablecoins like Dai that probably can't be used in this way because they can be liquidated - at least, this would add an additional risk to the option buyer because instead of the amount in Dai he could get an Eth amount.


Thanks, will take a look at it.

Yep, what would be important is obviously to design tools which would make the process straightforward. Perhaps I can propose the feature to the AtomicDex people ...

You have to differentiate two things: the price of the option (which is what you pay to the seller, or you get from the buyer) and the strike price of the option.

The strike price is the price where you can buy or sell your BTC according to the option contract.

For example, at this moment we have a spot price of approximately $9100. At Deribit (see here, take into account that this is a centralized platform), we see the following interesting option prices for a very tight collar:

Call option with expiration of June 5 at strike price 9250: ~$480-510
Put option with experiation on June 5 and strike price 9000: ~$490-520

So you can approximately finance a collar where you are guaranteed a price of $9000, selling the $9250 call option.

Now this means obviously that you will lose the opportunity to profit from all price increases of more than 9250 $. You can opt instead to take a little bit more risk with a "wider" collar: e.g. you could finance a $8000 put option with a $10000 call option. So you can make a profit of up to ~10%, but instead you risk to lose up to ~13.75%.

A corollary of all that is that put option prices are higher when the market is bearish. That means: If you are in a bear market and want to protect yourself from a deeper crash, you will lose almost all potential profits of a "backswing". However, if the market is bullish, put options are cheaper. That means that in a bull market you can often sell a call option with a higher strike price to finance you put option "just below the spot price". Currently I estimate the market to be slightly bearish.

To answer your question if you can re-setup the collar: You can of course re-buy your call option if you see the price going up and think that it will go up above the strike price of your call option, but that would mean that you lose the money you got for it, if you don't sell (at the same time) the put option, and then you won't protected anymore for a crash. Also, the more the price goes up, the call option will become more expensive. So if you wait too long, you would have to pay more than you got for it.

It's all a set of tradeoffs, and thus, it's something mainly suitable for business owners where stability is more important than a possible profit from a Bitcoin price increase. Another use case I have in mind is a worker who gets his salary in BTC at the start of the month and wants to protect a part of it until the end of the month so he can pay for his living costs, without having to change it directly to fiat.

This would be optimal, but it could cost you a lot of transaction fees if you re-calibrate the collar too often. But yes, it would be very attractive to have the opportunity to set up a longer term strategy and automate the individual atomic swaps.

You're totally right - I think my fault was to write "hedge against volatilty" if what I should have done was "hedge against a crash", which is exactly what "buying" a collar does - it protects you from volatility to the downside, while the collar seller will profit from it. I'll change the title accordingly - it also may be catchier then

I'm not an options expert, I discovered them a short time ago, when I was researching for hedging strategies and instruments for that. So I think you can give valious input to this thread.

---

I have looked at Tier Nolan's Atomic Swap model again, above all to try to answer the question "where to put the premium payment?". The authors of the paper I linked above argue that to do that in the correct way with Bitcoin a new opcode would be needed. But looking at the model, this seems not necessary if you pay the premium:

1) in stablecoins if you are buying a "put option atomic swap" (=allowing you to buy stablecoins for the strike price),
2) in Bitcoin if you are buying a "call option atomic swap" (=allowing you to buy Bitcoins for the strike price).

Explanation below.

From the perspective of the "collar buyer", i.e. the person who wants to protect from crashes buying a put option while selling a call option, this means: they have to pay stablecoins and will get Bitcoin for it. This is of course not ideal because an additional exchange operation would take place.

---

Explanation: In Tier Nolan's Atomic Swap model you have four transactions, two on each blockchain. A is the option buyer - the person who "set up" the contract and can chose to execute it or not, B the option seller. So we need a mechanism to pay from A to B.

Simplified description of the transaction in Tier Nolan's model:

TX1 on Blockchain X: Transaction paying the value in Coin X of the exchange from party A to B
TX2 on Blockchain X: Refund transaction with timelock if the exchange doesn't take place.
TX3 on Blockchain Y: Transaction paying the value in Coin Y of the exchange from party B to A
TX4 on Blockchain Y: Refund transaction with timelock (shorter than the timelock of TX2).

From my interpretation, the premium payment can only be inserted into TX3. TX3 is signed by party B (option seller). If it included an input from A going to B for the premium payment, the deal would be approximately: "B only gives the option to get Coins B if A pays the premium."