Crypto Derivatives, Lending, and a touch of Stablecoin

Let’s discuss the potential for on-chain crypto derivatives. In a nutshell, an on-chain derivative is a derivative contract whose payouts occur completely on-chain through a smart contract. They likely will rely on off-chain components — Oracles to deliver information about off-chain events or data — but the rest of the infrastructure need not reach off-chain.

If you’re not familiar with derivatives, they’re simply contracts that derive their payout from something else (an underlying), via some settlement mechanism. The simplest is physical delivery, i.e. at the end of some period, something is delivered from seller to buyer (a barrel of oil, a Bitcoin, etc). As a result, you’d expect the price of the derivative to roughly track the underlying asset. Alternatively, you can just settle to cash, which means paying the differences in price of the underlying asset from the price the contract is entered into to the final settlement price(or whatever the pricing formula is relative to some price or event).

“A derivative is a financial security with a value that is reliant upon or derived from an underlying asset or group of assets. The derivative itself is a contract between two or more parties based upon the asset or assets. Its price is determined by fluctuations in the underlying asset.” (https://www.investopedia.com/terms/d/derivative.asp)

Derivatives come in many forms and with many neat names: futures, forwards, NDFs, swaps, options (see: Deribit for off-chain crypto options). When you participate in a betting market, or trade a CFD (Contract for Difference) you are trading a derivative. When you participate in a prediction market, you are trading a derivative!

On-chain derivatives aren’t a new idea. Prediction markets, which were some of the most hyped early proposed uses of Ethereum, are old hat: Augur (which just went live, woo!), Gnosis, and Stox are the largest projects. These are general purpose derivatives markets — you can create a prediction market for pretty much anything. A more canonical form of standardized derivative is a Futures or Forward contract, which lets you bet on the future price of some asset. This can be as simple as the price of a cryptocurrency itself.

BitMEX is one of the most popular [centralized off-chain] crypto trading venues and it’s a derivatives exchange. You aren’t actually trading BTC — it’s not like Coinbase where you can buy 1 BTC and transfer it out to your wallet. Instead, you are making bets on the future price of BTC (denominated in US dollars). In addition, we also have the major regulated off-chain derivatives markets that list BTC futures contracts: CME and CBOE.

Off-chain hedges

If you’re already comfortable with how derivatives actually work, feel free to skip this section

Consider the situation where you own 1 BTC and want to hedge that exposure — you want to short 1 BTC worth of futures, thereby neutralizing your portfolio to the price fluctuations of the underlying BTC. If constructed correctly and the market is functional, this will allow you to hold your BTC physical (aka spot) position without having any exposure to the price volatility.

Why use derivatives when you can just trade the underlying?

Derivatives can allow you to: avoid taxable events, keep more assets in cold storage, lock in prices for future sales by selling forwards (e.g. if you’re a miner), diversify indirectly (e.g. buy an index fund swap), and speculate (e.g. shorting and leverage)

Source: https://www.kotaksecurities.com/ksweb/Research/Investment-Knowledge-Bank/what-is-derivative-trading

To put on a derivatives exposure, you typically use leverage. To be exposed to short 1 BTC, for example, you may only need 0.3 BTC in collateral (giving you 3x leverage, approximately). So you have your long 1 BTC physical exposure, ideally in cold storage, and then you move 0.3 BTC into a margin account at BitMEX to short 1 BTC future and hedge your position.

Consider what happens when BTC/USD goes up by 10%. Your long 1 BTC has appreciated in value by 10% (0.1 BTC worth, at the old prices). Your short 1 BTC future has lost 0.1 BTC worth (again, at the old prices — at new prices it would be closer to 0.091). As a result, this amount is moved from your margin account into someone else’s (kind of — depends how the margining system is set up) and you are left with about 0.2091 BTC in margin collateral at the futures exchange. But wait, we were trying to neutralize our exposure but it seems like we’ve lost, going from 1 BTC to 0.9091 BTC (0.7 left in our wallet, and 0.2091 in the margin account)? Ah, but we’ve neutralized it in terms of something else, probably the US dollar (depending on how you setup the derivative — the BitMEX BTC futures are BTC/USD contracts). So if BTC was worth $1000 and went up to $1100, our total portfolio value has gone from $1000 * 1 to $1100 * 0.9091 — no change!

But consider what happens if the price keeps going up. Eventually, our margin balance will approach zero value and we will be automatically closed out of our hedge if we don’t add more collateral. To prevent this, we’d have to continuously move our (physical, underlying) BTC, probably from cold storage, into our margin account at whatever exchange we have our futures exposure on. If you do this trade on something like CME or CBOE, it’s even worse because they do their accounting in fiat rather than crypto. So, you’d have to convert your BTC to USD and then wire that to the exchange. Slow and expensive. However, consider that CME/CBOE are regulated entities with various levels of protection for parties placing funds into their accounts, whereas BitMEX could disappear one day and you will have little chance of getting your assets back.

Staying on-chain

Enter on-chain derivatives. Imagine being able to put on this derivatives exposure — short BTC/USD — via a smart contract. The simplest implementation is a decentralized p2p model: you find a counterparty willing to take the exact opposite exposure and enter into a smart contract that references an off-chain price oracle.

To replicate the above hedge, let’s say you find someone that wants to be long 1 BTC/USD future, and you will be short. You agree on using GDAX BTC prices as your underlying, and an initial margin of 30%. You configure it so that the contract can be settled at any time by either party, or automatically when either party’s margin collateral balance falls to zero. The smart contract would accept a margin deposit of 0.3 BTC (just keeping this simple, but currently there is no easy way to do this with BTC collateral — you could use ETH pretty easily, though) from the two parties, and then would update those balances periodically by referencing the oracle’s prices. In the above example, when BTC/USD goes up by $100 (from $1000 BTC/USD), the smart contract can convert this amount to equivalent BTC value of 0.091 and transfer that from the “short” margin account to the “long” margin account, leaving them with respective balances of 0.2091 and 0.3909.

This ignores a bunch of details but the benefits are obvious. There is no longer a third party that you give up custody to, and all cashflows are kept on-chain and instantly-ish accessible via smart contract.

Source: https://marketprotocol.io/about

This is the promise, with various implementations proposed, from the likes of: dYdX, LeverJ, EMX, and MARKET Protocol. In a roundabout way, the on-chain prediction market platforms also offer this capability.

Centralized vs Decentralized — Counterparty Risk and Fungibility

Just like with exchanges themselves (see my post on DEXs), derivatives trading can be done through mechanisms ranging from purely peer-to-peer decentralized to completely centralized. The tradeoffs are a bit different than with spot exchanges, though.

You can imagine a centralized derivatives exchange that avoids taking custody, at least in the off-chain sense, of people’s coins. Instead, they would be on deposit in centralized smart contracts. This is akin to the structure of a hybrid DEX, like IDEX.

The main factors driving the decision between centralized and decentralized derivatives trade is counterparty default risk and contract standardization/fungibility benefits.

Typical “real world” derivatives contracts don’t just disappear when someone runs out of margin — typically they have a predefined expiry date or other specific underlying price-related clauses that allow early close-out. As a result, if your counterparty — whoever is on the other side of your trade — fails to keep their margin collateral balance positive, you face the risk of default. You might not get paid what you’re owed, as stipulated by the terms of the contract!

This centralized vs decentralized tradeoff exists in the current derivatives world, as well. The decentralized approach, where traders deal directly with each other on custom terms, is common for major financial institutions. You may have heard of “swaps markets”, the murky underbelly of finance. There’s nothing nefarious going on, but rather a big mess. A tangle of customized agreements and bilateral contracts between a large set of institutions. For an institution that wants a non-standard (and non-fungible) derivatives exposure, such as an interest rate swap designed specifically for a particular bond issuance of theirs, the best way to accomplish this is to go to a swaps dealer (typically a top tier investment bank) and have them construct a custom contract for you. This gives you the exact exposure you want. In return, you face the risk of your counterparty failing to pay up when necessary, among other costs. The big messy tangle of bilateral swaps stacked on top of each other — contracts are rarely closed out since it’s often easier to just enter into another one that cancels out the original exposure, increasing the mess — was one of the big issues raised following the 2009 financial crisis and reformed with Dodd Frank.

By studying how centralization and decentralized works, and doesn’t, in traditional markets, we can make better decisions about how on-chain markets will develop. You can be sure that regulators will step in and have their say on this matter.

Centralized trading occurs on exchanges or regulated execution facilities, like the CME. Centralization provides extra layers of protection against default, specifically through central clearing which provides an additional layer of protection. Trades on a centralized derivatives exchange all “face off” with the central counterparty who will step in if the trader on the “other side” defaults. In fact, there’s also a separate clearing firm on top of that. These layers add complexity, but can be quite useful when extreme market conditions cause individual traders and even brokers to go bankrupt.

In a centralized model, Alice and Bob do not directly face off each other. Instead, Alice goes through Brokerage/Clearing Firm B who interacts with the exchange clearinghouse C, in turn dealing with Brokerage/Clearing Firm D whose customer is Bob.

Source: http://www.iflr.com/Article/3437700/Derivatives-after-the-crash.html

We will see the development of both approaches in the crypto derivatives space. EMX seems to be taking on the centralized clearing model. On the other end of the spectrum, we have pure protocol plays like dYdX and MARKET Protocol which are seeking to enable completely decentralized trade (think “swaps markets”, for better or worse). One thing they will all face is the specter of…

Derivatives trading regulations

Given my background in algorithmic trading, much of it in the US futures market, I was interested in building something for the crypto derivatives space. About a week on the phone with attorneys with CFTC expertise has thoroughly dissuaded me of that. Derivatives trading regulations, in the US and other developed markets, are often more rigorous than for trading currencies, commodities, and in some cases even securities. This is for good reason. Derivatives trading involves leverage, and often unusual non-linear payoff structures (see: options contracts), which can result in you losing your shirt much quicker than you thought possible.

Suffice to say, centralized projects like EMX will have a hard time finding an offshore jurisdiction that will be friendly to them. I am very surprised major legal action hasn’t been brought upon BitMEX and others who continue to not do much to bar US participants (quite likely, an investigation is well underway). Criminal charges are possible! The CFTC has not hesitated to act on illicit crypto derivatives trading in the past.

Furthermore, I’m not convinced that even the fully decentralized approaches will be immune to CFTC scrutiny. They really, really don’t like people trading derivatives, or betting on things, or trading with leverage, without their oversight. How can they stop an open-source protocol? I suspect they will go after whomever/whatever they can, be it people behind websites that list (and especially if they match) open orders, or the developers of the protocol to the extent they are deemed to be contributing to (and especially if they’re profiting from) unregulated derivatives trading. If the CFTC doesn’t bring actions, you can bet that civil lawsuits will be coming down the pipeline once the first major thefts (from smart contract bugs, for example) or defaults occur.

Lending markets in crypto

There’s a lot of talk over lending markets in crypto. Basically, pledge some crypto assets as collateral and you receive a loan against them, either in fiat or more crypto. The main off-chain players are SALT and Unchained Capital, providing fiat in exchange for holding your crypto in a multi-sig wallet with the ability to liquidate it if the value of the collateral falls enough. The on-chain market is not quite as functional, but the biggest player is MakerDAO with over $50 million in outstanding ETH-collateralized Dai loans. The UX is basically non-existent, but the loans are cheap (0.5% paid as a “governance fee”), so… tradeoffs. Newer entrants include Dharma, Compound, and Lendroid.

(As an aside, there’s the fascinating growing world of fully collateralized counterparty-risk-free BTC lending that drives the Lightning Network. This is less applicable for my current topic, but very interesting. See Nik Bhatia’s writing on the topic for more.)

What does this have to do with trading derivatives? These lending facilities, especially the on-chain ones, are and will continue to be used largely to facilitate shorting and leveraged speculation. Similarly to how the securities lending market works in traditional markets, it functions as a healthy counterpart/alternative to derivatives. For example, rather than entering a short futures (derivative) position to hedge a long BTC exposure, you can borrow more BTC against your long and sell that, thereby effectively hedging yourself. In fact, from a cursory glance it looks like dYdX’s main approach to constructing short exposures is via a similar structure where collateral is locked up in a lending contract. In addition, this is currently the most common way of taking large short positions in crypto trading. The OTC lending market — off-chain peer-to-peer/bilateral between well-capitalized counterparties — is currently thriving.

One catch is that these lending facilities will generally require overcollateralization — you can only borrow against a percentage of your collateral. This is especially necessary when lending to effectively anonymous borrowers through a smart contract, from whom it will be quite difficult to collect in the event of a default. As a result, relying on being able to liquidate collateral to cover the loan is pretty standard. This is in contrast to how derivatives contracts are typically margined — they tend to be undercollateralized (like in the previous example where you get short 1 BTC with only 0.3 BTC worth of margin collateral), where you get more exposure than your margin deposit is “worth”.

A potential benefit of using borrowed crypto to enter positions, rather than relying on derivatives, is you can drop the dependency on an external oracle. You make actual trades in the physical crypto — with all the attendant custody risks, of course — rather than having to rely on an off-chain oracle to deliver prices reliably, and a smart contract to correctly settle the outcomes. Similarly, you no longer face the counterparty risk of someone defaulting on the other side of your derivatives contract. Once you borrow some crypto, you are free to sell/keep it as you please and no one can force you out of that position during the life of your crypto loan. These are potentially major benefits of using borrowed crypto to modify your exposure, as opposed to relying on derivatives. In many ways, crypto derivatives and lending markets are complimentary and competitive — they are often used to achieve similar outcomes, and will be traded off against each other for specific use cases.

Stablecoins?

I won’t say too much more since this post has gotten long enough already, but if you’ve been following along you may have already figured out the relationship to stablecoins (disclaimer: I am working as an advisor and helping with protocol simulation analysis at Reserve, a stablecoin project). Stablecoins seek to bring a low volatility cryptocurrency to market — many are pegging to the US dollar. The current champion is Tether with a market cap of over $2.5 billion. What do stablecoins have to do with crypto derivatives and lending markets? Access to these financial instruments allows for the creation of stablecoins from arbitrary instruments. Notice the original example where we own 1 BTC and short 1 BTC future against it — this effectively gives a USD neutral exposure…a stablecoin! Similarly, you can use a lending market in various ways to end up with stable exposure. In fact, the whole point of the MakerDAO system is to create the Dai stablecoin.

Wrapping up

• Crypto derivatives are important, very useful, and coming soon whether we like it or not. Let’s be smart about how we structure them so we avoid massive waves of default, and/or people going to jail for breaking futures trading laws
• Keeping everything on-chain has nice benefits for cash flow management and mitigating some counterparty risks — this was the big promise of blockchain, after all
• Crypto lending markets are already popular and will only grow. They can serve similar purposes as derivatives, with some benefits (reduced counterparty risk and oracle reliance) as well as costs (more collateral is typically required along with the increased cost of capital)
• Crypto derivatives and lending allow for many neat tricks, including the synthesis of exposures that stay stable in a fiat currency (a common goal of “stablecoins”)