- Set Protocol, and its first user-facing interface, TokenSets, leverages existing Open Finance protocols to provide seamless, global access to censorship-resistant algorithmically adjusted structured products
- The possibilities for these exotic structured products are quite literally endless, and, with some tinkering, could pave the way for the proliferation of competitive, trustless investment funds
This free preview of The Block Genesis is offered to our loyal readers as a representation of the highly valuable journalism & research our Genesis members receive daily. If you’d like to receive all Genesis content on our site and via daily newsletter, join here.
Open Finance infrastructure continues to proliferate. MakerDAO laid the initial foundations with the launch of Dai, the dollar-pegged decentralized asset. Uniswap, 0x, DutchX and a host of alternative protocols have paved the way for trustless asset exchange. UMA, Rainbow Network, and dYdX have since opened the decentralized design space to derivatives and synthetics.
Now, in a clear demonstration of the power deriving from the intersectionality of open finance systems, Set Protocol, and its inaugural user-facing interface, TokenSets, is consolidating these innovations into a single platform, providing seamless, global access to censorship-resistant algorithmically adjusted structured products.
Set Labs Inc.
Set Labs Inc, the six-person, San Francisco-based team behind Set Protocol and TokenSets, launched in November 2017 and proceeded to raise a $2.2 million seed round led by Craft Ventures in March 2018. TokenSets is the first user-facing application on Set Protocol, a series of smart contracts that allows users to mint baskets of assets in a single click.
The first class of sets was released in 2018: Ethereum 10, which provided exposure to the top ten erc20 tokens weighted by market capitalization; Stablecoin Set, which contained a mix of DAI and TrueUSD, and DEX Set, a thematic basket featuring tokens from 0x, Kyber, and Airswap. These initial sets can still be accessed via the legacy TokenSets interface.
In recent months, Set Labs Inc. has shifted focus away from these static sets to a product that is capable of rebalancing without a trusted counterparty. To that end, the team has invented a means through which to integrate rebalancing logic into Ethereum smart contracts, essentially replicating the role of robo-advisors in a trustless, censorship resistant fashion.
Today, TokenSets offers users two core products: a Buy & Hold strategy and a Range Bound strategy. Within these products there are sub-strategies catered towards those wishing for ETH exposure and those more inclined towards BTC.
Buy & Hold
The Buy & Hold strategy is probably most familiar to today’s crypto investors. A BTC Enthusiast set will maintain a 75:25 balance between BTC and ETH, rebalancing on a 30-day basis if the price allocation moves by more than 2%. At present, the BTC Enthusiast set has a price allocation of 77.65% and 22.35% between BTC and ETH. If this allocation remains constant for the next 11 days the set will be rebalanced, with enough BTC sold for ETH in order to return to a 75%-to-25% ratio.
Range Bound Strategies
The Range Bound strategies are considered to be more amenable to neutral, range-bound markets. Each set is comprised of a mix of the volatile asset (i.e. ETH or BTC) and a stable asset (at present, DAI). The set automates buying the bottom and selling the top within a particular range, thereby capitalizing on intermediate volatility even when the underlying asset remains fairly constant over a longer period of time.
In order to execute these range-bound strategies, the set designer must define a static lower and upper percentage bound that triggers a rebalance once the volatile asset exceeds either threshold. Two questions then arise: how best optimize these bounds to perform profitably over different market cycles, and how to evaluate performance in the first place.
As for optimizing bounds, Set Labs Inc. ran Monte Carlo simulations, approximating potential price histories to test for the best parameters. Price histories were limited to those where prices ended within -30% and 50% of the starting price. The Low Volatility (LOVOL) version of the Range Bound product sets an upper and lower tolerance of 58% and 42% ETH/BTC, respectively, before rebalancing back to a 50:50 distribution. The High Volatility (HIVOL) set uses upper and lower bounds of 58% and 33%, protecting against higher price moves, which are generally more prevalent in today’s crypto markets.
As for measuring performance, Set Labs has decided to optimize for risk adjusted returns, formally quantified through the Sharpe Ratio, which measures returns minus the risk-free rate divided by the standard deviation of the return. A higher Sharpe Ratio is favourable, indicating better returns per unit of variance of those returns.
The chart below illustrates the relative performance of the ETHHIVOL and ETHLOVOL sets versus pure ETH exposure in the period from May 12, 2016 – Jan. 16, 2017. Starting with $100 in each product, pure ETH exposure would net a loss of 2.18%, while the ETHHIVOL and ETHLOVOL sets net a gain of 10.45% and 12.61%, respectively.
Source: Set Labs Inc.
Comparing Sharpe Ratios across the three products using monthly returns produces the following results:
Source: Set Labs Inc.
Thus, in a relatively flat market, where ETH ranged between lows of ~$6 and $20, the ETHHIVOL set outperformed on a risk-adjusted return basis, whereas ETHLOVOL outperformed on a total return basis.
When examining the time period between May 2017 and the present, returns and Sharpe Ratios across the three different strategies look as follows:
Source: Set Labs Inc.
Source: Set Labs Inc.
While visually the ETHLOVOL and ETHHIVOL products appear much less volatile — and more profitable – than a pure ETH strategy, the Sharpe Ratios calculated on a monthly basis tell another story.
However, if you consider that these range bound strategies perform best over longer time horizons, and adjust the volatility calculations to a three-month basis accordingly, the Sharpe Ratios change to the following:
Source: Set Labs Inc.
On a three-month basis, the range bound strategies fall further in line with a pure ETH strategy, with all three products sitting between ~0.36-0.40.
The table below indicates the optimal holding periods for the various product strategies:
Source: Set Labs Inc.
In the immediate term, Set Labs will be looking to introduce even lower volatility range bound sets, providing a product best suited for holding on shorter time frames — months or quarters.
Further, Set Labs will look to incorporate Compound v2-based tokens. These tokens are backed by supplied assets, allowing Set holders to earn interest from their underlying coins and tokens, similar to the manner in which ETF providers lend out stocks on behalf of their owners.
Most exciting, however, is the potential future design space for sets. Similarly to platforms like Motif, which offers users the ability to invest in specialized baskets of equities — think High-Yield Dividends, (Consumer Staples, Utilities, Communication Services), Millennials #IRL (Internet & Social Media, E-Commerce, Apartment REITs), Couch Commerce (E-tailers, Travel, Auctions & Marketplaces, Discount E-tailers), Rising Interest Rates, Precious Metals, Bear US Market etc. — Set Protocol permits third parties to create any variation of a set themselves.
However, Set Protocol goes far beyond platforms like Motif in that any asset of any class can be included. If a user was particularly bullish on Andrew Yang, known for, among other things, his predilection for blockchain technology, they could buy a Set composed of a 4x long interest Ethereum token from dYdX, a Andrew Yang Presidential Election Veil share, and a synthetic asset via UMA pegged to the NASDAQ-100 ETF. If a user were convinced a global recession was around the corner, they could buy a set composed of a mix of bitcoin, a synthetic asset pegged to gold, and a synthetic asset pegged to the Short S&P500 ETF, all of which rebalance periodically depending on pre-set criteria. An investor bullish on MakerDAO stakeholders’ ability to return Dai to its peg could buy a set containing DAI and USDC, with automatic buys of DAI any time it trades below a dollar. Combine this with some MKR exposure and Compound DAI-pegged tokens and you suddenly have a algorithmically-driven bullish MakerDAO strategy. The possibilities for these exotic structured products are, quite literally, endless, under the assumption that good oracles can be sourced for each synthetic asset.
Using zero-knowledge technology, third parties with sophisticated investment backgrounds could even offer up sets of their own design for some kind of origination fee, using zero-knowledge technology to obfuscate the contents of the baskets while still regularly posting Net Asset Value, like the non-transparent ETF developed by Precidian Funds. If you throw in some kind of performance fee, this essentially allows any person worldwide to become a fund manager in a trustless, competitive, and provably fair fashion, perhaps completing the vision originally set out by Numerai.
The tax implications of buying a Set will vary by jurisdiction, with some nations possibly requiring Set owners to pay capital gains on a per-rebalance basis, while others may choose to treat the basket of assets as a single asset — buyers should be sure to consult with tax and financial advisors. If, however, the latter scenario were to be prevalent then Sets present a monumental step forward for crypto investors, allowing them to execute trading strategies that would typically net a hefty tax bill without the burden of paying capital gains on each trade.
As with most existing Open Finance products, the Set target market is currently restricted to retail customers, due to both the relative sophistication required to navigate the cryptocurrency industry combined with, largely by design, a lack of Know Your Counterparty procedures, the latter preventing institutions from participating in deregulated markets. However, in the near-term, sourcing liquidity for Set assets from permissioned pools of 0x relayers and having sets trade on regulated non-custodial secondary markets like IDEX could pave the way for institutions to participate at a later stage. User-facing interfaces building on Set Protocol could also seek to acquire an Alternative Trading System license.
Under the hood
Set Protocol’s one-click purchase process belies the enormous volume of activity happening behind the scenes.
An order submission initiates an ETH transaction to a Set Protocol smart contract as well as information as to how it should source liquidity and the type of Set to mint. This data is then encoded into a bytestream and sent to an exchange — currently Kyber – which proceeds to execute all orders in one atomic step. Price slippage from these market orders is a function of order size and users are presented the expected cost of execution, as well as gas fees, before purchase.
The rebalancing process shirks exchange orders for a Dutch Auction mechanism, which is similarly favoured by protocols like MakerDAO and DutchX for its ability to maximize revenue for sellers, avoid front-running from miners, and source liquidity from a variety of venues. Any third party can participate in these auctions, where the protocol smart contracts dispose of one asset for another in order to meet the Set distribution criteria.
At present each auction runs for 24 hours but as more third-party ‘keepers’ start participating — thereby bringing more liquidity and better price execution – the auction time can be cut to hours or potentially minutes. The incentive for these third parties is to bid somewhere below the fair market value of the undesired assets, netting several basis points via arbitrage in the process. As a result, set holders can always expect some degree of slippage, but once again, this should decline over time as the auctions gain traction and competition. In the near term, Set Labs will be participating in these auctions and complete sales at fair value. In the future, they may participate as a competitive bidder, using the auctions as a source of revenue.
As for oracles, Set Protocol currently relies on MakerDAO for price feeds. In fact, Set Protocol is the first instance to deploy the Multi-Collateral Dai price oracles. The decision to leverage MakerDAO oracles comes from the notion that the Maker Foundation team has proven themselves to have longevity: if a stronger, less-exploitable oracle were to appear down the line it is likely that Set Protocol would transfer over. Set Protocol does not currently use a price feed for Dai, with Sets instead honouring the peg despite Dai’s recent trade history below one dollar.
A new paradigm
According to the European Structured Investment Product Association (EUSIPA), European structured product exchanges alone generated 27.8 billion euro in turnover in Q4 2018. The total market size for these kinds of instruments is not to be underestimated. Factoring in the global nature of Ethereum, which will alleviate concerns around the fragmented liquidity across specialized products, Set Protocol represents a fascinating new paradigm and, considering the limitless constructions, perhaps the most remarkable addition to the Open Finance movement yet.