INTRODUCTION

Resolv Labs is building the protocol that employs a novel stablecoin design backed by delta-neutral strategy. While the delta-neutral collateral concept is not new and has been explored in the past, we suggest a new approach to structuring economic terms of the protocol. In Resolv, the stablecoin (USR) is supported by the protection layer, which absorbs exposure to funding rates and centralized infrastructure. This layer is designed to be a fungible yield-bearing investment instrument for investors with moderate risk appetite — Resolv LP (RLP).

The tokens are backed by treasury consisting of staked Ether and short Ether futures positions. Both of these components generate profits, which are then distributed to the Resolv tokenholders.

In this article, we provide an overview of where this yield comes from, how sustainable it is, and how Resolv distributes it to token holders.

YIELD SOURCES

Resolv benefits from two major sources of yield — ETH staking and funding rates generated by perp futures positions.

ETH staking yield is a more predictable part of the equation. Since migration of Ethereum Mainnet to Proof-of-Stake, the staking yield has consistently been floating within 4–6% p.a. range with mean value just above 5%.

Lido stETH APY chart

On the other hand, funding rates are far more volatile. Also, despite similar patterns driven by market-wide movements, their behavior changes materially across trading venues where the futures are traded. This is not only due to supply/demand structure specific to each exchange, but also because of technical factors related to calculation of funding rates.

Let us start by looking at historic performance of the strategy (stETH + model futures portfolio: 40% Binance; 40% Bybit; 20% Deribit). This portfolio is chosen as a simple model close to current base-case Resolv treasury composition. For the avoidance of doubt, we did not optimize the composition to make yield look more solid.

We observed 30-month period spanning across 2021 and 2023 as the timeline with relatively established futures market in terms of liquidity and rates (spoiler: earlier periods saw massively bullish trends with delta-neutral yields in 20–60% APY regions!).

A few figures to put things into perspective:

• Year-to-date APY of the strategy is 10%;

• 2022 saw an APY of just 4.06%, affected by a row of black swan events (Terra, Celsius, FTX), rather than general bearish sentiment;

• Observation period APY is at 8%.

SUSTAINABILITY

Needless to say, historic performance does not guarantee future returns.

To evaluate how market performs depending on the cycle, we have developed a statistical model and tested it against the real data. The key requirement for the model is to realistically model the structure of the historical funding rates (mean tendency, standard deviation, tails etc.) but, at the same time, be flexible enough to allow changes in parameters appropriate for different market conditions.

Which scenarios can be simulated with different model parameters? Let us look at an example from the experiments we conducted. The histograms below show the distribution of real funding rates (orange) on different exchanges (data from 2021–2023) and what we simulate under the assumption of a bearish market which results in bias towards near-zero or negative funding rates. Note that the model still captures the overall shape of the market but at the same time provides the change in the structure.

The primary objective here is to conduct stress-testing for bearish markets since consistent negative funding rates might drive the strategy to negative APY. In fact, this is not something unrealistic — in 2022 the market saw near-zero mean funding rates.

Model vs actual funding rates histograms for Binance, Deribit, Bybit

How would Resolv perform in such scenario? Imagine Resolv delta-neutral strategy is live for 365 days in 1000 parallel universes. Let us calculate overall APY of the strategy for each year and plot the distribution of one thousand APYs.

Model annual performance of Resolv delta-neutral strategy, stress-test scenario

Median yield is at ~5.5% and 99% of yields are above 2%, meaning that even in such a bearish scenario it is highly unlikely that protocol will suffer losses.

Still, can the overall yield be negative? In principle yes, but it is unlikely. In practice, arbitrage trades of holding short ETH and long ETH futures, would then be substantially profitable. We have also covered negative funding rate considerations in our previous article dedicated to risk assessment.

HOW RESOLV APPROACHES YIELD DISTRIBUTION

Using that approach we are also able to evaluate protocol metrics such as performance of staked USR and RLP tokens.

But first, let us remind how Resolv yield distribution works:

1. Protocol profits (or losses) are observed over each of the 24-hour reward epochs;

2. Epoch profits are distributed in three parts:

• Base Reward — to holders of staked USR vaults and RLP;

• Risk Premium — exclusively to RLP;

• Protocol Fees — to the protocol treasury.

3. If the protocol realizes a loss over reward epoch, it is absorbed by RLP. No distributions are made from the protocol then.

4. Worth mentioning — USR holders need to stake their tokens in order to receive the yield. Plain USR is a ERC-20 token with the utility of value retention.

At launch Protocol Fees will be set up in a range of 10 to 20%. Weightings of the Base Reward and the Risk Premium will be calibrated so that the staked USR yield is competitive with money market rates across DeFi. Changes to weightings mechanics during the life of the protocol will be a governance matter.

Now, with that in mind, let us illustrate how yields of each instrument are expected to look like in base case scenario. We will also look at optimistic case to estimate how the tokens will perform in bullish cycle.

BASE CASE

Let us again run our simulations to arrive at the results. We note here that yield distribution weightings were calibrated to result in 6% average yield for USR staking.

Histograms for yield distribution in base case, 1000 simulations

Resulting metrics are shown in the table below:

An important highlight is that in base case, RLP will be allocated 23% APY assuming its TVL comprises 15% of the aggregate protocol.

BULLISH CASE

Modelling optimistic yield scenario is a pleasant exercise, but it also provides insights for what to expect in bullish cycle, when average funding rates gravitate towards 10% p.a. area. In our case, we have modelled a moderately bullish scenario based on funding rates observed so far in 2023:

Histograms for yield distribution in bullish case, 1000 simulations

These are great results giving USR holders at ~7% mean APY; while RLP yields are in 30+% area.

Taking this into perspective, RLP yields in a range of 20–30% represent a competitive compensation of interest rate and counterparty exposure. We are truly looking forward to launching this instrument!

CONCLUSION

Funding rates are indeed volatile and can materially affect daily performance of delta-neutral strategies. Still, over the longer-term, such portfolio on average performs better than other sources of money market yield, e.g. US Treasuries.

It is fair to say that even in the bearish scenario modelled results reflect less than 1% probability of generating overall losses over a 1-year period and median result of 5.02% APY.

Moderately bull market model assumptions result in solid performance with 99% simulations ending up within 9%-12% APY with further upside to 15% APY area in more bullish scenarios.

Resolv adjusts revenue distribution between USR and RLP, so that USR stakers enjoy stable money-market yield (as we are convinced stablecoin designs should aspire to!). Yield volatility is allocated to the RLP holders in exchange for better expected returns, making it a competitive yield instrument for investors with moderate risk appetite.