Data on DerivaDEX

Data for traders, Data for the DAO!

DerivaDEX has a completely transparent orderbook and execution flow. Everyone has access to the same data, at the same time: exchange operator nodes, liquidity providers, traders, and the DAO.

If you are interested in doing your own data analysis, you can run the auditor at apidocs.derivadex.io and get started. DerivaDEX combines the transparency of on-chain trading with the speed and efficiency of centralized exchanges. Do you have ideas for new analysis or charts after reading this article? Hop in the Discord and ask!

📊 Data on DerivaDEX can enable:

• Shadow-trading of 🐋 whale accounts (all trade data is publicly available)
• Real-time transparency into all open positions on the exchange
• Total transparency of the insurance fund capitalization and drawdowns in real-time so users can evaluate ADL* risk.
• Visibility into total capital inflows and outflows from the exchange

Reminder: ADL stands for “automatic de-leveraging”, where a positive position is forced to close (de-leverage) because the liquidated counter-party does not have the capital to pay the winning trader what she is owed.

🙋 All users can answer data-driven questions like:

• How does market-maker behavior change under positive and negative funding rates?
• Is shadow-trading a whale or suite of whales profitable across different time intervals?
• Do changes in the funding rate attract large capital inflows or cause outflows to the TVL of the exchange?

Normally, data at this level of detail is not available to traders using centralized exchanges. Users of on-chain exchanges, such as AMMs, have used this level of data transparency to create DEX-specific strategies (like MEV). While MEV is mitigated on DerivaDEX due to order encryption, new strategies that make use of this level of data transparency will likely emerge.

⬇️ After about two weeks of trading, this article takes an initial look at some of the data from the exchange. ⬇️

✅ Session 2: Quick totals

• Number of traders — 921
• Overall volume — $10,738,208,247.16
• Insurance fund capitalization — >$16mm
• Checkpoints — >1,737

💰 Insurance fund

The insurance fund is capitalized through trading fees and positive liquidations. A positive liquidation occurs when a trader’s position is closed after the liquidation point is crossed, but before the bankruptcy price is met. This requires a swift liquidation engine and deep liquidity.

In Session 2, you can see one very large drawdown event that was handled smoothly. In Session 1, market making bots had reduced liquidity, so the system saw (and handled correctly!) significantly more liquidation events.

Combined, both sessions give strong evidence that the system is robust to price volatility and liquidation cascades, and demonstrate the performance of the insurance fund capitalization mechanism.

Insurance fund capitalization

Insurance fund drawdown

You can see a significant liquidation event at around 1 million requests (the huge vertical line). Under ideal liquidity conditions (market makers with a lot of money), the insurance fund had completely re-capitalized post-loss within half a million requests. Pretty neat!

The DAO will need to decide what size the insurance fund should be. Too large, and the capital could be better used elsewhere. Too small, and ADL risk is higher. The DAO can use data like this, as well as information about the exchange’s current liquidity and volume, to determine the ideal size of an insurance fund. In this scenario, the maximum drawdown was -1.5 million USD.

These fill charts show another view of the liquidation event, and demonstrate liquidation clustering (large liquidations often happen alongside, or exacerbate, dramatic price movement on an exchange, causing a liquidation cascade).

Why should traders care about this data?

• Traders can understand their risk of an ADL, and overall systemic risk based on the capitalization of the insurance fund
• Traders can take advantage of secondary effects of major liquidation cascades such as: a change in the funding rate, beneficial wick trades, expected reversion to mean or trend lines.

Why should the DerivaDAO care about this data?

• The DAO can manage the size of the insurance fund based on observed behavior and drawdowns in different liquidity conditions
• The DAO can make decisions about product listings (and access of certain trading pairs to the insurance fund) based on data about drawdowns and liquidity.

The DAO is in charge of new product listings on DerivaDEX, so ensuring that products on balance benefit the insurance fund (and at least don’t expose it to undue risk) is an important calculation that the DAO must make when undertaking listing decisions.

📈 Position information

Over time, we see a wider spread of notional positions in both ETHPERP and BTCPERP, with ETHPERP significantly outpacing BTCPERP. In the testnet environment, all users start with the same deposit size, but the size divergence between ETHPERP and BTCPERP is interesting.

Why should traders care about this data?

• Is the overall size of open positions in a trading product spread across many traders, or just by one or two whales?

Why should the DAO care about this data?

This data can drive profitable and interesting experimental questions like:

• Is ETHPERP participation higher because it is the default product on the exchange, or because it is a more desirable product?
• Is this affected by relative liquidity on ETHPERP versus BTCPERP?

⚖️ Funding rates

The funding rate charts aren’t super interesting in Session 2 because of the deep liquidity the bots are providing. On main-net, liquidity will likely ebb and flow based on other parameters, and the funding rate will come into play more often. In Session 1, the funding rate was much more dynamic.

As you can see, there was one interval where the funding rate diverged from 0 in ETHPERP and one in BTCPERP. In trading sessions with less liquidity, we can expect this to move. Noticeably, the major liquidation event that is visible in the insurance fund chart is not associated here with a change to the funding rate, meaning the mark price was consistently tracking the price feed of the underlying. This is what we expect to see in a market with efficient and liquid market makers.

Why do traders care?

• Funding rates can provide very profitable opportunities for getting yield on a market-neutral position.
• Funding rates can affect your bottom line if you’re holding an open position, especially if you hold that position across multiple funding rate intervals (a fee is assessed or a payment is distributed every 8 hours)

Why does the DAO care?

• A funding rate that varies too widely can negatively affect trader experience on the exchange
• The DAO may wish to explore productizing funding rate yield opportunities, and observing this data over time and understanding what factors influence it can help evaluate this opportunity.

🔍 Data and Derivatives

Data transparency is a necessary step in the evolution of derivatives markets. Right now, this level of transparency is new, and has many unexplored implications. In the future, we at DEX Labs believe that this model will be demanded by all users.

Expect more DerivaDEX Data posts moving forward, and if you’d like some support getting started or putting together your own analysis, feel free to reach out.

Join the DerivaDAO!

Join the Discord and email list to learn more about upcoming products and early access opportunities.

Ainsley Sutherland is the Product Lead at DEX Labs, lead R&D contributors to DerivaDEX.