HackMD
Is it worth to take an individual RAI position? Static Simulations & Automated Strategies
Up to date by October 2022
Authors: Danilo Lessa Bernardineli (BlockScience) and Michael Zargham (BlockScience)
This document describes some of the prior work done in order to measure the rationality of creating and manipulating an individual RAI position based on the market and controller conditions. Static Simulations were done in order to provide readily consumable benchmarks, while work on describing the space of repositioning strategies and their relationships to the system were initialized in order to collect requirements for how an comprehensive automated trading strategy should be.
Static Simulation of an RAI position
For measuring the baseline rationality of creating an RAI position, we've done an static simulation on which we execute an specific strategy and create benchmarks using counter-factual scenarios in order to compare the relative ROI.
Conclusions
- An RAI position can be partially understood as an leverage on the USD/ETH price. Positive or negative swings on it can generate outsized gains or losses as it allows the asset size to move relatively elastically while the liability size is relatively fixed.
- For instance, for a ~234% CR position, having the ETH price going up or down 1.4x fold can generate about 1.2x fold gains or losses vs an ETH hold position, or 1.4x fold gains or losses vs an USD hold position.
- For absolute numbers, have a look on the spreadsheet.
- For instance, for a ~234% CR position, having the ETH price going up or down 1.4x fold can generate about 1.2x fold gains or losses vs an ETH hold position, or 1.4x fold gains or losses vs an USD hold position.
- The discrepance between the Market Price and Redemption Price creates opportunities for one-shot arbitrage gains. Together with an stable and negative redemption rate, this can reduce the risk associated with an USD/ETH leverage.
Key Assumptions
- Stability Fees were not taken into consideration.
- SAFE liquidation is not handled explicitly
Position Strategy
- Creating the RAI position
- Allocate an fixed amount of ETH
- Lock all the free ETH & Borrow RAI on given ratio (eg. 2:1)
- Sell the borrowed RAI for ETH
- Lock all the free ETH & Borrow RAI on the same ratio again
- Sell the borrowed RAI for ETH again
- Lock all free ETH
- Withdrawing the RAI position
- Allocate the exact amount of ETH for paying off the RAI debt on the position
- Swap it for RAI on the secondary market
- Pay the RAI debt
Counterfactual Scenarios
- USD Hold: What would happen if all the allocations for creating / closing the position were simply stored as USD somewhere?
- ETH Hold: What would happen if all the allocations for creating / closing the position were simply stored as ETH somewhere?
Spreadsheet for the Static Simulation
| State | Benchmarks |
|---|---|
Automated Strategy
We've been collecting requirements for what should an automated strategy achieve in order to fullfill the full potential that the RAI admissible set of actions can achieve.
On the most fundamental form, an complete bot should be able to execute 6 types of repositioning strategies, as described on Fig 1 which are as follows.
- A (Neutral to Short)
- B (Short to Neutral)
- C (Neutral to Long)
- D (Long to Neutral)
- E (Long to Short)
- F (Short to Long)
Fig 1. Potential strategies for an RAI position automaton
More specifically, strategies A and B are illustrated by figure 2.
Fig 2. Strategies for the transitions A and B
More generally, the relationships between an individual SAFE, the secondary market, and the controller are described by figure 3.
Fig 3. Relationships between the Position and the Primary and Secondary Markets
