Gain DAO’s Proof of Concept: stress testing the trading algorithms

Have you heard of the pioneer of Hybrid Finance already? Meet Gain DAO! Gain DAO is an Ether-based pool, powered by machine learning optimized trading algorithms that are operating in traditional financial markets. This way, Gain DAO serves as a bridge between centralized and decentralized financial systems, leveraging their strengths and creating Hybrid Finance. As a GAIN token holder, you benefit from the possible appreciation of the underlying base asset (Ether) and from algorithmic trading strategies intended to grow the amount of Ether in the Gain Pool. The current trading algorithms are called Zen, Neo and Rey. They each have their own specialty, namely scalping, mean reversion and profiting from momentum.

But are these trading algorithms effective, even under challenging circumstances? Gain DAO recently concluded a Proof of Concept, where the three algorithms were put to the test. Three phases of testing were performed using the algorithms under live market conditions. For these tests, IC Markets was used as the Broker. You can read about the results below.

Phase 1
Period: Approximately 1 month.

The objective of this period was to stress test all three strategies working together with a very low balance. The low balance stress test was designed to see if the strategies could maintain profitability and at the same time limit drawdown to less than 20% despite being severely undercapitalized. The lack of capital forces the systems to adapt their risk parameters to compensate for being undercapitalized.

We observed that Neo did not place any trades during Phase 1 due to the capital restrictions. Zen only managed to place one trade. The rest of the trades were denied due to lack of available margin. Rey was able to place trades which resulted in a 87.83% gain during phase 1.

Conclusion: PASS

The portfolio adjusted to the reduced capital by restricting the use of Neo and Zen and nearly exclusively traded Rey. The end result was an 87.83% return during phase 1 while staying under the 20% drawdown threshold. While the returns were impressive, the risk of operating under such conditions does not fall in line with the risk profile for the Gain DAO Fund.

Phase 2
Period: Approximately 6 months

The objective was to observe all three trading systems operate while properly capitalized. The 20% drawdown limit was maintained. Neo, Zen and Rey all traded with frequency due to the increased trading capital. Rey, which trades the GBPCAD was affected by excessive volatility caused by the United Kingdom exiting the European Union. This resulted in a 20% maximum drawdown limit to be triggered on November 10th, causing the portfolio to incur its first negative month. Due to these adverse conditions for the Rey strategy, we disabled this strategy for the majority of December. Neo on the other hand was able to capitalize on the increased volatility by producing some impressive gains in the month of December. All three trading strategies were then enabled for January forward.

Conclusion: PASS

Despite one of the three strategies struggling the portfolio as a whole grew at a robust rate, which further demonstrates the importance of using a portfolio of trading strategies that excel in different conditions as opposed to only using one trading strategy. This smooths out the equity curve and leads to a better investor experience.

Phase 3
Period: Live Trading with 1000 Ether

For phase 3, the objective was to test the scalability of the system when trading with larger amounts of capital using the base currency Ether. The trading characteristics of the fund seem identical to Phase 2. Much of this is due to the liquidity providers ability to execute these larger orders and scale with the demands of higher volumes.

Conclusion: To be determined

To avoid poor order fills and other such downsides to trading with higher volumes, we conclude it is only appropriate to trade with financial instruments that have sufficient liquidity to execute orders effectively. So trading will be limited to major currency pairs. This is not an issue for Rey or Zen, however Neo does trade some instruments with limited liquidity, so those instruments should be dropped.

In summary, the three strategies together are complementary and lead to a smooth equity curve. Scalability is not an issue as long as only major currency pairs are traded.