EndoTech System Performance Use Case
12-25-2018

We set up the ET system to reduce risks in the event of slow December volatility. It introduced several return targets: 

  • First target of 5-8% return for 25-30% lots
  • Second target of 10-12% return for 15-25% lots
  • Third target of 20% return for 10-15% lots

In specific trades executed between December 17th and December 25th, 2018, the system decided to place hard-level limits instead of dynamic limits based on the unconstrained mean (UM) model, the unconstrained mean model with correlation (UMC), and more.

Subsequently, the entry was as follows (based on Kraken prices):

XRPUSD (19% return from invested capital)

  • Bought at $0.30647
  • Sold 25% at $0.33007, sold 25% at $0.36129, sold 13% at $0.37812, and sold the remaining 27% at the potential end of the trend at $0.3555
  • Reentry occurred with a buy at $0.36579 and a sell at $0.38038

BTCUSD (8% return from invested capital) –

  • Bought at $3515.00
  • Sold 25% at $3715.30, sold 19% at $4018.40, and sold the remaining 56% at the potential end of the trend at $3773.90

ETHUSD (21% open return from invested capital) –

  • Bought at $91.22
  • Sold 29% at $94.54, sold 18% at $101.03, sold 13% at $108.94, and remaining 40% are still open
Graph highlighting EndoTech’s ETH/USD order execution performance between 17 Dec and 25 Dec, 2018
Endotech ETH/USD Orders (17-25 Dec, 2018)

The placement of buy and sell orders in ETHUSD between December 17 and December 25, 2018 (based on Bitfinex prices)

Note that:

  1. The EndoTech system is set to be stable and maximize returns within a time horizon of one year. It does NOT attempt to optimize each trade. This is statistically correct. Optimizing systems to catch each trades’ highs and lows leads to overfitting — a situation demonstrably resulting in higher risks and lower returns.
  2. The system is not optimized to handle extreme cases (74% return is considered an extreme move). Arguably, ET Bot was supposed to exit the remaining lots at 70%. However, nobody could prevent the market from returning 100%-120% and 150%, and then we would be in the same position — exiting or not exiting optimally.
  3. The system performed correctly — it entered at the very start of the trend, locked in the returns, and exited the position when the trend stopped, resulting in significant returns.

Leave a Reply