Ctrlk

Prehistoric Vaults — RevertX

Risk Level: Medium

📓 Lab Journal Entry #03

Title: RevertX: Mean Reversion Under Observation Author: Byteceratops, Vault Systems Engineer Recovered From: Strategy Terminal Δ-9, Signal Behavior Unit

Purpose

This log outlines the trading behavior and mechanics of the RevertX Strategy. This vault is designed to hunt for mean-reversion between $ETH and $BTC using a dynamic, signal-driven trading model.

Overview

RevertX was developed to hunt price symmetry between $ETH and $BTC — a behavior we’ve observed frequently in the wild. When the relationship between the two assets drifts beyond a defined threshold, the vault opens long/short positions to capitalize on the expected snap-back.

  • Target Assets: $ETH / $BTC

  • Style: Pairs Trading (Mean Reversion)

  • Core Signal: Deviation from regression-based fair value

  • Trade Entry Rule: Signal > 1.5× standard deviation

Signal Mechanics

The strategy opens a long or short position when $ETH's price meaningfully diverges from its expected value relative to $BTC. This expected value is calculated via a regression that updates daily using the past 20 days of price data.

Signal Formula

Signal = ETH Price − (β₁ × BTC Price + β₀)

Where β₁ ≈ 0.077 and β₀ captures longer-term drift.

Figure 1: Trading signal fluctuates around the fair value (black line). Trades are triggered when the signal breaches the red (long) or green (short) thresholds.

The RevertX-trained Rexes are conditioned to strike only when this signal breaches ±1.5σ — filtering noise, conserving energy, and maximizing trade precision.

Backtest Results

To observe RevertX in the wild, we deployed it across historical terrain spanning late 2017 to early 2024. Using real market data and a consistent $1,000 trade size, the system completed 372 full excursions — each representing a signal-triggered round-trip trade.

Metric
Result

Annualized Return

~53%

Max Drawdown

$112 (11.2% of trade size)

Average Trade Duration

~5 days

Win Rate

33.6%

Despite a low win rate, the strategy remained net-profitable due to its strong reward-to-risk dynamic, favoring frequent small losses and infrequent large gains.

Figure 2: RevertX performance over time. Red line shows cumulative PnL. Signal entries are aligned with ETH/BTC divergences.

Notable Behavior

  • RevertX prefers frequent small losses to wait for occasional large wins

  • All trades are stat-tested for signal stationarity

  • Optimized for short trade cycles and capital rotation

🧪 REX Technician’s Log

We fed the RevertX-trained Rexes a steady diet of $ETH-$BTC residuals. When the signal roared above threshold, they struck with precision. Most excursions resolved in under 5 days — unless the jungle changed its rules.

End of Strategy Log

Signed: Byteceratops — authenticated Reviewed by: Dr. Morales, Risk Oversight Unit

Last updated