Macro Regime Engine (MRE v06)
aka MRE · MRE v06 · Macro Regime Engine
A 26-voter cross-asset system that classifies the global macro environment into one of four regimes — Goldilocks, Reflation, Inflation, Deflation — and outputs a binary risk-on / risk-off decision for SPY. 23-year backtest: +10.86% CAGR, −14.35% max drawdown vs SPY buy-and-hold −56.47%. Seven independent forward tests passed.
Full breakdown on the MRE v06 product page.
Arsenal Bitcoin Trend System
aka Arsenal BTC · Arsenal Bitcoin
A three-state trend system for Bitcoin built on VAMS (Volatility-Adjusted Momentum). Position sizing maps to a single composite vote: 100% on +2, 50% on 0, 0% cash on −2. 10.6-year backtest: +88.10% CAGR, −34.46% max drawdown vs BTC buy-and-hold −83.40%.
Full breakdown on the Arsenal BTC product page.
The 8th Rule
aka 8th Rule
DurdenBTC's original two-engine trend system for Bitcoin and Gold. Uses GVTS for fast trend detection and VATS for volatility-adjusted confirmation. Position sizing ramps from 25% on a probe entry to 100% on confirmation. Free on TradingView — no paid TradingView tier required.
Full breakdown on The 8th Rule page.
VAMS
Volatility-Adjusted Momentum
A composite signal that z-scores price distance from a 63-day quarterly-cycle baseline. The output classifies the market into Bullish (+2), Neutral (0), or Bearish (−2). The architecture underneath Arsenal BTC.
GVTS
Gaussian Volatility Trend Signal
The fast trend trigger inside The 8th Rule. Uses a Gaussian-filtered Hull moving average to identify early trend reversals. Opens trades at 25% allocation when it flips.
VATS
Vol-Adjusted Trend Score
The slower confirmation engine inside The 8th Rule. An ATR-based trend score that scales position size from 25% to 100% once a trend confirms beyond the noise floor.
Goldilocks regime
Growth rising, inflation falling. Equity bull, bond bull, credit-spread tightening. Historically the most reliable risk-on backdrop.
Reflation regime
Growth rising, inflation rising. Equity bull, commodity bull, real-asset friendly. Cyclicals and value lead.
Inflation regime
Growth falling, inflation rising. The toughest regime for nominal assets. Gold and short-duration usually lead; long-duration nominal bonds collapse.
Deflation regime
Growth falling, inflation falling. Long-duration bonds win; risk assets bleed. The regime MRE v06 is most aggressive about avoiding.
Drawdown
Peak-to-trough decline in portfolio value, expressed as a percentage of peak. A −50% drawdown requires +100% to recover; a −83% drawdown requires +488%. Truncating the left tail is the entire DurdenBTC thesis.
Deep dive: Bitcoin Drawdown Management guide.
Vol-adjusted sizing
Scaling position size inversely to recent realized volatility, so total portfolio risk stays roughly constant even as the underlying asset's vol changes. A core tool for drawdown management on high-vol assets like Bitcoin.
Sharpe ratio
Annualized excess return divided by annualized standard deviation of returns. The canonical risk-adjusted return metric. > 1 is good, > 2 is excellent. MRE v06 sits at 1.01; Arsenal BTC at 2.33.
Sortino ratio
Like Sharpe, but only penalizes downside volatility. A truer measure for asymmetric, trend-following strategies where upside vol is the goal.
Calmar ratio
Annualized return divided by absolute max drawdown. Single best one-number summary of return-vs-pain tradeoff for trend systems.
Profit factor
Gross profits divided by gross losses across all trades. > 2 is solid; > 4 is elite. The 8th Rule sits at 4.62 on BTC.
Forward test
Running a strategy on data the parameters were never trained on. The cleanest proof a system isn't curve-fit. MRE v06 ships with seven independent forward tests; all pass.
Methodology: Forward Testing page.
Out-of-sample (OOS)
Data the model has never seen during training. An OOS Sharpe near or above the training Sharpe is the gold-standard signal that an edge is real. MRE v06's OOS Sharpe (1.219) was 62% higher than its training Sharpe (0.754).
Monte Carlo block bootstrap
A statistical test that resamples the actual trade sequence in blocks to generate thousands of plausible alternative orderings, then asks where the actual sequence sits in that distribution. If your actual result is at the median, you weren't lucky. MRE v06 lands at the 50.9th percentile.
Three-state positioning
Position sizing that maps a single signal to one of three allocations — 100% long, 50% long, or 0% cash. Used by Arsenal BTC. The middle state lets the engine stay partially engaged in choppy regimes where neither trend nor vol fully confirms.