This is an independent implementation inspired by publicly described portfolio concepts. It is not affiliated with, endorsed by, or an exact reproduction of 42 Macro. The official Global Macro Risk Matrix and VAMS formulas are proprietary. This implementation uses open-source proxy signals where official signals are unavailable.
Loading…
Signal engine
Global equities
Gold
Bitcoin
Cash / T-bills
Cash return
Advanced parameters
Backtest start
Backtest end (blank = today)
V1 bullish threshold
V1 bearish threshold
V1 confirmation days
Volatility lookback
Momentum weights 21/63/126/252
V2 equity thresholds bull/bear / confirm
V2 gold thresholds bull/bear / confirm
V2 bitcoin thresholds bull/bear / confirm
Costs bps eq/gold/btc/cash
Staged entry
Rebalancing method
Execution delay
Macro release delays (days) INDPRO / CPI
Benchmarks
Current portfolio
External signal entry (append-only log — previous signals are never overwritten)
Effective date
Macro regime
Global-equity momentum
Gold momentum
Bitcoin momentum
Notes
Drag = pan · shift+drag = box-zoom · scroll = zoom X · drag Y-axis = zoom Y · double-click = reset
Regime model diagnostics (V2 top-down)
Daily relative regime weights over the backtest window. These softmax weights measure relative strength, not statistically calibrated probabilities. Target roughly two to three major risk-on/risk-off transitions per year — diagnostic only, not an objective.
Development 2006–2013 · Calibration 2014–2017 · 2018+ evaluation. Parameters were not numerically optimized against the displayed 2018+ test period. However, the methodology was designed with knowledge of modern market history, so these results are not a true prospective out-of-sample record. A prospective record begins when the model parameters are frozen and published.
Full revised strategy vs existing methodology, static 60/30/10, 60/40, buy & hold, simple 10-month SMA, macro targets without momentum, and momentum overlay without macro targets.
The official KISS regime and momentum signals are proprietary. Therefore, an independent implementation cannot be expected to reproduce the official portfolio's trades or performance. Differences may arise from signal construction, historical data, Bitcoin proxies, execution timing, transaction costs and portfolio-accounting conventions. The objective of this model is robust out-of-sample risk management, not matching a published backtest.
Robustness by subperiod
These are descriptive subperiod slices using identical data and accounting. The first row is partial when the selected backtest begins after 2017. They are not independent prospective tests and should not be interpreted as parameter validation.
Attribution (contribution to cumulative return)
Time in each macro regime
Signal history
Data sources & historical proxies
Methodology & disclosure
Strategy description
The Macro Trend 60/30/10 portfolio is a long-only, event-driven strategy combining global equities, gold, Bitcoin and Treasury bills. A macro regime first determines each asset's maximum target allocation. An independent volatility-adjusted momentum signal then determines whether the portfolio holds 100 percent, 50 percent or 0 percent of that target. All unused exposure is allocated to Treasury bills, which earn the historically available three-month Treasury-bill yield.
Two signal engines are available: Version 1 — Economic Regime plus Simple VAM. A publication-date-aware economic regime (industrial production and CPI) sets the targets; a simple volatility-adjusted momentum score sets the multipliers. Version 2 — Cross-Asset Regime plus Enhanced VAM (default). The primary allocation regime is generated from a daily cross-asset model that measures market-implied growth, inflation and defensive-risk conditions. A separate publication-date-aware economic regime based on industrial production and CPI is displayed as supporting context. The daily cross-asset model is an independent approximation of a top-down market-regime process and is not 42 Macro's proprietary Global Macro Risk Matrix.
The asset momentum overlay combines medium-term price momentum, trend agreement, volatility momentum and drawdown conditions. It is an independent volatility-adjusted momentum model and is not the official 42 Macro VAMS formula.
During the growth-friendly Goldilocks and Reflation regimes, the portfolio can hold as much as 60 percent global equities, 30 percent gold and 10 percent Bitcoin. During Inflation and Deflation regimes, risk-asset targets are reduced and the cash allocation increases. Bearish asset-level momentum can reduce risk exposure further.
Publicly described portfolio rules (used as specified)
Long-only portfolio; maximum 60% global equities, 30% gold, 10% Bitcoin
Four macro regimes: GOLDILOCKS, REFLATION, INFLATION, DEFLATION
Reduced targets during risk-off regimes (Inflation/Deflation)
Residual weight goes to Treasury bills; portfolio always totals 100%
Independent open-source approximations (NOT official 42 Macro rules)
Version 4 disclosure: V4 is an independent trend-aligned momentum extension with volatility scaling. It is not official KISS, VAMS or Global Macro Risk Matrix methodology.
Version 2 regime: 40 daily cross-asset features in three groups — growth/risk-appetite (16), inflation (12), defensive/liquidity (12). Feature momentum is computed over 21/63/126/252 trading days, volatility-adjusted and normalised with a rolling three-year z-score (no future observations). Growth, inflation and defensive-risk scores are equal-weighted group averages; regime probabilities come from a softmax over four regime scores.
Version 1 regime: industrial production (INDPRO) as the growth proxy; CPI (CPIAUCSL) as the inflation proxy; three-month changes in year-over-year macro rates define the regime. These are revised-data simulations with publication-delay assumptions, not ALFRED vintage backtests; revised historical observations may differ from information available in real time.
Version 2 momentum: composite = 0.55 × normalised price momentum + 0.35 × trend agreement − 0.20 × volatility momentum − 0.25 × drawdown penalty, with asset-specific thresholds (equities ±0.20, gold ±0.15, Bitcoin ±0.30), signal hysteresis and emergency risk-reduction rules.
Version 1 momentum: single composite volatility-adjusted return over 21/63/126/252 days with ±0.25 thresholds.
Confirmation periods and historical proxy instruments
Open-source proxy methodology. These are not official 42 Macro signals.
These parameters are an independent approximation and are not official 42 Macro parameters.
Look-ahead, taxes and overfitting warnings
Backtests using revised economic data without publication delays can contain look-ahead bias. Real-time vintage economic data should be used where available.
Taxes are not included. A taxable investor may experience materially different results.
Selecting parameters based on the period being evaluated can create overfitting and produce unrealistically strong historical results.
Timing, staging and execution conventions
Signals are computed after market close from that day's close; execution happens at the next trading day's open (never at the same close). Macro data only affects the portfolio after its publication date (default: INDPRO +17 calendar days, CPI +15 calendar days after the observation month-end); the adjustment executes on the next trading day. Bitcoin is evaluated on business days using the most recent business-day observation; no weekend trades; Bitcoin ETF changes execute in the next ETF session. Momentum changes require the composite score to remain in the new range for the configured confirmation period (default 3 trading days). Staged entry (optional, off by default) buys 50% of an increase at the first execution and 50% five trading days later, cancelling the remainder if the signal reverses first; reductions execute in full. Staged entry is an execution option and not part of the open-source signal formula.
Backtest example (Reflation, eq BULLISH / gold NEUTRAL / btc BEARISH)
Targets 0.60 / 0.30 / 0.10 → weights 0.60×1.00 = 0.60, 0.30×0.50 = 0.15, 0.10×0.00 = 0.00, cash = 1 − 0.60 − 0.15 − 0.00 = 0.25.
Automated tests