What correlation means
Reading the numbers
Correlation measures how two things move relative to each other, on a scale from -1.0 to +1.0.
If correlation is +1.0, two strategies always move together. When one gains, the other gains. When one falls, the other falls. This is what most traditional portfolios look like — all boats rising and sinking on the same tide.
If correlation is 0.0, the two strategies are independent. One has no knowledge of what the other is doing. This is the foundation of real diversification.
If correlation is negative, one strategy tends to go up when the other goes down. This is active hedging — the portfolio is designed to profit from its own drawdowns.
11 engines, 55 unique pairs
Moderate — both use VIX regime
Highest pair — both equity/gold rotation
Near zero — independent signals
Different asset class entirely
Negative — hedges equity exposure
Near zero — event-driven, uncorrelated
Factory leg — independent
Different asset class
Near zero
Measured at -0.03 — truly independent
Some overlap — both hold equities
Low — sector rotation vs mega-cap
Different asset class
Negative — Shield hedges equity concentration
Low — different signals
Low despite both touching tech
Near zero
Independent
Near zero
Low — ML vs macro signals
Near zero
Negative — hedging relationship
Low
Independent
Near zero
Low
Near zero
Near zero — completely different markets
Low — both macro but different signals
Low
Near zero
Low
Near zero
Independent
Near zero — different universe
Near zero
Independent
Zero
Near zero
Zero
Near zero
Slightly negative
Low — some commodity overlap
Near zero
Independent
Low
Near zero
Low
Independent
Low despite both being momentum
Low
Independent
Near zero
Independent
Independent
What the matrix tells you
When one engine struggles, the others do not know about it. This is not a metaphor. It is measured, pairwise, across 55 unique combinations. Average correlation of 0.04 means the engines are, for practical purposes, independent systems operating in parallel.
Pairs worth examining
| Engine A | Engine B | Correlation | Rating | Why |
|---|---|---|---|---|
| Sentinel | Meridian | 0.565 | Highest | Both rotate between equities and gold. Expected overlap. This is the only pair above 0.55. |
| Harvest | Meridian | 0.509 | Moderate | Both hold equities. Harvest is concentrated mega-cap; Meridian is ML-optimized. Different signals, some asset overlap. |
| Sentinel | Harvest | 0.337 | Moderate | Both use VIX regime signals. Different instruments and different timing logic. |
| Harvest | Shield | -0.22 | Hedging | Shield gains when equity-heavy Harvest suffers. This is by design. The insurance works. |
| Sentinel | Shield | -0.18 | Hedging | Same hedging dynamic. Shield offsets equity-linked drawdowns. |
| Carry | Everything | ~0.02 | Independent | Crypto funding rates are a completely different market. Zero relationship to equities, bonds, or commodities. |
| Lens | Everything | ~-0.01 | Independent | CEO psychology has no relationship to price momentum, volatility, or macroeconomic signals. |
| Factory legs | Core engines | ~0.03 | Independent | The factory pipeline discovers strategies precisely because they are uncorrelated to existing legs. |
Most investors own one risk: the equity market going up. When it doesn't, everything in their portfolio falls together. This matrix shows a different architecture — eleven independent engines, each watching different signals in different markets, producing returns that have nothing to do with each other. That independence is not a side effect. It is the entire point.