Correlation with Equities
ρ = +0.02
Grid vs. SPY (US Large Cap). Near-zero. The strategy moves independently of stocks — both up and down — providing the statistical prerequisite for diversification.
Sweet Spot Allocation
10–20%
Grid allocation in a forward-looking scenario where grid returns are modestly positive. At this range, Sharpe ratio improves marginally (+0.03) vs. a pure traditional portfolio. Above 30%, diminishing returns dominate.
Efficient Frontier — Traditional vs. Grid-Augmented Portfolios
Portfolio A (Traditional Only: SPY + GLD + AGG) vs. Portfolio B (20% Grid Fixed, 80% Traditional Optimized) vs. Portfolio C (30% Grid Fixed, 70% Traditional Optimized). X-axis: annualized portfolio volatility. Y-axis: annualized portfolio return. Full 2017–2025 history. All grid returns are backtested/simulated.
Reading this chart: Points further to the upper-left are better (higher return, lower risk). The Traditional frontier (teal) sits above the grid-augmented frontiers on the full 9-year history because the grid's geometric CAGR is impaired by the 2018 and 2022 crash years. Individual frontier points are labeled. The forward-looking scenario (dotted lines) shows where the frontiers move when grid expected return is modeled at +8% rather than the historical -16.3% CAGR.
Correlation Heatmap — Grid vs. Traditional Asset Classes
Annual return correlations, 2017–2025 (n=9 annual observations). Teal = positive correlation, dark = near-zero, red = negative correlation. Note: at n=9, confidence intervals on correlations are wide (approximately ±0.40 at 95%).
Key number: Grid vs. SPY = +0.02. This near-zero correlation is the mathematical foundation of the diversification argument. A perfectly uncorrelated asset reduces portfolio variance even if its standalone return is lower than the portfolio average. Caveat: This correlation may not hold in crisis conditions — stress events can temporarily spike correlations across all risk assets.
Year-by-Year Annual Returns — Grid Composite vs. Traditional Assets
Annual total returns 2017–2025. Grid composite = arithmetic mean of valid backtested fold returns across 17 cryptocurrencies per year. SPY and GLD 2017–2023 from public historical records; 2024–2025 from Supabase benchmark database. AGG all years from Supabase benchmark database.
Note: In positive years for grid (2020: +11.1%, 2021: +13.2%, 2024: +10.3%), the strategy generates returns comparable to or exceeding bond allocations and competitive with moderate equity returns. In crash years (2018: -62.9%, 2022: -58.9%), losses are severe and uncorrelated with bond gains.
Allocation Sweep — Sharpe Ratio vs. % Grid Allocation
Sharpe ratio (Rf = 4.5%) of a blended portfolio as grid allocation increases from 0% to 50%. Remaining capital optimized for maximum Sharpe among SPY, GLD, AGG at each level. Two scenarios shown: Historical (uses actual 2017–2025 grid CAGR of -16.3%) vs. Forward-Looking (assumes grid expected return of +8%, improved regime filtering).
Historical scenario: No sweet spot exists — every grid allocation from 0% upward reduces Sharpe. The geometric return penalty from 2018 and 2022 is dominant.
Forward-looking scenario: Sweet spot at 10–20% grid allocation. Sharpe improves from 0.92 to ~0.95. Above 30%, traditional assets dominate again.
Key Performance Metrics — All Assets and Blended Portfolios (2017–2025)
Geometric CAGR, annualized volatility, Sharpe ratio (Rf=4.5%), and maximum annual drawdown. Grid metrics are backtested/simulated.
| Portfolio / Asset | CAGR | Ann. Vol | Sharpe | Max DD | 2022 Return | Status |
|---|---|---|---|---|---|---|
| Grid Composite 17 coins, backtested avg |
-16.3% | 27.2% | -0.76 | -58.9% | -58.9% | Sim Only |
| SPY US Large Cap Equities |
+18.9% | 15.8% | +0.91 | -18.1% | -18.1% | Live ETF |
| GLD Gold ETF |
+16.3% | 20.4% | +0.58 | -3.6% | -0.3% | Live ETF |
| AGG US Aggregate Bonds |
-0.4% | 5.7% | -0.86 | -14.4% | -14.4% | Live ETF |
| 60/40 60% SPY / 40% AGG |
+10.9% | 9.8% | +0.65 | -13.2% | -13.2% | Blended |
| 80/20 Blend 60/40 + 20% Grid replacing proportionally |
+7.5% | 10.6% | +0.28 | -13.6% | -24.3% | Partial Sim |
| 70/30 Blend 60/40 + 30% Grid replacing proportionally |
+5.8% | 11.6% | +0.11 | -14.2% | -31.1% | Partial Sim |
For Advisors — Plain-Language Summary
What the data shows: Crypto grid trading strategies have demonstrated near-zero correlation with US stocks over the 2017–2025 backtest period. This is the statistical requirement for diversification — but it is not sufficient on its own. The strategy also needs to generate positive returns often enough to justify its place in the portfolio.
The honest trade-off: Grid trading is a short-volatility income strategy that earns consistent small returns in sideways markets and suffers large losses in sustained trending (bear or bull) markets. Over the 9-year backtest, two severe bear years (2018, 2022) compounded to produce a negative geometric CAGR. In the 7 remaining years, the strategy averaged modest positive returns.
The forward-looking case: If future grid performance averages +6–10% annually with improved regime filtering (the CoinRoc RXI™ and GSI™ systems are designed for this), a 10–20% grid allocation has the potential to improve a traditional portfolio's risk-adjusted return by approximately 3–5% in Sharpe ratio terms — a meaningful but not dramatic benefit. Advisors should size the grid allocation conservatively (10–15%) given the tail-risk profile.
Mandatory Risk Disclosures
All grid trading return figures in this document are derived from backtested simulation using the AdaptiveGrid v3DynamicMode strategy with retail-binance-us cost tier (0.40% maker / 0.60% taker). No live capital was traded. Backtests reflect hindsight — they cannot replicate execution uncertainty, liquidity constraints, or regime shifts not present in historical data.
Grid trading is a short-volatility strategy. It systematically generates income in low-volatility ranging markets and can suffer outsized losses in trending markets. Crypto markets can enter sustained trending regimes that last 12–24 months, during which the strategy may experience drawdowns of 50–70%.
Correlation estimates with n=9 annual observations carry 95% confidence intervals of approximately ±0.40. The reported correlation of +0.02 (grid vs. SPY) is not statistically distinguishable from correlations as high as +0.42. The diversification benefit assumes this correlation remains near zero in live conditions.
Past simulated performance does not predict future results. This material is for research and educational purposes only and does not constitute investment advice. Consult a qualified financial advisor before making any investment decisions.