Home / Strategies / Curve
Crypto · strategy report

Does anything beat buy & hold on Curve?

Every setup we tested on Curve — ranked out-of-sample, corrected for multiple testing, and forward-tracked in public from the day this page published. The honest answer is the headline.

MIXED

Beat buy-and-hold in both windows — but can't be told apart from selection luck.

Beat buy-and-hold in both the full window and out-of-sample but its OOS Sharpe 1.35 did not clear the 2.22 selection hurdle (best-of-N luck cannot be ruled out). Buy-and-hold benchmark: -38.2% CAGR over 8.5 years (-15.6% CAGR in the out-of-sample window).

Educational research from historical backtests — not investment advice. Past performance does not predict future results.

Curve (CRV): Beat Buy-and-Hold in Both Windows, Couldn't Beat the Selection Hurdle

Crypto sets a brutal baseline: Curve compounds at -38.2% annualized just for holding, and it charges drawdowns near -71.8% for the privilege. Against that backdrop we ran 598 indicator configurations on CRV. The best of them — Acceleration Bands on the daily timeframe — beat buy-and-hold in both the training and holdout windows, with a profitable out-of-sample trade profile across 76 trades. That is genuinely uncommon here. But 'best of 598' is exactly the phrase that should make you slow down, which is what the second paragraph is for.

When you pick the winner from 598 attempts, the result is partly signal and partly luck of the draw. Our hurdle corrects for that: with this many tries and 2.6 years of holdout data, an out-of-sample Sharpe needs to clear 2.22 before we trust it. This one landed at 1.35 — profitable, but statistically indistinguishable from the luckiest of hundreds of tries. Add that crypto structure keeps changing — new venues, new participants, shifting volatility regimes — and a pattern from the past has no obligation to repeat. Treat this as a research lead, not an edge.

Every figure above is computed from our own backtests — nothing is estimated or invented. Hypothetical results; not investment advice.

The rules

Top setups as mechanical rules

Exactly as the backtest defined them — no discretionary steps, no hidden filters.

#1 · Volatility · Daily

Acceleration Bands

Mechanical rule (exactly as backtested): Price Headley's volatility breakout bands — long on a close above the upper band. Signals are evaluated at daily-bar close, the position changes on the NEXT bar, 0.08% cost per side, long/flat only — no leverage, no shorting.

+167.8%
Total return
0.47
Sharpe
-71.8%
Max DD
39.5%
Win rate
76
Trades
+50.5%
vs B&H

Out-of-sample (last ~30% of the window, never used to pick this setup): Sharpe 1.35 · alpha +85.5% · 19 trades over 2.6 yrs.

#2 · ML · Daily

Lorentzian Classification

Mechanical rule (exactly as backtested): k-nearest-neighbor classifier over oscillator features (Lorentzian distance) — long when the most-similar past bars led higher. Signals are evaluated at daily-bar close, the position changes on the NEXT bar, 0.08% cost per side, long/flat only — no leverage, no shorting.

-52.5%
Total return
0.17
Sharpe
-95.4%
Max DD
47.4%
Win rate
371
Trades
+29.8%
vs B&H

Out-of-sample (last ~30% of the window, never used to pick this setup): Sharpe 1.3 · alpha +93.4% · 135 trades over 2.6 yrs.

#3 · Trend · Daily

Vertical Horizontal Filter

Mechanical rule (exactly as backtested): Trade trend only when the VHF says the market is trending (and price is up). Signals are evaluated at daily-bar close, the position changes on the NEXT bar, 0.08% cost per side, long/flat only — no leverage, no shorting.

+387.6%
Total return
0.61
Sharpe
-71.2%
Max DD
52.7%
Win rate
55
Trades
+58.7%
vs B&H

Out-of-sample (last ~30% of the window, never used to pick this setup): Sharpe 1.3 · alpha +80.6% · 13 trades over 2.6 yrs.

Forward test

Since publication — including if it loses

0.0%
the published setup, since 2026-07-02 (0 market days)
0.0%
buy & hold, same window

The forward record is just getting started — the gap between the two is the honest score. Marked to market nightly from real prices, rules frozen at publication, as of 2026-07-02. Currently FLAT.

How this verdict was computed (mode: out-of-sample)

We tested 598 setups (indicator × parameters × timeframe) on Curve. Only setups with ≥30 trades qualify (332 did). Setups are ranked by out-of-sample Sharpe — the last ~30% of history, which standard-parameter rules never saw during selection. Because picking the best of 598 tries mines even the holdout, the VALIDATED verdict additionally requires the top setup’s OOS Sharpe to clear a selection hurdle of 2.22 (√(2 ln N)/√T) AND positive alpha in both windows. Of the eligible setups, 81.0% had positive out-of-sample alpha (median OOS Sharpe 0.39) — the table below is truncated, but this summary covers all of them. Full recipe: methodology · the engine’s contract lives in the repo as STRATEGY_METHODOLOGY.md.

Ranked table

Top 20 of 332 eligible setups

Ranked by out-of-sample Sharpe. Full + out-of-sample columns, costs included. Hypothetical.

#SetupTFTotal retSharpeMax DDWinTradesα vs B&HOOS SharpeOOS αOOS trades
1Acceleration BandsDaily+167.8%0.47-71.8%39.5%76+50.5%1.35+85.5%19
2Lorentzian ClassificationDaily-52.5%0.17-95.4%47.4%371+29.8%1.3+93.4%135
3Vertical Horizontal FilterDaily+387.6%0.61-71.2%52.7%55+58.7%1.3+80.6%13
4Ulcer IndexDaily+339.4%0.58-80.7%33.3%54+57.3%1.23+83.4%14
5Chaikin VolatilityDaily+493.1%0.64-81.3%46.7%92+61.5%1.18+80.8%26
6Relative Volatility IndexDaily+284.3%0.56-68.8%41.0%134+55.4%1.12+80.3%34
7Bollinger BreakoutDaily+569.8%0.66-65.0%39.4%33+63.3%1.06+69.8%10
8Keltner 20 BreakDaily+78.6%0.35-69.8%45.8%48+45.3%1.06+57.8%11
9Bullish EngulfingDaily-33.8%0.27-96.5%47.3%110+33.5%1.04+72.8%37
10Hull MA 100 TrendDaily+26.9%0.37-90.8%32.7%55+41.1%1.03+70.7%11
11Geometric MADaily+221.5%0.54-82.7%37.4%99+53.0%0.97+64.9%27
12Median MADaily+292.3%0.56-84.2%41.4%116+55.7%0.96+61.2%30
13Ehlers Roofing FilterDaily-81.8%0.24-95.6%35.1%37+20.1%0.95+69.4%10
14Elder Ray (Bull/Bear Power)Daily-20.8%0.3-95.1%34.3%137+35.5%0.93+61.4%36
15Even Better SinewaveDaily+92.7%0.48-91.4%34.1%44+46.3%0.92+63.2%13
16KAMA 30 TrendDaily-23.9%0.3-91.1%30.1%83+35.1%0.91+59.4%22
17Volume-Weighted EMADaily-9.2%0.33-94.6%39.1%138+37.1%0.89+59.7%37
18TRIMA 30 TrendDaily+56.4%0.41-85.9%40.4%57+43.6%0.89+57.7%13
19HLC TrendDaily-0.1%0.35-91.4%36.9%122+38.2%0.88+58.7%32
20WMA 10/40 CrossDaily+192.9%0.54-82.9%33.3%36+51.7%0.87+58.1%10

Hypothetical backtests with 0.08%/side costs. Not investment advice — see the full disclaimer.

Read this before acting on anything

These are historical backtests of mechanical rules. They are educational research, not investment advice, not a recommendation, and not tailored to you. Educational information only — not investment advice. Hypothetical backtested results; past performance does not guarantee future results. Trading involves risk of loss.

Keep digging