Data statis (sesuai desain)— Bootstrap atas data historis tetap — bukan real-time, sesuai desain.
Block Bootstrap — Sharpe Confidence Interval
Crypto bisa jadi investasi aman + passive income — kalau kamu pakai platform legitimate
(Bappebti-registered exchange) + strategi 9-tahun validated, bukan "bot Telegram 1% per hari".
Realistic return: 10-35% annual + 4-7% DeFi yield untuk modal nganggur.
UPDATE (Meta Optimizer findings — 2026-05-12, v2 koreksi):
Setelah exhaustive grid-search 210 kombinasi, ditemukan trade-off:
- 75R/25H = absolute return tertinggi (ROI 2.5× vs baseline 30/70) — pilih kalau prioritas compounding
- 30R/70H baseline = Sharpe lebih tinggi (1.29 vs 1.16) — pilih kalau prioritas risk-adjusted + tidur nyenyak
- DD: 75R/25H punya max drawdown lebih dalam (~62% vs 37%) — bear market lebih brutal
Method: Sample blok consecutive returns (size mo) untuk preserve autocorrelation.
simulations per strategi. 95% CI = (P5, P95).
Real edge: P5 > 0 = robust positive Sharpe across resampling.
Sharpe Confidence Intervals
Sharpe distribution histogram
Why Bootstrap? Returns ≠ Normal Distribution
Standard parametric metrics (Sharpe ratio, t-test, VaR-normal) assume returns are
normally distributed — symmetric bell curve, thin tails. Crypto melanggar asumsi ini secara brutal:
daily log-returns BTC menunjukkan fat tails (extreme moves jauh lebih sering dari prediksi normal),
negative skew (crash lebih tajam dari rally), dan volatility clustering.
Inilah kenapa kita pakai block bootstrap — non-parametric, tidak butuh asumsi distribusi.
Bukti empiris ada di 2 chart di bawah.
Histogram: BTC daily returns vs Normal overlay
Bar = empirical BTC log-returns (~9y). Garis merah = normal distribution (μ=0.001, σ=0.038) di-scale sama N. Lihat ekor kiri & kanan: empirical jauh lebih tinggi dari normal = fat tails.
Q-Q Plot: Sample vs Theoretical Normal
Titik biru = sample quantiles. Garis abu-abu putus = y=x (perfect normal). Deviasi di tails (>|2σ|) = fat-tail evidence. Bootstrap tidak butuh asumsi ini.
Mau Lebih?
Premium subscriber dapat akses Live Trading Bot + Telegram Signal + Command Center.
Referensi Ilmiah:
- Primary: Politis, D.N., & Romano, J.P. (1994). "The Stationary Bootstrap." Journal of the American Statistical Association, 89(428), 1303-1313. DOI: 10.2307/2290993
- Künsch, H.R. (1989). "The Jackknife and the Bootstrap for General Stationary Observations." Annals of Statistics, 17(3), 1217-1241. — Moving Block Bootstrap origin.
- Aronson, D. (2006). Evidence-Based Technical Analysis. Wiley. — MC permutation tests untuk significance.
- Cont, R. (2001). "Empirical properties of asset returns: stylized facts and statistical issues." Quantitative Finance, 1(2), 223-236. — Fat-tail evidence di financial returns.
- Bouri, E., Molnár, P., Azzi, G., Roubaud, D., & Hagfors, L.I. (2017). "On the hedge and safe haven properties of Bitcoin." Finance Research Letters, 20, 192-198. — Crypto fat-tail & non-normality.
Past performance does not guarantee future results. Crypto is high-risk; size positions accordingly. Industry sources (Glassnode, CryptoQuant) flagged sebagai methodology praktisi, bukan peer-reviewed.
Gratis · Cukup Login
Masuk dulu buat buka tool ini
Semua tool Pro & Premium gratis — kamu cukup login pakai WhatsApp. Tanpa biaya, tanpa kartu.
Gratis selamanya • Cuma butuh nomor WhatsApp