Replay 223 tickers and 17 built-in strategies day by day over historical closes. Five families: split buying (A), dual mode (D) that switches between attack and defence on the weekly RSI, overnight breakout (B) that buys a move above the open and sells at the next open, two GDXU-only strategies (G), and Laoer's infinite buying V4.0 (I). You see not just the return but how deep the account fell, and how many years ended negative when each year is restarted from scratch.
| Preset | Weights | Buy | Target | Timecut |
|---|---|---|---|---|
| Standard | 10·15·20·25·20·10 | −0.01% | 1.5% | 10d |
| Defensive | 5·10·15·20·25·25 | −0.01% | 0.01% | 10d |
| Aggressive | 16.7 × 6 | −0.1% | 2.0% | 12d |
| 연도 | 선택 전략 | 단순 보유 | ||
|---|---|---|---|---|
| 수익률 | MDD | 수익률 | MDD | |
Pick a ticker, set the period and the seed capital, then press Run Backtest. The tool replays the split-buying rules day by day over the closing prices and reports what the account would have looked like.
Results show total return, maximum drawdown, annualised return (CAGR), win rate and average holding days, with the equity curve drawn against buy and hold.
Maximum drawdown here is measured on the whole account value including cash. Split buying leaves most of the seed uninvested most of the time, and idle cash does not fall. That is the main reason the figure looks better. Read it as low exposure rather than as protection — check the exposure number alongside it.
Neither. Every day the tool adds cash plus the market value of open tranches, builds an equity curve from that, and measures how far the curve fell from its running peak.
Whole-share buying. If one tranche is allocated $1,000 and the share costs $1,200, it buys nothing. Raise the seed or lower the number of tranches.
Only for a single day. Over longer periods volatility drag pulls the result below 3x in choppy markets and above 3x in a sustained trend. The arithmetic is worked through in the guide on volatility drag.
Yes. SOXL and TQQQ listed in 2010 and SPYU in 2023, so the years before listing are index-based: the daily return of an older 1x index or ETF tracking the same market (for SOXL the PHLX Semiconductor Index), multiplied by the leverage, minus the fund fee, the borrowing cost and a tracking cost measured on the years where real and index-based prices overlap. The screen tells you when a result uses this range and marks those years in the yearly tables. SCHD has no free 1x source, so it starts at its 2011 listing. How the series is built and how closely it matches the real ETFs is in the guide Backtesting leveraged ETFs before they existed.
It closes tranches that have been sitting at a loss. Without it, a falling market leaves every tranche open and the account has no cash left to buy the decline. It is the rule that keeps the cycle turning.
Prices are split and dividend adjusted, so dividends are reflected in the price series. Taxes are not. Figures on this page are pre-tax, and a high-turnover strategy loses more to tax than a buy-and-hold one.
Move each parameter 20% either way and run it again. A strategy worth using sits on a broad plateau; if only one exact setting works, it was fitted to the past. There is a full checklist in the guides.