Run the strategy. Inspect the result. Keep the code.

Buy on SMA(20) crossing above SMA(50), sell on the reverse cross.

Buy oversold liquid ETFs when RSI(14) is below 35, sell when RSI is above 60.

When RSI is weak, place a passive buy limit below the close.

In an uptrend, wait for a 1% pullback before entering.

Place buy-stop orders above the recent 20-day high.

Enter on SMA(20/50) trend and attach a 5% protective stop.

Enter breakouts, but refuse to pay far above the stop trigger.

Enter an SMA trend, then protect the downside with a stop-limit sell.

Enter on SMA trend and let a 4% trailing stop manage the exit.

Buy oversold RSI readings, then let a 5% trailing stop handle risk.

Trend entry with a trailing stop that converts to a limit order.

Enter on SMA trend and attach take-profit plus stop-loss exits.

Buy oversold RSI dips with a predefined reward/risk bracket.

Submit two competing entry orders: buy a dip or buy a breakout.

On 5-minute bars, buy oversold RSI(14) dips and wrap each entry in a tight intraday bracket (+0.6% take-profit / -0.4% stop-loss).

On 30-minute bars, place buy-stop orders above the recent 12-bar high so entries only trigger on confirmed intraday breakouts. Positions are held for a fixed number of bars and then exited at market.

An event-driven monthly rotation executed with explicit orders. On the first bar of each new month, rank the universe by 6-month momentum, then sell names that dropped out of the top set and buy the new entrants at market.

An event-driven strategy with no calendar at all. Each name has an SMA trend signal; the strategy trades only when a signal flips — buy on a flat->long flip, sell on a long->flat flip. Between flips it does nothing.

Trade four USD-quoted spot-FX pairs in the direction of six-month momentum. Position size is an integer number of standard lots, constrained by ATR risk and a per-pair notional cap.

A diversified 55-day Donchian breakout across index, energy, and metal futures. Orders are whole contracts sized by ATR risk, gross notional, and the reference initial margin carried in ContractSpec.

Hold each sector ETF only when SMA(50) is above SMA(200).

SMA(50/200) trend filter on a diversified ETF universe.

Enter when RSI(14) is below 35, stay long until RSI rises above 60.

Rank ETFs by 12-month return, hold the top two only if return is positive.

Allocate more to ETFs with lower recent volatility.

If SPY is above its SMA(200), hold risk ETFs; otherwise hold defensive ETFs.

Use yfinance-backed 15-minute bars from /bt/prepare and hold an equal weight basket when RSI(14) is oversold.

Hold each name only while its fast EMA(9) is above its slow EMA(21) on 1-minute bars. A fast, high-turnover trend-follow / scalp template.

On hourly bars, hold each name only while its SMA(10) is above its SMA(30), equal weight across active names. A slower intraday-to-swing trend template.

SMA(50/200) trend on a diversified ETF basket, rebalanced monthly, with a circuit breaker that flattens the book on a large drawdown and waits out a cooldown before re-engaging.

Quarterly momentum rotation on a broad ETF universe, but layered with a tracking-error trigger (rebalance early if the book drifts too far from the benchmark) and an annual-turnover budget (a cost gate that suppresses trading once the rolling turnover budget is spent).

A daily-updated momentum tilt across large-caps, but instead of fully rebalancing every day, only trade the positions that have drifted outside a 10% band — a partial rebalance that keeps turnover low.

Trade the mean-reverting spread between two closely related names using a log-ratio z-score. When the spread stretches, short the rich leg and long the cheap leg; unwind when it reverts.

Same mean-reversion premise as the ratio pair, but the spread is built from a rolling OLS hedge ratio (beta of A on B) instead of a 1:1 log ratio, so the pair stays balanced as the relationship drifts.

Each month, rank the universe by 12-month price momentum, go long the top names and short the bottom names in equal dollar amounts. The classic cross-sectional momentum factor.

The mirror image of momentum. Over short horizons, recent losers tend to bounce and recent winners tend to give back. Each week, rank the universe by 1-month return, LONG the biggest losers and SHORT the biggest winners.

A simple SMA(50/200) trend basket, but instead of holding fixed weights, scale total exposure so the portfolio targets ~10% annualized volatility. In calm markets the strategy leans in; in turbulent markets it de-risks.

Hold a momentum-selected basket (top names by 6-month return) and scale it to a 15% annualized volatility target, allowing modest leverage. A different base strategy and a higher target than the trend example, to show how the same overlay adapts.

Hold a diversified multi-asset basket, but size positions so each asset contributes equal risk (equal risk contribution, ERC) rather than equal dollars. Low-volatility assets (bonds) get more capital; high-volatility assets (equities, commodities) get less.

Equal-risk-contribution across US sector ETFs, then a hard per-position cap so no single sector dominates. This shows how to CHAIN risk models: risk parity first, position limit second.

Buy a name when its price closes below the lower Bollinger Band (a stretched-cheap signal) and hold until it reverts back above the moving-average midline. A classic band-based mean-reversion template.

Hold each name while its MACD line is above its signal line — a momentum trend filter that reacts faster than a simple long-window SMA crossover.

Trade each USD-quoted spot-FX pair in the direction of its six-month momentum and scale active signals by inverse 63-day volatility.

Rank four USD-quoted spot pairs by three-month return, go long the strongest base currency and short the weakest. Gross exposure is one and net exposure is zero before the engine's cash buffer.

Diversified long/short trend following over equity index, rates, energy, metals, and grains, with inverse-volatility weights.

Run a normal SMA(50/200) trend strategy, then inspect its temporal robustness with a ROLLING walk-forward and an explicit pre-OOS purge gap.

Same SMA(50/200) trend strategy as WF 01, but validated with an EXPANDING walk-forward — the training window grows over time (anchored start, moving end) while the test window slides forward. This mimics a strategy that keeps all history as it accumulates.

Inspect a weekly RSI mean-reversion strategy with an ANCHORED walk-forward, using a fixed purge plus a percentage extension before each OOS window.

Find the best SMA fast/slow window pair for a trend strategy with an exhaustive GRID SEARCH. Each candidate is scored by running a real backtest and reading its annualized Sharpe.

Use Optuna's Tree-structured Parzen Estimator (TPE) to search SMA fast/slow windows over continuous integer ranges, then validate the winning parameters with a rolling walk-forward.