Trading strategy case studies: rules, returns and drawdowns

What happened when we coded the idea?

These reports start with a specific published idea, make its rules executable, and inspect the resulting trades. Some changes improve a control. Others reduce losses by surrendering too much return. Each study keeps the less favorable periods and explains the execution choices behind the comparison.

The figures show historical simulations, with separate return and drawdown panels where available. Downloads preserve the reviewed measurements and observations. An author's name identifies the source idea; it does not imply participation or endorsement. Native NexusTrade backtests and independently audited local simulations are labeled separately.

Value portfolios and stock ratings

Follow three portfolio experiments and a stock-screener study through their charts, rule changes and later results.

Creator ideas and their tested adaptations

Create, backtest and deploy trading bots

ET

ETH trading bot: create, backtest and deploy a crypto bot

Compare a 42.43% development return with ETH holding after fees, then inspect two later historical windows and complete exit evidence.

Open page →

Options trading strategies: create, backtest and deploy a bot

Compare six SPY options bot variants, a 52.68% modeled development result, 94 audited fills and a no-trade holdout caused by the fixed premium budget.

Open page →

Leveraged ETF trading bot: create, backtest and deploy TQQQ rules

Build a TQQQ bot from a QQQ trend signal and test the frozen rules across 15 years. Inspect the 3,182.36% strategy return against TQQQ's 27,375.08%, the 51.68% versus 81.97% drawdowns and the losing reserved test.

Open page →

Walk-forward optimization: a TQQQ trend bot that survived its own out-of-sample test

The full optimization procedure published as evidence: three declared rounds, the losing-trade audits that drove each repair, validation on untouched dates, and a five-fold walk-forward that chained to +2,895% out-of-sample - within reach of TQQQ buy-and-hold at a fraction of its 82% drawdown.

Open page →

Intraday selection and a longer test

NV

NVDA intraday versus buy and hold

A limited lead disappeared when the same quarterly selection procedure was extended. Inspect walk-forward folds, cost reruns and the unresolved replay difference.

Open page →

Use the evidence to define your next test

Start with the rules, cost convention and same-date baseline. Read both the broad and recent results before copying a portfolio. A historical improvement is a reason for another declared test, not a forecast of future returns.