APPLIED QUANTS · EPISODE 2 · FILLED-IN FICTIONAL EXAMPLE
Version 2026.11.04

This invented Buy Bot-style exercise illustrates precise rules. It does not reproduce the owner's private Buy Bot, represent its development history, or recommend a trading strategy. Numbers and prices below are fictional.

MARKET AND DATA
One imaginary stock, DEMO; one-minute completed bars during a teaching session from 09:30 to 16:00 America/New_York. Use the first bar's opening price as the session reference. No after-hours entries.

ENTRY SIGNAL
Threshold = session opening price times 1.01. While flat, signal once when the previous completed close was at or below the threshold and the current completed close is above it. Both closes must be from this session. A first bar alone cannot establish a crossing.

EXECUTION AND SIZE
Buy one share at the next bar's opening price, never at the already-observed signal bar's open. Start with $1,000 cash. Permit only one position and one entry per session. Skip if cash cannot cover the next open plus the stated fee. No leverage, shorting or real orders.

EXIT
After entry, signal an exit when a completed close is below the session opening price; sell at the next bar's open. Also schedule a close at the 15:59 bar's open (a fixed time known in advance). Do not create an entry that would execute at or after 15:59. If required data is missing, mark the run incomplete and report any open position; never invent a fill.

PRACTICE COST MODEL
Charge $0.01 per filled order. The deliberately simplified first fixture has zero spread/slippage and enough liquidity for one share. These assumptions are not market realism; later lessons must test explicit costs and alternative fills.

KNOWN-ANSWER CHECKS
Reference open $100; threshold $101.
Previous close $100.90, current close $101.10: entry signal. If next open is $101.20, fill there; cash becomes $898.79 after the fee.
Previous close $101.10, current close $101.20: no new crossing.
Previous close $100.90, current close exactly $101.00: no entry (the rule says above).
Already in a position: ignore another entry signal.
Missing next bar: no invented fill; report incomplete data.

WHAT THIS LEAVES TO TEST
Whether a crossing rule is useful at all; realistic data quality, costs, fills, session calendars, and performance on unseen periods. A rule being precise does not make it profitable.

Educational content only, not investment advice. Backtest results are hypothetical and don't predict future results. Trading involves risk of loss.
Applied Quants retains ownership. Reuse terms are being finalized.