A MarketXED playbook turns a committee direction into numbers you can execute: limit entry, stop, target, share count, and an exit clock. It is built for small retail tickets (think a fixed dollar risk, Robinhood-style limit orders, and an end-of-day backstop). It is a checklist, not a promise, and not financial advice.
How the prices are built
Stop distance uses a floor of about 1.5 times ATR, never tighter than about 1.5 percent of price (and a few cents on very cheap names). A one-ATR stop often dies inside a single bar’s noise. Target distance is set so that after a simple friction/slippage haircut, reward still clears a configured R multiple (about 1.2R by default). Shares come from that risk-per-share versus your risk dollars, with a half-Kelly-style cap so a hot probability cannot explode size.
- Limit slightly through the print so a marketable-limit fill is realistic.
- Day time-in-force, not a forgotten GTC.
- Stop first after a fill; hope is not a working order.
When the regular session is already open, the default validity window is on the order of 90 minutes, and never past a few minutes before the cash close. If the idea is staged while the market is shut, the morning window can be longer so an opening gap has room. A green timer exit is still a timer exit, not a target hit. That distinction matters for the calibration labels.
What the playbook will not invent
It will not assume you can short on a cash app that cannot short. It will not raise risk because sentiment is loud. It will not ignore PDT or unsettled cash. If the committee is a learning-only print, there is no live ticket to copy. If you change risk dollars in the UI, the share math should change with it; if it does not, do not “fix” it by typing a share count from memory.
How to use it with the vote tape
Read the committee split first. Fractured agreement plus a full-size playbook is a mismatch. Tight agreement plus a name that fails your universe filter is also a mismatch. The playbook is the last translation step, not the thesis.
Set a phone alarm for the exit time the card prints. If neither stop nor target prints, flatten. That habit is the whole difference between a labeled experiment and a story you tell at the close. Overnight holds are a different product; this playbook is session-scoped unless you explicitly change hold rules (and then you must relabel).
Education, not automation
The steps on the card (search the symbol, limit buy, attach stop, alarm) exist because most losses in this style of tape are process losses: market orders on a spike, no stop, or a hold past the clock. Follow the card or pass. Do not mix the card with a second undocumented plan. MarketXED will not place the order for you.
Friction, R, and why tiny accounts feel different
Round-trip slippage is larger as a percent on cheap names, so the playbook adds friction into risk-per-share and still demands the target clear an R floor after that haircut. That is why a “1.2R” card is not a daydream measured from mid. If your account cannot buy a whole share without blowing the risk cap, the honest output is one share or a pass, not a fractional fantasy the venue cannot honor.
Rehearse the attach-stop step once on a quiet name you do not care about (or in paper). Most process losses are: market-in, no stop, no alarm. The card already writes the alarm. If you skip it, you left the experiment. Relabeling that as “discretion” just poisons calibration.
Worked example: $25 risk on a $90 name
ATR stop floors at about 1.5 percent, so risk per share is roughly $1.35 plus a friction haircut. $25 risk is about 18 shares before the half-Kelly cap, then the card prints a day limit and a 90-minute clock. If you only have $200 settled, you may still be fine on notional, but a cash account cannot recycle the exit into a second name today. The card does not see that. You do.
FAQ
Does the playbook place the order?
No. It prints prices, shares, and a clock. You submit at the broker. MarketXED will not swipe for you.
Why is a green exit not always a win?
A timer flatten that is slightly green is a time-stop label. A target tag is a different event. Mixing them breaks calibration.
Can I hold past the clock if I still like it?
You can, but that is a new experiment. Relabel. The old probability belonged to the 90-minute rule.
Print the card or screenshot it before you swipe. After the exit, mark which line you broke if you broke one. Most people break the alarm. A few break the stop. Almost nobody admits they broke both. The playbook cannot save a process you will not audit. One clean card a day beats three edited memories.