Putting it to work

The method suits a routine that takes minutes a day: end-of-day data in, a short list of signals out, orders placed before the next open. Nothing needs watching during the session.

An evening routine

Signals are published after the US close, so a routine fits the evening. A typical one has four parts:

1. The signals arrive. The day's list is one call to the API (see the quickstart). Each signal has a stock, a type, an entry and a stop.
2. What's new stands out. Continuation signals are re-issued daily until they fill or end, so the list repeats stocks. The moveCount field in the signals guide says which day of a set-up it is.
3. Each position gets a size. The arithmetic from page 6 turns the stop distance into a share count.
4. Orders are in before the open. A signal's entry is a level that price must trade through, which is what a stop-entry order does. The protective stop follows once the position is open.

The orders do the watching. That suits anyone with a day job, and it was one of the original reasons Marc Rivalland took to swing charts: the method needed a mechanical component, so orders could act while he was busy elsewhere.

Using the API from a spreadsheet or a script

The data comes back as JSON. A spreadsheet can pull it in with a few lines of Apps Script, as the quickstart shows, and a script in any language can do the same.

A simple spreadsheet might have one row per signal, with columns for the stock, signal type, entry, stop and the risk per share (the difference between them). With the account size and the risk percentage in two cells at the top, the share count becomes a formula. The same layout in a script is a loop over the response.

Keeping the API key out of the sheet's cells, and in the script's properties or an environment variable, is a good habit whatever the tool. The docs have the field-by-field detail.

Keeping a record

A trade log turns a method into something that can be checked. For each trade, a line with:

• the date, stock and signal type
• the entry, the stop, and the risk per share
• the exit, and the result in dollars a share (or points)
• one line of notes

After a few dozen trades, the log answers the questions that matter. How large is the average winner, compared with the average loss? How do the results compare with what the backtests suggest? Is the method being followed as written?

That last question is the one that most often needs answering.

Mistakes the method warns against

Marc Rivalland ended his book with a short list of errors that experienced traders still repeat. In paraphrase:

• Skipping signals after a poor run. Nobody can tell in advance which signal will be the good one, and a method can only be judged on all of its signals.
• Raising the size to recover. A bigger position after a loss changes the risk, and the method's arithmetic stops applying.
• Forcing the pace in quiet markets. Sideways months are normal. Added risk to make up for them is not part of the method.
• Trading against the trend. The method follows swings; it doesn't call turns on instinct.
• Reading too much into the latest move. Trends run in waves, and every strong run or sharp fall has had corrections. Headlines tend to extend the latest move forever.

Where to go next

The free signals show all four signal types on real Dow stocks, and the pricing page lists the plans for the full data and history.