METHODOLOGY

How SEPT Tests Trading Ideas

Short answer: SEPT research checks a trading idea against years of real US market history before we say anything about it. Rules are fixed in advance, results must hold up in two separate time periods, costs and realistic stops are included, and we publish what failed as well as what held up. Results describe the past, not the future.

Our rules

  1. Rules first. What we test is fixed before we count anything. We don’t adjust it after seeing the results.
  2. Two periods must agree. Every result is checked on two separate stretches of history. If they disagree, the result doesn’t count.
  3. Realistic trading. Stops act like real stop orders, and trading costs are included.
  4. No hindsight. A test only uses information that was available at that moment.
  5. Liquid stocks only. Tests use actively traded US stocks, re-checked over time.
  6. Failures are published. If an idea doesn’t hold up, or we find a mistake in our own work, we say so and withdraw it.

Every study gets a label

LabelMeaning
ValidatedHolds up in both periods
Tested, not profitableMeasured honestly; no edge after costs
Not confirmedLooked promising, failed a further check
Not validatedNot checked yet; not used
WithdrawnWe found a problem in our own measurement; never used

Important limits

  • Hypothetical results from past data: not a forecast and not advice.
  • Real-world fills can differ, and small edges can disappear after costs.
  • Results cover many stocks together. A single stock can behave differently.

Questions: support@septinvesting.com

FAQ

Does SEPT predict where a stock will go?

No. SEPT research describes what happened in similar past situations. It is not a forecast.

Why check two separate periods?

An idea can look good by luck in one period. Requiring two periods to agree filters out most flukes.

Why publish failures?

Most trading ideas don’t hold up, and knowing what doesn’t work is as useful as knowing what might.

Not financial advice.

Weekly: key levels to watch + one trader question answered. Free.

We email a confirm link first. Unsubscribe any time. Tools and data, not investment advice.