Trying to get into swing trading and develop a strategy please give your thoughts!
**(OK before people get mad yes I used claude to summarize my strategy. No, this is not just a strategy that I asked claude to just make for me. Im working on coding with claude to create a swing trading widget that will send alerts and do automatic sl and tp calculations based on my rules when certain requirements are met to enter a trade! Please also keep in mind this is a new strategy and dosent have alot of data to back it up yet).**
**The three-timeframe stack**
**• Daily sets the bias.** Price has to be cleanly above both the 20 and 50 EMA with the 20 above the 50 (uptrend), or the mirror for a downtrend. Anything else is “no bias” and gets skipped. The daily trend is the tide — you only fish in its direction.
**• 4H finds the setup.** Given a daily bias, the 4H is where an entry takes shape: a pullback to the 4H EMA20, or a breakout past an established level.
**• 1H confirms the trigger.** The setup isn’t a trade until a *closed* 1H bar confirms it — a close back above the 1H EMA20 for a long, below for a short. This stops you front-running a setup that never actually turns.
**The two setups**
**• Pullback:** in an uptrend, price dips back toward the 4H EMA20 (within \~1.5%), then a 1H bar closes back above its EMA20 — the dip is being bought. Enter on that reclaim. Stop below the pullback low (or the 4H EMA50, whichever’s structurally sounder); target the prior swing high. Shorts mirror this in a downtrend.
**• Breakout:** price *closes* a 4H bar beyond a level it’s been respecting — and the level is measured excluding the most recent bars, so the current move can’t define its own breakout. Stop is the base of the move that produced the break.
**The gates every trade must clear (all AND conditions)**
**• Bias separation** — the daily EMAs must be at least \~0.25% apart. A few cents between them is noise, not a trend.
**• 4H agreement** — the 4H can’t lean against the daily bias.
**• VWAP agreement** — price on the correct side of an anchored VWAP (anchored to the last earnings or swing).
**• Reachability** — the entry can’t be more than \~2% from current price, so you’re not chasing an already-extended move or a resting order that’ll never fill.
**• Volume** — breakouts need conviction (≥\~1.1x average, since thin breakouts are the classic trap); pullbacks just need life on the reclaim (≥\~0.8x, since pullbacks legitimately form on low volume).
**• Earnings** — no trade within 48h of a report.
**• R:R ≥ 1.5** — after structure sets the stop and target, if reward isn’t at least 1.5× risk, skip. The order is fixed: structure → stop → R:R. The R:R check never reaches back to pick a friendlier stop.
**Position sizing**
Fixed fractional: risk a set % of capital per trade (with a hard ceiling). Shares = risk dollars ÷ stop distance, so dollar risk stays constant whether the stop is tight or wide.
**The logic tying it together**
The whole thing is built to keep you *out* of bad trades, not to find more trades. Each gate removes a specific way people lose: no-trend chop, counter-trend entries, chasing extended moves, failed breakouts, earnings surprises, and trades where the math doesn’t pay. What survives is a small number of setups a year where trend, structure, timing, and payoff all line up — which means a lot of “wait” days and few actual entries. That’s the intent, not a bug.
**Where I’d genuinely like input:** the thresholds (volume multiples, the 1.5% pullback zone, the 2% reachability cap, the 0.25% separation) are starting hypotheses, not values proven by a track record yet. Curious whether people think any of these are too strict, too loose, or measuring the wrong thing — and whether the earnings/reachability filters are worth the setups they cost.