How to find uptrend pullbacks with ChatGPT or Claude

A starting list for research. It does not recommend any trade.

Seven stocks matched at the Nov 4 close; these five closed above their average most often. The next session WS fell to $35.90, well below its average.

US stocks, ETFs and futures. The free plan runs 20 queries a day on daily bars; Pro adds intraday.

Paste these into the same chat, after the first answer, so your AI knows which results you mean.

• Chart the top three with their 20-day and 50-day averages.

  Copy

• Which of these reported earnings in the past month?

  Copy

• Any insider buying in these over the past three months?

  Copy

• Keep only the ones within 10% of their 52-week high.

  Copy

A pullback is a brief dip during an advance. This scan looks for a low that reached a rising 20-day average, followed by a close back above it on lighter volume. The 20-day average must also sit above a rising 50-day average.

The rule Monk.trade uses: common stocks averaging at least $20M traded a day, checked on the last completed session. The 20-day EMA must be at least 2% above the 50-day SMA, and both must be higher than five sessions earlier. The low must reach or fall below the EMA, the close must hold at or above it, and volume must stay under 80% of the preceding 20-session average.

• Pullbacks to the 50-day average

  “Using Monk.trade, check the last completed daily session for common stocks averaging at least $20M traded a day. Require a 50-day SMA at least 2% above the 200-day SMA and higher than 10 sessions earlier. Find stocks whose low reached the 50-day SMA and whose close held at or above it, on volume below 80% of the prior 20-day average.”

  Open in ChatGPT ↗Open in Claude ↗

  MonkQL

  from stocks
  let s50 = sma(close, 50)
  let s200 = sma(close, 200)
  include type == "common" and sma(close * volume, 20)[2] >= 20000000
  where s50[1] >= 1.02 * s200[1] and s50[1] > s50[11] and low[1] <= s50[1] and close[1] >= s50[1] and volume[1] < 0.8 * sma(volume, 20)[2]
  show close[1] as prior_close, s50[1] as sma_50, volume[1] / sma(volume, 20)[2] as vol_ratio
  sort close[1] / s200[1] desc
  take 5
  

• Pullbacks within 10% of a 52-week high

  “Using Monk.trade, find completed-session 20-day EMA pullbacks with a close within 10% of the 52-week high. Keep common stocks, the $20M average daily trading value floor and the 2% gap between the EMA and SMA.”

  Open in ChatGPT ↗Open in Claude ↗

  MonkQL

  from stocks
  let e20 = ema(close, 20)
  let s50 = sma(close, 50)
  include type == "common" and sma(close * volume, 20)[2] >= 20000000
  where e20[1] >= 1.02 * s50[1] and e20[1] > e20[6] and s50[1] > s50[6] and low[1] <= e20[1] and close[1] >= e20[1] and volume[1] < 0.8 * sma(volume, 20)[2] and close[1] >= 0.9 * highest(high, 252)[1]
  show close[1] as prior_close, e20[1] as ema_20, volume[1] / sma(volume, 20)[2] as vol_ratio, count(close > e20, 30)[1] as days_above
  sort count(close > e20, 30)[1] desc
  take 5
  

• Nothing on this page is a signal to buy or sell. Prices can keep moving against you, and you can lose money.
• The query reads the bar before the newest, which is the last completed session while the market is open. Before the open or after the close the newest bar is already complete, so this example, run before the Nov 6 open, read Nov 4. Ask your AI to check the newest session instead.
• A moving average may fail to hold as support. A change in momentum or news can push a stock below its 20-day or 50-day average.
• Stocks with fewer than 56 sessions lack the history needed to compare the completed 50-day average with its value five bars earlier. The scan reports them as undecided.
• Earnings releases, ex-dividend dates and merger agreements can bring a price to its moving average for reasons unrelated to a routine pullback.