Market’s been dead. I’m hoping the summer thing is done.
Traditional story is Labor Day, desks fill back up, tape wakes up. I tested it instead of nodding along.
September is usually the soft month. Covered that. This one’s flat so far. Fine.
The useful bit: Sept/Oct candles run about 20% bigger than summer. SPX average bar sits around 1.2-1.3 in the quiet months then you get that spike. December slumps back toward summer. Christmas.
It’s not just red bars being fatter because September is weak. Up days get bigger too. Down days get bigger. Just more action. July’s the tightest. August a little more. Then Sept/Oct.
If you eyeball stops or risk a flat 1% on QQQ every trade, that math changes. Same dollar stop in a bigger-candle month means you get tagged more.
I ran a dummy system. Percent stops = messy year to year. ATR stops = smoother. Because some months stocks just move more.
ATR is on TradingView. Last 5 days, what’s the average candle. That’s your yardstick. Half a normal day for a tight stop. A full day for a target. Number updates. You’re not stuck with summer sizing in October.
Every stop in the StatsEdge book is ATR. Bids too. It flexes.
Free list is still one idea a night. New systems are up. statsedgetrading.com. Get away from the screens.
The ATR versus percent point is the one that costs people the most money and gets the least attention.
The part I would put next to it is what range expansion does to entries, not just to stops.
If an honest stop has to be at least a day's range wide, then in a big-range month the breakout print gets expensive in a hurry. The nearest level you can actually defend is further behind price than it was in July, so the same trade needs a wider stop and smaller size to carry the same risk.
That is why I take fewer breakout prints and more pullbacks once ranges open up. When price comes back to the level, a one-range stop still sits under structure instead of out in space.
Same yardstick you are using. It just changes where I am willing to buy, not only where I am willing to be wrong.