It’s a simple PA system based on reversals, A reversal is a break of support if long, Or a break of resistance if short. It’s a long/short system, It’s either long or short 1 contract. How to avoid going short at #3/#5 then buying back higher at #4/#6 ? Any rule of thumb ? Could be using thighter stops, If close > entry candle’s high then exit short. Here it doesn’t enter on bar close, But just assume it does. A guy said “A trendline should never be steepened but always flattened”. There’s something similar with horizontal lines. Just asking because … Everyone, but me, makes money here.
One buy one sell is a very brittle system. Imagine you are trading this with nine figures. What do you do?
You mean reversing thousands of contracts ? Well … I stop trading and buy an island But that’s a good point. A reversal might not be a sign to reverse. But still it’s a danger signal Therefore I should get out, at least. For example I exit a long on a reversal (Short), Then I either buy the next reversal (Bull) Or I sell the next breakout (Short)
A poster by the name Ammo, ~10 years ago, successfully traded against the ES trend for a while using multiple buys / sells along the range inside the trend. I imagine he would be buying that island by now if he had just gone with the trend.
Supports are purple and resistances are orange. The levels are drawn programmatically, Agree it’s not beautiful but … The levels are here.
if you read all these stories about futures traders, ones who made it were those swings/postions carry trades with no additional leverages with defined stop loss. in another word, you don't have big enough balance to trade more than you can afford.
Hi Sekiyo, One of my systems has multiple conditions for entry, including a rule that price MUST close on or beyond an empirically determined "best" EMA line. The TOS range chart below doesn't appear to be a perfect match to yours, but does suggest that this rule might have prevented the loss at #3 if a 10 EMA line was applied. My system currently uses a 14 EMA line, which also would have worked here. Might be something worth testing...