What I’m learning about trend days is that they cannot be predicted per se. they materialize from unexpected news or events. Those gonzo moves, not measured moves I mean. While measured may be anticipated at higher probability. Gonzo is found at low probabilities.
The biggest issue I have is with the counter-spikes of volume that trash the directional breakout trades on occasion. What I have done to mitigate that is to adjust for breakout and wait until a solid return to that 2 SD breakout level to be long or short, depending on net volume direction and price momentum. The breakouts using that method work, but you have to be keen around those edges beyond the breakout point. I think if you could survive a wider stop level, they work over time, but that kills the win/loss ratio when the targets are predefined. If I find more time, I am going to do the distribution analysis I think you have seen for that automated model I used before volatility got wild in late 2018-early 2019. You can actually anticipate what the median level of time in market should be before you decide to leave the market. It helps define the boundaries a lot better than just winging it.
What I’m learning about trend days is that they cannot be predicted per se. they materialize from unexpected news or events. Those gonzo moves, not measured moves I mean. While measured may be anticipated at higher probability. Gonzo is found at low probabilities.
The biggest issue I have is with the counter-spikes of volume that trash the directional breakout trades on occasion. What I have done to mitigate that is to adjust for breakout and wait until a solid return to that 2 SD breakout level to be long or short, depending on net volume direction and price momentum. The breakouts using that method work, but you have to be keen around those edges beyond the breakout point. I think if you could survive a wider stop level, they work over time, but that kills the win/loss ratio when the targets are predefined. If I find more time, I am going to do the distribution analysis I think you have seen for that automated model I used before volatility got wild in late 2018-early 2019. You can actually anticipate what the median level of time in market should be before you decide to leave the market. It helps define the boundaries a lot better than just winging it.