IN looking to exploit a trend, is past data relevant for the trade you are looking to enter? For example, if all of the data to left is what happened, is there any use in looking at the data before entering?
I trade on a purely mathematical model. This model identifies a big number of scenarios that are grouped around different strengths of the trend. The system identifies, based on past data, which scenario is most likely to develop. Basically it starts by eliminating all the scenarios that dont fulfill the conditions. At the end there will be only 1 scenario left over. Thats the scenario that will be initiated. But as markets are dynamical, your trading should be dynamical too. So the scenario is checked continuously every tick, and if necessary (when conditions for the scenario are not fulfilled anymore) a scenario will be closed or replaced by another. So either the trade can be closed or the trade can continue following another scenario. So past data is, within certain time limits, essential for feeding the mathematical model with data. The data is needed to calculate which scenario it is developing and to see if this is the scenario we are in.
Next, assume one was looking at trends and defined them in say... 8 different strengths or so... and the assumption was daytrading the ES for example, would it be prudent to define various days as a particular trend delimiter. For example a short squeeze day, will have two major trends associated with it, a strong down trend, followed by an even stronger uptrend, these two would encompass the entire day. A strong uptrend day, would just start in the left lower corner of your chart, and end on its highs in the upper-right hand corner with minimal pullbacks. Clearly two different types of days. There could be several other types of days as well. But in these examples, would it make sense to define trends in this way? Or can a trend exist that may encompass several of these days together?
All depends of what you are able to build as a system. So I can only speak about my own experience, but there are surely other valid opinions too. Each day is a new day, there are no predefined opinions of what will happen that day. The mathematical model calculates in an eternal loop and gives me information about what to do. I just follow the instructions. When the day is finished it is always clear what kind of day it was. Then the checkup starts: did my model guide me thru this day in an optimal way?
The most difficult part is to obey your model without any doubts. You are always influenced by your brain that might want you to do something else and not to follow the mathematical model. It is like going long when everybody is sure you should stay short or go short. Like the market making each time a new low and then suddenly in one of these new lows I should go long. And vice versa. Very tough until you are used to it and have build enough confidence to follow the model and not the market.
You mention OHLC information. Again assuming you wanted to "day trade," for your studies, would daily OHLC data be sufficient? Or would you need to see what happened INSIDE the period you wished to trade? Like take your days, and figure out how the day developed?
All I need are the consecutive ticks. There are so many potential scenario’s that can develop and change during the day that, to me, it is useless . I don’t figure out, I obey my system. Tests have shown that if I try to improve the mathematical model, the results will go down and the drawdowns will go up. So the choice was not difficult for me. My personal experience or intuition cannot beat the math. And it makes trading also much more relaxed. Maths are doing the difficult part of trading. I often have periods of 1-2 hours, and even more, that I am sure I can stay in my trade and shouldn't even look at it. Most daytrades have to watch hours and hours to their screen.
Is time of day significant? Is duration of a swing significant? Or is it just distance that matters?
It is very simple. The model tells me what to do, I don’t have to break my head on the time of the day, duration, swing or not, support or resistance.
About the number crunching: I had to replace my computer because in very active markets he was not able to update the information quick enough. I had an Athlon II X4 630 and replaced it with an Intel 4770K running at 4.3 Ghz. When I bought this new PC the man from the shop did not believe that my Athlon could freeze up, till I showed him in realtime what was happening. Just watching the opening was already enough to freeze up.