Let's look at a real example of how TA is created by real world actions. For example, CL (oil) went up due to the USA killing an Iranian General. So obviously, trend was up. We then also look at Price action for the following trade that I made.
The Fall of 2018 20% down move did not violate the primary monthly trend line of the SP5500 from 2009 to present (2020)
I like reactive TA so I liken it to a surfer riding a wave. The waves tell you when to paddle (adjust the position), when to wait for the next swell (stay on the sidelines), and when to ride (...the trend)
I have a couple of absolutes - "cut your losses short" "position size is key to survival" But I guess that counts as a few...
But shouldn't one be removing "belief" from consideration? Combine multiple methods by doing each independently, then seeing when they align or contradict, and learn how to apply a weight to each to learn how to make/develop an informed judgment call? (now code that... lol)
"Most Price Action is Random Technical traders tend to attach plenty of meaning to each bar or candlestick, agonizing over its relation to others, searching for chart patterns. Adam takes the view that the vast majority of price action in any market imitates a random walk. In these market conditions where price action is random, trading is, theoretically, a zero-sum game. However, in the real world, it's actually a negative-sum game due to execution errors, commissions, etc. Taking this view, achieving positive expectancy in your trading is as much about identifying conditions not to trade within as it is about identifying setups where price action is less than random. Adam holds that he’s never encountered a consistently profitable trader who trades a mechanical system with no discretion, and the randomness of price action is the reason why. If a trader doesn’t have a working understanding of price action and market structure, they won’t be able to identify the market conditions to avoid taking trade signals within their mechanical system. Even though Adam provides trade setups in his book, he refers to them as rough guidelines for identifying conditions where there is sometimes buying or selling imbalances. Because entries, exits, position-sizing, and units to risk per trade are relatively arbitrary, these setups cannot be traded mechanically without an understanding of price action and market structure." This is why you do need to go with a belief and then and then only then use PA for the trade setup. This is also why most purely coded systems fail over time. I was able to add more contracts to my trade today due to my belief in the direction on price before I EVEN opened my NT chart. Then I found a TA or PA setup that looked valid and was able to place the trade which worked out.
This is my win percent from the last week including today using my most updated understand of PA and TA combined with my belief on where price is headed. If you can achieve a high enough win % on equal risk vs reward, then you can over time have a successful trading system.
This is very interesting. I'm going to have to spend some time on it to understand fully. I think I do something similar to "If a trader doesn’t have a working understanding of price action and market structure, they won’t be able to identify the market conditions to avoid taking trade signals within their mechanical system.", but not nearly as specific. I don't have what I consider a good "working understanding of price action and market structure", but substitute that with having developed a "feel" by watching book and price action for the security I'm trading. When I see conditions that don't match what I validated my trading rules against, or doesn't "feel" right, or "feels" wrong, I won't enable Auto Entry. When the price/market is looking like its behaviour is "normal", then I watch for one of my setups, and when I would have previously manually entered a position, I instead enable Auto Entry. It then follows my entry & exit rules, acting upon my defined signals. It follows those rules acurately and reacts much faster than a human can. Upon Entry it disables Auto Entry. If the conditions stop looking "right" while I'm in a position, I can select some tightening rules (click on their radio buttons), or I can throw those shares into a tighter category of rules with a single click (I have Running, Entry & Tighter. My default upon entry is the middle, Entry.). If conditions start to look wrong, I Close the postion before my rules Close it. I can't find a way to code that recognition of conditions, if they can be, because I don't understand them anywhere near well enough. This misses letting some Positions Run, but if the position hasn't gained a decent amount of room after entry, I'm not comfortable letting it run with rules managing that position that weren't validated against those conditions. So I think I'm heading in a similar direction to what you describe, but I'm going to need to reread a number of times and think some more. Thank you
%% Sounds right dr Cruz; actually I'm taking your word on that. I remember SPY,QQQ sure Went below 200 day moving aVerage plenty.That DEC 3 week 2019 down move violated the 50 dma , on QQQ, but closed nicely above it, same day. JAN 3 week 2020,QQQ down move did not violate 50 or 200 dma, or close below them .................................................................................................
I don't buy that the market moves randomly. But in the past decade or two, there have been more and more factors that can and do influence market moves, and the digital world means they come into play significantly faster than in past times. I believe there are so many, that it is valid to say that this makes the market Effectively Random. But all of the factors have to be summarized & consolidated in one place - the price. A few years back I spent six months exploring DSPs and developing some for my use. Which one with which parameters, would only work for a part of a day. There were too many overlapping 'signals', coming and going with varying amplitudes. Running multiple DSPs with multiple parameters in parallel looked interesting, but in the end looked very much like varying curve-fitting. Pick the one(s) that provides the confirmation bias you want. I couldn't come up with a way to weigh, or select which to use/trust right-now. I suspect this could be a good task for AI. Not within my capabilities. However, within an Effectively Random market, there are periods in which there are clearly discernible dominant actions. An action could be from: a particular factor that has become momentarily strong, a factor or factors that countered others have subsided temporarily, a number different factors have become aligned (independently or due to a common influence), for a unknown period of time, with a sudden or gradual alignment, and a sudden or gradual eventual loss of alignment, a volatile churn of factors that coincidentally align, with new factors aligning as supporting factors churn out, manipulation, any of the above resulting in decisions to join in that action. The list of those factors is huge. Even a cursory attempt at a list of the major factors is problematic (for me anyways). There is no way I can identify and gather data for all of the factors, let alone grade, weigh and come up with a net anticipated influence. Again, could be a good task for AI. But I can use TA to identify discernible price actions and identify: when they are tradable, have an expectation of profit, and a low expectation of loss. Up to this point, I've developed my custom indicators with a view to identify price trends, entry conditions and conditions with a high expectation of profit. Your post brought to my attention that I haven't paid adequate attention from the view of when NOT to enter a position, counting on this-is-a-good-time as being sufficient. They sound equivalent, but taking that specific point of view gave me ideas for three new indicators to say "stay clear". Trying to code mathematically what I see in a chart that says 'stay clear'. Can't hurt to see what value they bring.