They say that the market is constantly evolving and one should adapt to it, but to which extent? Some traders say that they are looking for new patterns some part of their time and the other part they're following their system. I trade for about several month and want to build a persistent system, and a bit puzzled about the degree of impermanence that implied here. What kind of system is solid enough? To which extent I may want to modify it during the time?
Patterns and behavior that existed in 1908 still exist today because human nature does not change. What has changed however are the logistics in the market.
Get used to the idea of impermanence -- it will help you survive. As far as how to deal with that impermanence day-to-day? Get used to the idea of "telescoping" -- where you switch from 1min to 1hour to 1day to 1week (or whatever perspective-changes work best for you)...and back down -- a practice that will alert you to what the greater timeframes hold. DO NOT operate under the misapprehension that whatever you look at in December will hold for you forever, or through to next December, or even into next week. You must *always* have other irons in the fire, and be testing both what you're operating, and what you *may*be* operating next. What looks great today may be obsolete next year, and in deep trouble by just next summer. And likewise, what looks like it should hold promise right now but *sucks* -- that may be your bread&butter next year. Yup yup yup...!
That would be the Holy Grail ... Not even worth searching for it, imo. In one of Perry Kaufman's books he said that he wants his system to be like a cockroach ... Survive on very little in any condition (it is believed they can survive nuclear apocalypse). Well, that sounds great but very unlikely to achieve. I think it can be achieved but requires large investment in tech (hft). So for regular folks, I think to start, you need to create simple systems to fit different market regimes. Then you can either chose when to run which, or run them in parallel but varying how much to allocate to each. Like these guys(not a recommendation, just some good info): https://algorithmictrading.net/
Can you please give an example of such thing? Say, I trade rebounds and I can imagine that it requires 5 consolidating candles near border now instead of 3. How fast the changes requires to adapt to? Is it enough fire for irons if I'm starting and trade part-time, say, 2 hour per day?
Nobody is talking about candles or price patterns. If you're still in this stage, I'd suggest you papertrade for another year.
Could you elaborate on this pls, what kind of patterns do everybody talk about? Purpose of this post is to clarify this things. I understand that it can be vague a bit, but nevertheless
"Patterns" as named entities are meaningless. A Japanese Candlestick is a *wonderful* graphical representation of 4 pieces of data -- 4 observations -- in 1 piece of time. But that's all it is. There is no magic; there are no secrets. Consider this DIRTY LITTLE SECRET: if you take an hour's worth of 1min data, and set up 12 5min candles, you will undoubtedly perceive "patterns". However, if you take that same data, and move the definition time 1 or 2 minutes, and lay out 5min candles at 02:00 and 07:00 and 12:00.... you will have an entirely different set of so-called "patterns." (And the trades supposedly based on those patterns would likewise have changed. Just by scootching the defining window by a slice or two of time.) That is not science, that is chicken guts, tea leaves, and rolls of bird-bones. SO! What IS rigorous enough for trading? For whatever entity you're trading, look for long tails. That is a loud announcement that the market could not hold that price extreme during that time period; that is an irrefutable fact. Check out the volume -- does it fall off or spike dramatically? This are facts of finished trades. If the time window you've chosen does not exhibit long tails (wicks) at market turns, telescope in or out until you can see the market's agenda plainly demonstrated. This is what they wanted to do. (And is what, in fact, they *did* do!) Whether they keep it up during your trade? That's another cup of coffee. End of story.
IMHO, no matter market or algos evolve, this very basic principle of trading won't change: Buy at cheap price and sell at expensive price. My job is to gauge how market is cheap or expensive based upon my fundamental and technical analyses.