For averages, smoothing will cause "lag" as well as reducing high frequency amplitudes and distorting the signal some. May be handy as a tool to compare latest prices toward a smoothed "old" price, or for other uses one may dream up. Though, smoothing is not "pure" lag, only shifting the data forward in time will be "pure".
All indicators are old news, as price and volume are old news. But price isn't random, it's very often a kind-of "continuous" signal with small amount of gaps. random(1000) should be what we can call random, and would gap almost all the time. So liquidity, market mechanisms and transparency are causing a signal to trade on: price per tick/time.
Even though price and volume are old news, does not mean most participants are reading and playing the markets perfectly according to TA. On the contrary, participants have all sorts of reasons to participate, many timeframes, goals, greeds, fears, limitations and errors, thus creating chaos. Even the bots will do that to any market, so is something basic when participants behave differently. TA is every possible analysis on price and volume, so does not really dictate any one behaviour either.
Now, indicators based by price and volume by themselves, are useless compared to raw price and volume data. What really matters is what you do with those data, and for that you may use mathematical formulas and/or logic. Some call these indicators, but they're just math, formulas that do nothing by themselves.
So, beware of indicators, not just because they're inferior to the raw data, but if you want to trade it, you also need to figure out why, as well as how!
