And yet, it's not. That's the bottom line. Those who don't understand fundamentals and why trends exist are doomed to posting that everything's random. Because they don't understand shit about what stocks really are. And by the time you've figured out a trend is underway, yes it might be over. But that's you man, you are basically lazy. The reality that it's your lack of awareness and research that is at fault won't phase you when you can excuse your lack of effort with quaint theories that have no basis in reality.
Looks to me the random market believers are a bunch of curve fitters. They have no other way of getting a handle on the market than through random correlations that are merely incidental and have no statistical validity, nor would they project forward in time. So I can see why marketsurfer is looking for statistical proof, because try as he might he never found it himself. A clue I can offer is that there is a direct (and beyond statistical) correlation between demand and price going up (up trend), supply and price going down (down trend). What I can't offer is where to find these things. I had to put in the time, effort, money and the kitchen sink to find them for myself. So it must be for everyone else. Therein lies salvation. I believe someone who is good with a simple business, such as running a fruit and veg stall, or a hotdog stand are better placed in understanding the markets. There are lots of similarities.
This proves that no matter how smart you think you are, there will always be people who will be smarter. Even if they have a lower degree and should theoretically be more stupid. Many times thinking out of the box is more valuable than become PhD or anything else. Intelligence alone cannot always solve the problem. I know Phd's who are beaten in trading every day again. PhD is not a miracle solution and all they"achieve" should not always be interpreted as the only and undisputable truth. I have seen verifiable proof of real trading results that by their size and their statistical probability destroy your complete statement about randomness. One day these things will become public. But for specific reasons these things are kept secret at this moment. Making it public now can damage the economical value of these "systems".
You don't need to predict. It is clear that you cannot see a trend before it has started, so you are always too late. You should try to see as soon as possible "in hindsight" that a trend is developping. A trend is a big move from which you should try to take the biggest part. If a 5 hours trend is spotted 10 minutes after it started and you get out 10 minutes after the end of the trend you can make a lot of money. This can be done. If you take 9 points out of a 12 points trend it proves that markets are not random.
In statistics, there is a clear difference between deterministic trends and drifts. And it is possible to have a random walk combined with a drift or with a deterministic trend or both or by itself or etc. In trading, it probably does not matter to most as terms like drift and trend are conflated, especially for those trading with technical or order book set-ups. But if you run a statistical framework it probably is wise to be aware of this.
I think it is very easy to find out who is right, no matter if it is trend or not trend, random or not random, trend or drift........ If you have a certain opinion, this opinion should be visible and confirmed in the results of your trading. If your trading is succesfull it means that, for you, your opninion is the correct one. All the rest doesn't matter.