Is price movement really random and unpredictable?

[engineering]

If asset prices are primarily driven by information arriving in the market and since this information is by definition unknown until it arrives it is fair to view prices as random.

This is nonsense. Just because you personally didn't figure something out in advance doesn't mean it was random. It especially doesn't mean that future price movement is completely independent of past price movement.

For example, there was a time period where personal loans to Elon Musk backed by his stock holdings would have been subject to a collateral call if the stock fell below a certain price. The founder of a large company having to liquidate a portion of their holdings would significantly affect the price of the stock.

This nonsense is like saying "I couldn't predict the results of a sealed bid auction therefore the results are random."
The result of the auction were determined 100% by the bids and thus not random at all. Each one of those bids was submitted by bidders making and deliberate choice and therefore also not random. If you made an effort to identify the bidders and their economic situations, you might have been able to predict the resulting price.

I would actually predict that as t approaches infinity, the value of a stock approaches it's shareholder retained value plus discounted future earnings divided by outstanding shares.

 

[BlackPhoenix]

If the markets are not random, how do you explain why more than 90 percent of traders end up going belly up? That figure has always been consistent. Go figure.

Exactly because it's only 90% If the markets were truly random this number should be very close to 100%

 

[BlackPhoenix]

The law of large number do dictate that in the majority of cases, if you play a statistically significant number of times, you are correct. But trading in general is far from large numbers so an occasional "genius" trader will appear or an occasional "method" will work even if the market is a Markov. Look at this stock price chart, generated by GBM, very "tradable"

Yes, I have been through this exercise of generating truly random charts that look to a human as if there was a tradable signal and that look like as if it was a real chart. While we can agree that these generated charts do not have a signal doesn't mean there isn't signal in real charts just because they look similar to a human. While humans have tendency to find patterns where there is none (e.g. apophenia) the opposite is also true, i.e. humans inability to see signal in noisy data (signal detection theory). My thinking is that there's a signal that can be traded but which gets increasingly difficult to discern from random noise the shorter timespan you go.

 

[schizo]

Exactly because it's only 90% If the markets were truly random this number should be very close to 100%

Do you believe herd mentality exists in the market? Then you can't say 100% randomness can exist.

 

[padutrader]

Random walk theory claims that prices move randomly and are not influenced by their past. Because of this, it is impossible to use past price action to predict the future. But is it? You might not be able to predict what would happen in the distant future, but what about the next bar or candle? How many of you have a pretty good track record when it comes to guessing the next bar correctly?

i do not trade price so if it is unpredictable it does not matter to me. if it random or not is not important

i trade market conditions which can be seen determined and is predictable.

if it is range then buy low sell high and scalp and use very very wide stops -stay in position until the markets shows clearly that the market condition-range in this case- has changed then buy high and sell high

 

[schizo]

i trade market conditions which can be seen determined and is predictable.

Ohhhh, so you trade the "market conditions", eh? Then tell me just how those conditions are formed. Does it form out of thin air?

 

[BlackPhoenix]

Then you can't say 100% randomness can exist.

I'm saying the opposite

 

[Aisone]

If the markets are not random, how do you explain why more than 90 percent of traders end up going belly up? That figure has always been consistent. Go figure.

So this leads me to believe that the market is random only because 90% of those who ultimately blow up are trading randomly (and their trades are showing up as random on the chart).

Price movement for ES and NQ is not random, but the system that drives it is extremely complex and not easily understood, which is the biggest reason imo why most traders may not be successful. Stocks probably have more randomness, but similar manipulation.

 

[Slope Trader]

...Now, the curious question is what people are using to make predictions? Looking at a chart? How can you gauge probabilities simply by looking at a chart unless it's a specific set-up or similar which you back-tested and have the statistics for? If not, it seems more like guessing or shooting from the hip as opposed to prediction...

.

I like this comment quite a bit, it's my own preference too (text was underlined by me).

Agree with a comment heard elsewhere: Markets are often chaotic, but they are not random.

It seems important to recognize that trends persist, both short-term trends and long-term trends. When I began consistently trading in the direction of the current trend as defined by looking at medium-period moving average slopes my win% and other system metrics all improved. I use colored lines to simplify visualization and slope directions are also coded into my strategies.

Momentum/inertia is also an important concept here (shout out to Volpri who mentioned the concept in this thread too). At times, prices in motion upward or downward are more likely to continue in motion in the short-term and we can identify patterns/setups where it is more likely to continue.

Frequently chaotic but not random...

 

[panzerman]

If market prices are not random, then make them random, if that suits your trading style.

1. Model price data as pink noise
2. Whiten the data
3. Peak limit the data
4. Perform a Fisher Transform on the whitened and peak limited data

John Ehlers' work explains all this in detail.