Trading today, I realized that the EMH must be false. I respectfully disagree. The reason is that if new information becomes known, traders/algos can't just instantly submit huge buy or sell orders. This would devour standing orders on the book, and they'd just put the price up or down on themselves--way beyond its true value. This is why we have VWAP and other pacing algos, to balance the flow of orders with market supply and demand. It takes a bit of time for new information to be digested into the markets. Are you as excited as I am?!! 1. https://en.wikipedia.org/wiki/Efficient-market_hypothesis
This is the wrong interpretation of the efficient market hypothesis. EMH is a framework that is useful for market participants to evaluate their strategies. The idea is that if prices reflect known information, then you cannot make excess returns analyzing known information. Analysts/traders/portfolio managers everywhere analyze data in order to spot market anomalies and arbitrage opportunities, by using EMH as the framework. You should read into factor pricing models like Fama-French or APT.
But most of the time 'new information' is new only to you and me. Most information is known to those who move market. How do you factor that. Do you really think ppl wait for a news before buying 500k aapl shares.
Also, those big investors are not trading the news lol. When they submit their order for 500k shares, they probably spent 1-4 weeks analyzing the company and speaking to management. And no big investor wants to have a big market impact, which is why they use algos...
Good thought... I did some work last month on a model for flow of market-moving information, from source to public.
EMH is clearly false. Otherwise, it would be impossible for prices to change without new information becoming 'available.' Yet they do all the time, often dramatically. The premise assumes that information, and only information, moves markets. But other forces move markets as well. If a whale hits the sell button, the price drops because supply overwhelms demand. Unless you define 'information' as 'price action based on whatever happens,' EMH has holes, but if you define it that way, then it collapses into a tautology or a definition, rather than a useful theory. ie, the price reflects the information that a whale sold 50mm shares in 5 minutes. However, that reverses the causality. The information didn't cause the decline in price, the fat sell did, and that was reflected in the available information, which showed up on the volume print. Also can be applied to market panics (often false ideas, not information) and manias. Big players triggering stops to harvest profits. How could prices possibly spike by 20%, then reverse if EMH were true?
For me EMH has always been false, but it's false in an interesting way. Take a market with 100 participants, each participant has a fixed IQ of 100 and similar historical backgrounds. New data is release to the participants and they act accordingly, buying or selling to move the price up or down to their perceived value position. This is EMH. ----- Now let's just give those participants different life backgrounds, and suddenly we have extra information for some percentage of participants that others do not have access to. Perhaps they work in biology and so have a better understanding of new biotech stocks. In this case, data is released, and over time assimilated by those who can transform that into information, while others have no play in the situation (but may buy or sell anyway). ----- Now let's give those participants different IQ levels, some as low as 60, and some as high as 140. The higher IQ participants can more quickly parse the data into information, and can draw inferences and connections where the 60 IQ participants cannot. In the case of a highly specialised participant (General AI), most interconnections can be established, and therefore, an optimal price level fairly easily calculated. ----- In short, tldr, EMH is false as new financial data is transformed into information by participants. New data is released, yet only some people have access to this data because of their background and ability to connect the dots.
In your example, the whale took action based upon a signal. When reversals occur w/ low volume, they have a greater chance of failure. When reversals occur w/ high volume, they have a greater chance of success. There are moments where certain types of volume is anticipated to come into the market for the current market cycle to ‘continue.’ When this anticipated volume fails to materialize then ‘change’ is on the table. As for causality, effects are always perceived first. As for EMH, the velocity of which information propagates and the ‘meaning’ of that information within the present market cycle and context is where smart money positions itself. One either approaches their participation in the markets w/ logic and rationality or they don’t. The markets reflect more of a participants bias than anything else. The framework and methodology by which one generates that bias is a durable edge.
You are basically correct. It is this limitation in the order book that many theorize is responsible for the square root law of meta order impact. That is if you are trading a really big order that is split up, there's a limit to how much impact you can have on the market. Instead of your impact being linear it follows a square root law. This can prevent markets from fully pricing in new information quickly, and is yet another falsification of the efficient market hypothesis. The market tries to be efficient, but there's many reasons including structural reasons why it can't always do a good job at that.