Trying to rationalize intraday price moves was always an exercise in futility. Since algos have dominated market activity, rationalizing price action moves is an irrational activity that is prevalent mainly among retail traders.
Rationalizing price action means placing it inside the bounds of simple logic. What can simple logic infer from random activity other then the fact that it is random?
Algo activity is mostly random, just like the motion of particles suspended in water due to collisions, also known as Brownian motion. In the short to medium term, the market may have a drift, usually it is upward, but that is attributed to position and passive investors actions rather than to algo activity.
Most algos attempt to buy at the bid and nearly simultaneously sell at the ask, which is equivalent to market making, then repeat this many times during the day and profit from the spread that retail is willing to pay to participate in the game. Other algos just create random swings in an attempt to shake out weak counter-parties. There are also more sophisticated algos in high frequency domain that attempt to profit from exploiting order flow. The list is long depending what edge the developers target but one thing is certain:
Algo activity cannot be analyzed with simple (mainly propositional) logic and therefore cannot be rationalized.
Yet, thousands of retail traders attempt to rationalize intraday moves every day. Some do that for fun, or a sort of recreational activity, but the majority takes the task seriously. But this is an exercise in futility. For example, notice the swings in the intraday 5-minute chart of S&P 500 index on April 3, 2019:
There is no way to rationalize the action due to random activity. The daily chart summarizes the activity as a small star-shaped candlestick, as follows:
In the daily chart traders try to rationalize daily moves and ignore intraday action. For example, traders care about things like the golden cross that occurred a few days ago and other patterns, such as RSI(14) indicator divergences. But even on this level it is becoming more difficult over the years to rationalize price action and for systematic trading we started already employing strategies that operate in the weekly timeframe. In this way we hope of capturing the drift rather than the noise although slower timeframes suffer from higher lags and this is a problem to tackle.
Unless one considers it recreational activity, or just having fun, it is difficult to understand why so many retail traders try to rationalize intraday moves and even use simple indicators to make forecasts. The first thing you learn in basic statistics is that randomness means lack of predictability. Algos make the market random and as a result the intraday price action. As algo activity becomes more dominant, the randomness starts affecting higher timeframes. Impact on daily timeframe is in progress and in the next few years that will also turn random and unpredictable. This “randomization” will affect all higher timeframes until markets gain maximum efficiency and forecasting in any timeframe will be impossible. This process may take a few decades. However, the intraday randomization is now completed and it makes no sense trying to forecast price action in those timeframes as retail did in 80s and 90s. Intraday price action cannot be rationalized due to the complexity and randomness.
If you found this article interesting, I invite you follow this blog via any of the methods below.
If you have any questions or comments, happy to connect on Twitter: @mikeharrisNY
Charting and backtesting program: Amibroker