Zero-Intelligence Trading Dominates Price Action
Zero-Intelligence (ZI) traders act randomly with minimal constraints. This year ZI traders have determined price action and are causing difficulties to traders with non-random models.
These ZI traders are mostly individuals with programming knowledge that are trying to capitalize on platforms that offer an integrated environment for algo developments, backtesting and trade execution. Most have no trading experience and use models that suffer from data-mining bias. These models show good performance in train and test samples but in reality are no better than a fair coin toss.
A crucial aspect of this recent trend is that ZI traders share their random algos and make minor modifications. Their objective is profit sharing in case they are able to develop and algo that will be used by the platform operators to manage money.
Needless to say that without major modifications and enhancements this model will fail since it generates low volatility noise that defeats its own purposes. Specifically, intelligent traders stay on the sidelines when trading becomes random or decrease exposure. Since trading is a zero-sum game, soon these noise traders will have no one to profit from other then their own group. Then losses will start accumulating.
The chart below shows how ZI traders have actually determined price action year-to-date in SPY ETF. We simulated 20,000 fair coin traders that buy at the close if heads and sell that the close if tails. This is long-only SPY trading in daily timeframe. We included commission of $0.01 per share and the initial capital was $100K.
The important result from the simulation statistics is that nearly 94% of the ZI traders have profited this year. This further means that these ZI traders have actually determined price action. This is one reason that the market has exhibited low volatility in recent months. This is the end result of having many random algos in the market buying and selling securities based on similar logic but in essence no more effective than a fair coin toss.
We recently contacted a few experts in quantitative trading and machine learning that use our software and we received interesting replies about market conditions and performance year-to-date. We will include those in another article because we think the subject is of interest to all traders.
The majority of ZI traders will be out of the market soon because they will not make any money from their activity and will have to find a real job in programming where they can actually contribute and even make a good salary. In the process, this will be a little painful for professional traders. We believe that if intelligent traders withdraw for a few months, most ZI traders will disappear. Patience is always a virtue in trading. The idea that someone without knowledge of the markets can write a program and profit from data-mining is ludicrous. At the same time, platforms will survive but they will have to attract real talent, pay for it and lower expectation because operational cost will increase substantially.
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
Technical and quantitative analysis of Dow-30 stocks and 30 popular ETFs is included in our Weekly Premium Report. Market signals for longer-term traders are offered by our premium Market Signals service. Mean-reversion signals for short-term SPY traders are provided in our Mean Reversion report.
via Price Action Lab Blog | Quantifying Market Price Action http://ift.tt/1hnoc7s
May 15, 2017 at 01:09PM