The misconception of Artificial Intelligent trading in the automated traders’ community of Africa

Photo Credit To Emmanuel Tweneboah Senzu, Director of Research-Frederic Bastiat Institute

Artificial intelligent trading is the effective application of Algorithmic trading. Then, what is Algorithmic trading? It is the implementation of trading rules into a mathematical program by a method of coding and using that program to trade. Almost 98% of the young retail trading community as well as 58% of the Investment market of Africa that believe in automated trading, perceive this style of machine trading system as omnipotent to investment maximization, and a solution to risk management techniques applied in human trading for the market analysis.

In favour of the advancement of trading technology is partly an expression of man by nature to use the easier pain to maximize benefit. This has resulted in a downplay of traditional technical analysis, perceived as a complex approach and difficult method to have a good earning from the market. This is how the escalation in the campaign of machine learning on alpha generation became accepted as the best alternative of profit-making in replace to human trading, who are perceived with emotional and inefficiencies in decision making at the trading room]. Especially with long-term trading, there was a common misconception that profit-making and maximization of earnings are best found in machine learning algorithms program usage.

While the developers of Algorithmic trading software’s has focused solely on the beneficial aspect of their trading software to sell their products by discrediting the traditional technical analysis as a method becoming obsolete, and the future of trading is about Artificial Intelligence trading: as a scholar in this field, I hereby caution the young trading community of Africa to realize the silence risk in the overconfidence and reliance on Algorithmic trading software application without quality market forecasting skills on chosen market instruments. Not forgetting the emphatical statement of Dysart (1967), which states, ‘there is no mathematical system devoid of human judgment, which will continuously work without error in the real market.’

In a twelve (12) months correlation test of the past closing price to the future closing price of both the short and long term period of the following securities market trading as in Forex, Commodity and Stocks, there were fascinating findings, which indicate, without a careful and effective technical application of algorithmic trading software’s it damage to trading accounts is lethal. 

The short term period as analysis for this article used (4)hours as a minimum short term trading period, and (24)hours as a maximum short term period. And it was observed as follows, using a standard deviation measurement range of 0 to 1.0

  • In the above stated standard deviation range for the instruments in the forex market, the past price pattern to that of the future price pattern in the short term period variedly deviated at a range of 0.3-0.7
  • In the commodity market, the standard deviation range of the instruments of the past price pattern to the future price pattern was within the range of 0.2- 0.5
  • Finally, within the stock market, the standard deviation range of the price pattern of the past and the future was within 0-0.2

I then came to long term period observation of the market, with my long term period defined as (24)hours as 1-day trading, classified as minimum long term trades and (4) weeks trading as 1-month trading, classified as maximum long term trade. The following were observed as a standard deviation measurement range of price of the past and the future from 0 to 1.0

  • At the forex market, it was realized, the standard deviation of the price of the past and the future varied within the range of 0.8-1.0
  • At the commodity market, the standard deviation was observed to vary from 0.3-0.6 in price pattern of the past and the future.
  • Finally at the Stock market, the standard deviation of the price pattern range between 0.1-0.5 of the past to the future.


None of the market analysed had a static variance and consistent standard deviation, which indicate, Artificial intelligence trading software applied to the real market requires a lot of the user moderation in market predictions on a regular basis to enable the software to function profitably. Secondly, the effective use of Artificial trading software is perceived to be relevant for long-term traders, which concludes that the quality use of Algorithmic trading programs seems to be profitable for commodities and stocks market due to their regular primary trend formations. And noted to be a bad instrument for forex trading market due to the presence of more of tertiary trend character, if not used as a hedging target. The final conclusion suggests that to opt for Artificial Intelligent Trading software without a rigorous educational understanding of technical analysis, and market forecasting, there is a higher probability to crash an investment or loss of funds as a fund manager. Hence, the perception to spend huge financial resources to acquire the best algorithm trading software to save you from the knowledge burden of technical analysis is a fallacy.


Emmanuel Tweneboah Senzu, (Ph.D.)

Dean of Research, University College of Management Studies, Ghana.

Director of Research, Frederic Bastiat Institute

[email protected]


  1. Dysart, Jr., Paul (1967) Bear Market Signal; A sensitive breath Index has just flashed one. Barron’s News Paper. Sept 4th 
  2. Senzu, T. Emmanuel (2019) The Emotional Stress and Psychological discipline required of an automated trader.

About The Author

Related posts

Leave a Reply

Your email address will not be published.

Thank you for your upload