2019 Summary Report The average cumulative gains and losses for AI agents across all currency pairs are positive!

This is the overall 2019 results of the “MAiMATE”, a forex trading AI that grows, released on August 14, 2019 by Invast Securities Inc. (Head office: Minato-ku, Tokyo; President: Takeshi Kawaji).
In addition, this report only extracts the “forward test results after the birth of the AI agent”.
MAiMATE is a free foreign Exchange signal delivery service available to customers who have opened an account with Invast Securities.

MAiMATE AI Agents Overall 2019 Year-to-date Results(August 14, 2019 – December 31, 2019)

As a result of the steady accumulation of profits through continuous learning by each AI agent, the MAiMATE service as a whole earned nearly 400,000 pips of profit! The number of registered AI agents also grew steadily to over 3,000, with the average profit and loss per AI agent rising to +129.4 pips.


(インヴァスト証券作成)

Final number of AI agents in 2019 3,013
Cumulative profit and loss for the period +389,803 pips
Average cumulative profit or loss per AI agent (individual) +129.4 pips (=389,803/3,013)

Only one AI Agent can be created for each registered customer.
The accumulated profit and loss is the cumulative total of the settlement profit and loss during the period.
An AI Agent can only hold a maximum of one position.
The pseudo-trade results are based on the rate of “Tri-Auto FX” of Invast Securities.

  • An “AI Agent” is a FX trading AI character created for each client based on four choices chosen by the client.
  • Each AI agent tries to understand the “trade to aim for” by being evaluated by the customers it creates, and through continuous learning after it is born, it grows into an AI agent that you like.

2019 Year-to-date results by currency pair(August 14, 2019 to December 31, 2019)

USD/JPY” struggled to find profit opportunities toward the end of 2019 with no sense of direction, but ended with accumulated profits maintained. EUR/USD” struggled in the September to mid-October 2019 period, when there was a lot of daily price swings, but finished with positive gains by successfully capturing the emergence of subsequent trends. The “GBP/USD” successfully navigated the strong market fluctuations related to Brexit and ended with a high AI agent individual average profit of +274.3 pips.

In addition, around 70% of AI agents across all currency pairs ended 2019 with a positive profit, suggesting that the overall MAiMATE service has been able to contribute to the acquisition of stable revenue opportunities.

Customers choose their own currency pair when creating an AI Agent, and of the three currency pairs, “USD/JPY” is selected by far the largest percentage, accounting for more than 70% of all currency pairs. Therefore, the number of AI agents in USD/JPY is high, and the absolute value of cumulative profit and loss pips is likely to be large.
At present, MAiMATE offers one currency pair out of three currency pairs.

(Prepared by Invast Securities)

currency pair Final number of AI agents in 2019 Cumulative profit or loss for the period (all) 平均累計損益(個別) 2019年をプラス利益で終了したAIエージェントの割合
USD/JPY 2,169 +266,771 pips +123.0 pips 69%
EUR/USD 517 +33,330 pips +64.5 pips 69%
GBP/USD 327 +89,702 pips +274.3 pips 71%

Only one AI Agent can be created for each registered customer.
The accumulated profit and loss is the cumulative total of the settlement profit and loss during the period.
An AI Agent can only hold a maximum of one position.
Pseudo-trade results are based on Invast’s Tri-Auto FX rates.

In MAiMATE, when creating an AI agent, the customer (master) is asked to provide the following information: 1. currency pairs, 2. technical indicators (moving averages, Bollinger Bands, RSI), 3. news (market conditions, market analysts’ market predictions, emotional expressions on the Internet, external factors such as regulations, etc.) to be used as references ) 4. Choose an investment policy (risk-averse, return-oriented, balanced).
You check the trading of the born AI Agent and educate the AI Agent with “praise” and “scolding” once a week to tell the AI Agent what you want it to do, and then develop the AI Agent to your liking through continuous learning after it is born.