My earliest attempt at active investing, carried out during the summer-break of my second year of Bachelor’s studies in physics.
I used the MetaTrader environment (mainly exploiting the MQL5 language, which is free and based on C++) for developing algorithmic trading strategies based on technical analysis, in particular price patterns. The strategies were subsequently back-tested for some in-sample data and optimized using parametric sweeps and genetic algorithms.
Some (out-of-sample) profitable systems went live for a little less than a year (achieved 1.23 Sharpe during the period). There has been no further development after that, due to the resumption of the studies and work, plus the quick realization that the combined effects of high fees, slippage and data quality (all factors limited by my capital at the time) were not able to surpass traditional passive investing returns.
In the attachment are present a couple of simple trading strategies, plus a sample of the pattern library I used for futures on equity indices (mainly DAX and E-mini S&P 500).