Episode 10: Further Validation of Backtesting
BuilderOU STORY|Episode 10
Backtesting,
verified further
When you break down a single win rate in detail,
the differences hidden behind the numbers begin to reveal themselves.
Hello.
This is
Mr. Ojun, of BuilderOU
In Episode 9,
on the backtest
the win rate exceeded 60%,
but the person who actually used it
found that they were not winning.
Backtesting can pick up every signal that appeared in the past.
However, the people actually trading have lives.
Work. Sleep. Meals. Time with family. Time with a partner. Interactions with friends. Hobbies.
Furthermore, there are times when it is said to be better to avoid—such as around economic indicators or right after the market opens.
If you subtract all of those, the actual time available for trading is far more limited than you might imagine.
Therefore, I realized that not only should I backtest indicators, but I also needed to further validate the backtest itself.
1. Start by doubting the numbers
and questioning them
In Episode 9, it became clear that the figure of a 60% win rate was not a lie.
However, I did not simply believe that figure as is.