Hi all,

How important is performance recency in backtests? I tried googling this (in both a quant and statistical perspective) and was not able to find anything.

One backtest iteration shows significantly higher performance in the most recent years however
Another backtest shows a better performance overall

I'm leaning towards overall performance because it is unbiased, but I keep thinking that the strategy could be more fitting to the current times in this exponentially evolving trading world.

What do you guys think?