Let's say I have designed a terrible trading strategy with a poor win rate of 20% or less. We are not risking more than 1% in a single trade and an excellent risk management and position sizing are in place. Can we flip it?
As per quantitative books on trading by good authors - We should not build accidental model/approaches.
However, I feel this is okay as it's contrarian?
"Win rate" is not a sufficient metric -- the size of gains and losses matters. A strategy with a win rate of 80% can still be bad if the losses are large. A simple example is writing naked out-of-the-money options ("picking up nickels in front of a steamroller") -- eventually you will go broke.
You say "we are not risking more than 1% in a single trade" but it's doubtful this can be maintained when flipping. If these are long stock positions, they turn into short stock positions, with unlimited risk (even with stop-losses, you could face a gap-up).
The opposite of a bad strategy is not necessarily a good strategy! The basic reason is that taking the opposite (or generally, using any chosen leverage) works arithmetically whereas returns compound geometrically. If a strategy can deliver 100% gains or 100% losses, the 100% losses will bankrupt you. If you flip it, you still have a strategy with 100% gains or 100% losses and it will still bankrupt you.
Here is an evaluation metric for whether a strategy is useful, taking into account that you can use it with various weights, including "flipped". In an efficient market, no strategy other than buy-and-hold is useful in either direction.
