Fooled by randomness: 1. Your strategy worked in the past 1 1/2 years just by chance, or, 2. Your strategy is indeed working in the 1 1/2 years, but probability messed you up lately with a string of losses. If so, eventually it will turn to a string of wins. 3. The herd found out about your winning strategy and all crowded into it making it no longer effective. How do you know which is real? My thinking is 5 yrs backtest and 1 1/2 years walked forward are just not long enough periods for verifying it to be a winning strategy. If you believe you truly have an edge, you should have the conviction to stick it out a few more years to find out.
You are not giving enough information to really tell. But the few comments already mention the very obvious mistkes.
Which comes back to bad testing. If your software does not have any option to consider it, you should consider moving on to a software that does. But who really things thepast 5 years are representative for the fear/greed ratio in the market?