That's very true, a strong trend works best. You'll see on the KMD example I posted that a 'fake-out' in the Aug/Oct crosses would have cost at least the brokerage fees, which would be material for the big buyers/sellers, if not a capital loss in that particular case. That's why I suggested 'confirmation', meaning for example the RSI, Money Flow, %R and various other indicators (depending on speed of indication) really should be well into the overbought/sold zones.
Anyway, I'm digressing from the action on KMD. What a mess! Looking forward to the fundamentals guys trying to figure out 'fair value'. Assuming this doesn't kill KMD, there will be some good buying in due course.