For security reasons, you should first sign your message and then encrypt
it with the recipient's public key. If you
perform the the reverse operation (encrypt then sign), then a threat
agent may
intercept you message, skip your signature and sign "your encrypted"
message. So the recipient will hence receive a
signed message from the threat agent and no more from you.
And how can the threat agent do this using the expected private key? And if
it can't, wouldn't the receving agent have a serious security hole if it
accepted such a message?
Curiously yours.