>Unless we know that deaths from the virus are greater than the deaths caused by the second-and-third-order effects of panicking

Do you believe that when the death toll raises from non-panicking and not imposing a lockdown (not to mention the total saturation of the healthcare system), that wont create even more panic and self-lockdown -- and have their own second and third order effects?

Yes, and this is why - the population of investors is honestly not all that panicky. If people start dying, investors will watch the death rate, and companies' share prices will reflect some instability but it will be somewhat in line with the actual economic impact, which traders are smart enough to price it at the small level it would actually be given the mortality rate.

By imposing lockdown, we have created a real economic slowdown, that traders are smart enough to deem catastrophic (because a month or restauranteurs going without wages is actually catastrophic and represents a much larger loss of future incomes that the still-small-but-blown-up-in-the-media death rate from the unchecked coronavirus.) Unlike the non-lockdown option, there is no optimistic case. You can't say, "earnings are slow, but people will be sorry if they panic sell because when the media frenzy dies down things will settle." You have to say, "earnings are slow, and might not actually pick up again because the lockdown lost a bunch of people their job."