Well, Minsky's financial instability hypothesis is at least interesting. His argument is, basically, that there's a cycle wherein: during normal times, people want to beat the averages, so they engage in more speculative bets; as that ratchets up, and speculative positions become increasingly leveraged, a point comes where debt is financing interest on speculative leverage; once enough people get suspicious, further debt isn't extended, so the speculative positions default and drive cascading defaults (since there's "blood in the water"); the after-crash period comes with a renewed sense of caution, and people accept lower yields as the price of earlier speculative frenzy; over time, the caution comes to seem outmoded and people believe that normal times are here again; and so on.
I'm not sure how I feel about this, but it's at least a perspective.
Basic differential equations: Any system with an effect proportional to the negative second derivative in it will tend to have cyclic behavior. That's a bit oversimplified, but in practice it often works out that way. And there's plenty of such things in the economy, so cycles in the economy are inevitable.
You can make solid cases that we don't need to have quite the crashes we actually do. But I don't believe in the existence of a real economy that doesn't have some substantial cycles in it that people will point to as evidence that something is wrong.
(Of course, if enough of those people get together and get put in charge of an economy, they often are successful at removing the cycles, by virtue of removing all instances of the economy going up at all....)
I'm not sure how I feel about this, but it's at least a perspective.