Although this is purely a matter of terminology -- the underlying logic of the argument is correct -- the positive feedback loops have not become "negative" feedback loops. A positive feedback loop is one in which the feedback from a change promotes even greater change in the same direction. In a negative feedback loop, the feedback from a change acts to oppose the change. Thus self-reinforcing phenomena -- spirals up and spirals down -- are due to positive feedback loops.

In general, positive feedback is destabilizing, negative feedback stabilizing, and it is important to keep the terminology straight (negative feedback loops in particular have been one of the most basic and important tools of engineering ever since the earliest days of the industrial revolution).

But it's also important to know (as control engineers have for more than a hundred years), that even nominally negative feedback can become destabilizing if improperly used. Because there is always some lag around a feedback loop, and at some frequency that lag will equal half the period of the sine wave of that frequency (a 180-degree phase shift), the effect will be that the nominally negative feedback becomes positive feedback at that frequency, and if the gain (amplification ratio) around the loop is greater than one at that frequency, the loop will break into oscillation, with the oscillations increasing in amplitude until some external constraint limits them, or they destroy the system.

To prevent such instabilities, control engineers always "tune" them with frequency-sensitive elements to get the loop gain down well below one at the 180-degree phase shift frequency (if it's close to one, the system will "ring" whenever it's disturbed).

So in some cases the instabilities we are seeing are not the result of positive feedback loops -- or at least not solely of positive feedback loops -- but rather of negative feedback loops (a classical market is a negative-feedback regulated system) that haven't been properly tuned. In an engineering sense, government regulation (if properly done) the equivalent of loop tuning, and an unregulated market is an untuned feedback loop -- which will be violently unstable unless by luck its loop gain happens to fall below one at the 180-degeree phase shift frequency.

Many of the unstable markets we see are due to interacting negative feedback and positive feedback loops.