From the Mind Over Market article:
Markets are tools that, relative to the alternatives, happen to have great strengths with respect to incentives, efficiency, and innovation. But they are not perfect...
Decentralized Networks of Increasing Complexity
We live in a world of largely decentralized networks of increasing complexity: electronic networks, networks of supply chains and trade, financial networks that link the balance sheets of disparate entities. Market incentives cause actors to operate or modify parts of the network in ways that maximize efficiency locally. But the presumption – often an article of faith – that the whole remains stable and resilient has no theoretical or empirical support. Indeed, it seems inaccurate.
Resilience is a public good, created by the right kind of redundancy.
In a decentralized structure, redundancy tends to be undersupplied in the process of local optimization. That is why the tsunami that hit Japan last year disrupted many global supply chains: they were (and still are) too efficient from the standpoint of withstanding shocks.
In financial markets, local optimization seems to lead to excessive leverage and other forms of risk-taking that undermine the stability of the system. Much research is needed to understand which interventions or restrictions on individual choice are needed to make certain kinds of market equilibria stable. But, clearly, markets do not do this well by themselves.
The risks associated with the scale, complexity, and interconnectedness of the financial system are clearly not well enough understood. A paper by James Rickards expands a bit on what Professor Spence says above. An excerpt from the paper:
...there can be little doubt that the current period of globalization from 1989–2009, beginning with the fall of the SovietUnion and the end of the Cold War, represents the highest degree of interconnectedness of the global system of finance, capital, and banking the world has ever seen. Despite obvious advantages in terms of global capital mobility facilitating productivity and the utilization of labor on an unprecedented scale, there are hidden dangers and second-order costs embedded in the sheer scale and complexity of the system. These costs have begun to be realized in the financial crisis that began in late 2007 and have continued until this writing and will continue beyond. - James Rickards
I'd say it's more than fair to say the financial system needs to be reigned in somewhat. Changes are sorely needed to reduce systemic risk. Though in itself an insufficient solution, less leverage and more limits on the use of derivatives seems a good place to start. Derivatives in their current form are, of course, a huge source of hidden leverage and hard to measure risk.
One beneficial circuit breaker is a clearinghouse of some kind for over-the-counter derivatives.
Clearing houses add needed transparency (increasingly banks are clearing credit and interest rate derivatives through central counterparties). Also, the kind of intermediation that a clearing house provides by its nature can reduce systemic risk substantially. Those that object to a clearing house claim nonuniformity of contracts make this impossible. That excuse seems like, well, an excuse. Contracts can be modified to create uniformity.
Many other safeguards and limits are almost certainly needed to assure system stability and it's better to do so before we're in the next crisis. I'm guessing, in time, we'll learn, one way or another, much more has to be done to protect the system against the risk of instability and collapse.
I've yet to read it but, in his book Currency Wars, my understanding is that Jim Rickards provides a useful foundation for understanding complex systems and he suggests what we should do proactively to prevent the next financial crisis. The paper by Rickards is a good way to get some background on his thinking.
Ultimately, a system this large and complex when under stress (what Rickards describes in the paper as the supercritical state) cannot be expected to behave predictably in any meaningful way.
Rickards makes that point very clear.
More limits and safeguards that protect the system (not the participants) seem inevitable. The question is whether we'll make sure this gets done proactively or not.
Rickards explains why complex systems by their nature behave unpredictably and are prone to collapse. He also seems to think that regulators and bankers are still not using the right tools or metrics to assure system stability.
In any case, the time to put up the firewalls is before there's another fire.