In his post on safe assets, Arpit Gupta references Engineering the Financial Crisis by Jeffrey Friedman and Wladimir Kraus, one of the most brilliant books I’ve read in years. At some point I hope to discuss the Friedman and Kraus book at length, as I think it deserves a proper essay.
For now, however, I want to draw your attention to a post by Ashwin Parameswaran that resonates with some of the themes Friedman and Kraus explore, though I don’t know if Ashwin has encountered their work. In “The Pathology of Stabilization in Complex Adaptive Systems,” Ashwin draws a parallel between macroeconomic stabilization and the increasing reliance on psychiatric medication to manage mood disorders, using the work of Robert Whitaker as his guide.
So we have these weird parallels: (a) financial crises have grown more frequent and more severe over time to the point where we now seem to be in a perpetual crisis, and we seem to need a bigger and bigger dose of monetary and fiscal stimulus to kickstart a recovery after each one; (b) the number of disabled mentally ill has increased dramatically at the same time that we’ve discovered these marvelous drugs that promise to “fix” chemical imbalances in the brain. Could it be that psychiatric drugs mitigate certain symptoms yet actually create dangerous imbalances that larger and larger doses can never truly fix? And could it be that our approach to macroeconomic stabilization has, in a similar vein, actually made matters worse?
One way of thinking about this is the idea of hormesis, the notion that a little bit of poison can actually be good for you. Living in extremely hygienic environments might actually make us more vulnerable to infection, etc., by not allowing us to develop sufficiently strong resistance.
The fundamental reason why interventions fail in complex adaptive systems is the adaptive response triggered by the intervention that subverts the aim of the intervention. Moreover once the system is artificially stabilised and system agents have adapted to this new stability, the system cannot cope with any abrupt withdrawal of the stabilising force.
Our systems grow more brittle, fragile, and crisis-prone the keener we are to guard against shocks and disruptions.
And as Ashwin explains, there are no bad guys:
Similarly, when a central bank protects incumbent banks against liquidity risk, the banks choose to hold progressively more illiquid portfolios. When central banks provide incumbent banks with cheap funding in times of crisis to prevent failure and creditor losses, the banks choose to take on more leverage. This is similar to what John Adams has termed the ‘risk thermostat’ – the system readjusts to get back to its preferred risk profile. The protection once provided is almost impossible to withdraw without causing systemic havoc as agents adapt to the new stabilised reality and lose the ability to survive in an unstabilised environment.
Of course, in economic systems when agents actively intend to arbitrage such commitments by central banks, it is simply a form of moral hazard. But such an adaptation can easily occur via the natural selective forces at work in an economy – those who fail to take advantage of the Greenspan/Bernanke put simply go bust or get fired. In our brain the adaptation simply reflects homeostatic mechanisms selected for by the process of evolution.
Ashwin’s post is well worth a close reading. He goes on to explain why throwing up our hands and accepting the logic of crisis after crisis and increasing stimulus each time will likely end in tears:
The structural malformation of the economic system due to the application of increasing levels of stimulus to the task of stabilisation means that the economy has lost the ability to generate the endogenous growth and innovation that it could before it was so actively stabilised. The system has now been homogenised and is entirely dependent upon constant stimulus. The phenomenon of ‘rapid cycling’ explains a phenomenon I noted in an earlier post which is the apparently schizophrenic nature of the markets, turning from risk-on to risk-off at the drop of a hat. It is the lack of diversity that causes this as the vast majority of agents change their behaviour based on absence or presence of stabilising interventions.
Moreover, it is not just craven financial elites that clamor for stabilization policies that invariably benefit craven financial elites: it is virtually everyone in the crisis-plagued societies, as financialization gives everyone a stake, direct or indirect, in the fate of asset prices, etc.
Interestingly, I noticed a parallel between Ashwin’s observations and a recent (brief) talk by John Cochrane of the Booth School:
For nearly 100 years we have tried to stop runs with government guarantees ‐‐ deposit insurance, generous lender of last resort, and bailouts. That stops runs, but leads to huge moral hazard. Giving a banker a bailout guarantee is like giving a teenager keys to the car and a case of whisky. So, we appoint regulators who are supposed to stop the banks from taking risks, in a hopeless arms race against smart MBAs, lawyers and lobbyists who try to get around the regulation, and though we allow – nay, we encourage and subsidize –expansion of run‐prone assets.
In Dodd‐Frank, the US simply doubled down our bets on this regime. The colossal failure of Europe’s regulators to deal with something so simple and transparent as looming sovereign risk hints how well it will work. (European banks have all along been allowed to hold sovereign debt at face value, with zero capital requirement. It’s perfectly safe, right?)
The guarantee – regulate ‐ bailout regime ends eventually, when the needed bailouts exceed governments’ fiscal resources. That’s where Europe is now.
And the US is not immune. Sooner or later markets will question the tens of trillions of our government’s guarantees, on top of already unsustainable deficits.
What financial system will we reconstruct from the ashes? The only possible answer seems to me, to go back to the beginning. We’ll have to reconstruct a financial system purged of run‐prone assets , and the pretense that nobody holds risk. Don’t subsidize short‐term debt with a tax shield and regulatory preference; tax it; or ban it for anything close to “too big to fail.” Fix the contractual flaws that make shadow‐bank liabilities prone to runs.
Here we are in a golden moment, because technology can circumvent the standard objections.
I imagine that Ashwin wouldn’t embrace Professor Cochrane’s account in all of its particulars, but he does seem attuned to some of the same flaws of the permanent stabilization regime.