‘The System Has Become Over-Optimized’

In “Underestimating The Biggest Risk Of All,” I contextualized the Bank of England’s rate hike trial balloons by reference to markets which, over a dozen years, have hyper-optimized around the post-Lehman policy dynamic.

Everything is self-referential and assumes established behavioral patterns by all parties involved.

That’s a more generalized way of describing an interconnected set of granular dynamics, including the post-GFC muscle memory (predicated on expectations of implicit or explicit policymaker “puts”) which prompts dip-buyers and vol-sellers in equities to reengage on any sign of weakness or vol expansion.

When that behavior is consistently rewarded, the classical conditioning optimizes around itself. Over time, traders attempt to front-run one another’s Pavlovian response function. That, in turn, interacts with modern market structure.

“As [volatility] resets lower, it incentivizes monetization of hedges and resumption of short vol flows, which creates delta to buy and gets dealers back into a market-stabilizing ‘long Gamma’ and ‘long Delta’ position, as options are again sold from overwriters and strangle sellers,” Nomura’s Charlie McElligott wrote this week, describing the self-referential feedback loop. “As realized vol softens, systematic strategies are able to mechanically re-allocate on a lagged-basis, while simultaneously, the reversal in spot sees CTA trend forced to cover nascent shorts and begin to re-lever long exposures.”

Recently, I’ve characterized this as a kind of self-contained, deterministic setup. In his latest, Deutsche Bank’s Aleksandar Kocic captured it better than I’ve been able to.

“The mechanisms that have been governing the markets [over the past decade] have largely lost any external points of reference and can now be evaluated only in their own terms,” Kocic wrote, adding that “as these mechanisms continued to expand, increase in size and become more efficient, everything has been gradually converging towards those modes of functioning, so they gradually accounted for all of market realities and cannot be exchanged for anything else.”

At the heart of this is monetary policy. Kocic juxtaposed “fragmented and passionately local” politics with “the most important problems” which are “becom[ing] increasingly global.” He cited LTCM, the financial crisis and, now, the pandemic.

Monetary policy, he said, is mostly attentive to global issues, as central bank independence (or relative independence, anyway) leaves policymakers free to focus on the bigger picture, while “fiscal policy remains straddled between global economic needs and the passions of local politics.”

Consider how that creates and then perpetuates the hyper-optimized, self-referential system described above and in the article linked here at the outset. As Kocic wrote,

Monetary policies across different economies remain coordinated, not because of some internal agreement between different central banks, but due to commonality of their economies’ exposure to the same global problems. With each crisis, as the stakes are growing and central banks behavior and reaction functions are becoming more aligned, the market expectations and reaction to those policies are inevitably showing high levels of coordination.

The BoE is at risk of upsetting things, which is why I spent what may have otherwise appeared to be an inordinate amount of time this week editorializing around the price action in gilts and the prospect of a policy mistake in the UK.

Coordination at the front end is “a reflection of the common circumstances that force certain central banks’ actions to be aligned,” Kocic remarked, noting the breakdown of that coordination amid the prospect of preemptive hikes from the BoE, which the market believes might constitute a policy mistake (figure below, from Kocic’s note).

Preemptive hikes for risk management purposes are obviously not consistent with established post-financial crisis behavioral patterns.

Although expectations for rate hikes over time are building, the prospect of forceful, preemptive action is seen as remote outside of the UK. As Kocic put it, “this possibility of departure from ‘unconditional’ accommodation by the BoE was received by the US market as a remote possibility for the Fed, and priced in by the risk premia and vol market as a way of widening the rates distribution.”

But the BoE’s nods to a possible departure from the norm are notable indeed. Kocic went on to write that,

Deviation of one central bank from the expected behavior, is capable of creating potential anxiety and triggering ripples across different economies, however, not because it reflects changes in the global environment, but rather as a potential challenge of the established pattern of central banks behavior and the risk that success of any such challenge might cause.

A central tenet of the post-Lehman system would be challenged (in spirit at least) by a decision from a major, developed market central bank to deploy preemptive hikes to manage risk — any kind of risk, be it inflation, financial stability or otherwise.

“The system has become over-optimized, and it is just when all uncertainty disappears that it also reappears – when one mode of functioning dominates, because there is no alternative outside of it, stakes become the highest,” Kocic said. “Even a slightest possibility of Fed’s departure from its expected script is perceived as a source of tail risk.”