—
*(A continuation of Metastasis of Economic Frameworks Augment Systemic Failures, and, before that, Playing With Fire: Animal Spirits Ensconce Progress and Trepidation in Economies)*
When I wrote about the metastasis of economic frameworks, the claim was that central banks were fighting supersonic markets with subsonic tools. That framing was already generous. It assumed a pilot in the cockpit — a human institution, however slow, still steering. The uncomfortable revision is this: the tools were never just subsonic. They were built for a world where the thing being governed *waited for governance*. That world is now the exception, not the rule.
The Lag Was Always the Real Risk
Every dislocation of the last two decades has lived in the same gap: the distance between when a policy is written and when the system it targets has already moved somewhere else. 2008 was not a failure of models predicting risk. It was a failure of transmission — reserve multipliers built for a slower credit cycle, applied to a system already trading at the speed of milliseconds. The lag was not a bug in an otherwise sound architecture. The lag *was* the architecture.
What has changed since is not the size of the gap. It is who — or what — is standing on the other side of it.
From Subsonic Tools to No Tools At All
Policy transmission was always designed around a human tempo: a committee meets, a rate moves, banks reprice, households and firms adjust over quarters. That tempo assumed the counterparties on the other end were also operating on a human clock — traders reading statements, analysts building models overnight, capital reallocating over days.
Agentic capital does not read the statement. It reads the order flow generated by everyone else who read the statement, and it repositions before the ink is dry on the analyst note. This is not high-frequency trading dressed up in a new name. HFT exploited latency within a single, still-comprehensible market structure. What is emerging now is capital allocation performed by systems that assess narrative, sentiment, and cross-asset correlation continuously, without waiting for a quarter, a news cycle, or a human being to notice the pattern first.
The central bank is no longer slow relative to the market. It is slow relative to a market that has stopped waiting for anyone.
Reflexivity Without a Human in the Loop
The reflexive economy I described previously — where expectations shape the very conditions they are meant to merely observe — assumed reflexivity ran through human psychology: fear, greed, animal spirits, the crowd. That is still true. But a second reflexive loop now runs *underneath* it, machine to machine, at a speed no committee can observe in real time, let alone govern.
This is the paradox worth sitting with: institutions built their credibility on the premise that policy could *anchor* expectations. Anchoring requires the anchored party to notice the anchor. When the dominant marginal actors in price formation are agents optimising on a timescale beneath human perception, the anchor is still being dropped — it is just landing in water nobody is swimming in anymore.
Policy as Theatre, Not Transmission
None of this means policy stops working entirely. It means policy increasingly functions as *signal for the humans left in the loop* — retail sentiment, longer-duration allocators, the institutions still operating on quarterly cycles — while the marginal price-setting activity has already moved on before the signal finishes propagating. Policy becomes theatre performed for an audience that is no longer the majority of the room.
This is not conspiracy. It is arithmetic. If the majority of daily price discovery is increasingly mediated by systems reacting in milliseconds to information most humans haven’t processed yet, then “the market’s reaction to policy” and “policy’s actual effect on the real economy” are no longer the same measurement. We have been reading the theatre and calling it the transmission mechanism.
What This Means for the People Paying Attention
If you take this seriously, the practical implication is not despair — it is a reallocation of *where you look for signal*. Three shifts worth making:
1. **Stop reading policy as prediction. Read it as lagging confirmation.** By the time a rate decision is public, the agentic layer has already repriced the probability distribution several times over. The decision confirms a direction; it rarely sets one.
2. **Track the infrastructure, not just the instrument.** The systems doing this reallocation need verified, trustworthy inputs to act on — which is precisely why the next competitive edge in markets is not a faster trade, but a more *provably true* one. Trust becomes the scarce resource once speed is commoditised.
3. **Treat volatility clustering as a structural signal, not noise.** When agentic reflexivity compounds on top of human reflexivity, volatility doesn’t just increase — it clusters differently, in patterns closer to coordinated cascades than the old normal distribution assumptions most retail risk models still quietly rely on.
The Closing Paradox
Central banks were built to be the adults in the room, slowing down a system that might otherwise overheat. That mandate assumed the room would wait for the adults to speak. The room no longer waits — it has already voted, repriced, and moved on to the next hypothesis before the adults have finished clearing their throat.
Animal spirits will always exist. I said that before, and it remains true. What is new is that they now have a machine-speed nervous system running underneath them — one that does not sleep, does not wait for committees, and does not care whether the institutions built to govern it have caught up yet.
They haven’t. The question, as before, is whether we build something that can. Remember, monetary policy is a blunt tool as I said years ago!