Inflation Calms, Global Liquidity Does Not: Why June’s PCE Print Was a Distraction

The US PCE deflator printed negative in June. Core PCE rose 0.1% month-over-month, the softest reading since 2021. The textbook response is immediate: the Fed has room to cut, risk assets rally, the cycle turns. The market did not behave. It priced a September cut without conviction, then traded sideways on every headline since. That divergence is not trader irrationality. It is a signal that the inflation print was never the operative variable. The Bitunix analyst framing this month gets the structure right: the source of pressure on global assets is not US CPI. It is the jointly restrictive financial environment engineered across the G3 central banks. I ran the full teardown, checked the data flows against my own macro risk models, and the conclusion holds. Rate cuts without global liquidity expansion are theater. The code was solid; the logic was not. Context: everyone is looking at the wrong dashboard. Since 2022, the market has been trained to read the Fed as the single point of failure. Fed hawkish, assets suffer. Fed dovish, assets recover. That model broke in August 2024, when the Bank of Japan’s policy drift triggered a global carry-trade unwind that a dovish Fed could not cushion. The Bitunix report correctly identifies the shift: we have moved from a unipolar monetary regime to a multipolar one, and the marginal price-setter of global liquidity is no longer Washington. It is Tokyo. Japan’s central bank held rates in July, but the internal vote count revealed a real faction pushing for hikes. The Bank of England has its own hawks. Meanwhile, Japan and Korea have both been caught intervening in FX markets, selling dollars to defend their currencies, a hidden channel of tightening that shows up nowhere in the standard macro calendar. Core teardown, channel one: rate divergence is not easing. The market keeps modeling a Fed cut as a global liquidity event. The math suggests otherwise. If the Fed cuts 25 basis points while the Bank of Japan hikes and the Bank of England stays on hold, the global weighted average policy rate barely moves. Worse, the dollar carry trade means Japanese yen funding flows into higher-yielding assets. When the BOJ tightens, those flows reverse. We saw the consequence on August 5, 2024: a violent repricing of risk assets driven not by US data but by yen-funded position unwinding. Volatility hides in the compounding fractions. The carry trade is a leverage stack built on a yield differential of a few hundred basis points. The funding leg moves, the stack collapses. Channel two: FX intervention as unannounced tightening. When the Japanese Ministry of Finance sells dollars to support the yen, it drains dollar liquidity from the system. Every intervention is a sterilized quantitative tightening event aimed at the currency pair. The Bitunix report flags this as evidence of coordinated intent between Japan, Korea, and tacit US tolerance. That is the right read. But the deeper implication is that intervention is finite ammunition. Foreign reserves are a stock; speculative flows are a flow. The stock depletes; the flow continues. At some point, Japan must choose between continued reserve drawdown and actual policy action. The market knows this. It is why the yen weakens into every intervention and why the dollar strength persists. Check the inputs, ignore the hype. Channel three: communication as a policy tool. This is the least understood element of the current regime. Central banks are not managing inflation data. They are managing inflation expectations and policy credibility. The Fed cannot cut aggressively the month after a positive core PCE print without signaling that it is politically captured. The BOJ cannot hint at exiting negative rates and then reverse. The market is not pricing the rates; it is pricing the volatility of the policy path itself. That is why a negative PCE print produced a muted rally. The data point met the expectation, but the expectation was always secondary to the credibility signal. Silence in the logs speaks louder than bugs. When the Fed goes quiet on balance sheet runoff while signaling cuts, that is a bigger event than any single CPI release. Now the part that most commentators miss: the AI investment cycle is the only genuine anchor holding up risk asset valuations. AWS revenue surprised to the upside. Oracle expanded its cloud partnership with Google. OpenAI is cutting prices. On its face, this suggests the technology cycle is in early expansion, not late bloat. The Bitunix analysis treats AI capex as a support factor for earnings, and that is fair. But from my audit work in the crypto trenches, I have learned that capital expenditure cycles and liquidity cycles are correlated, not causal. When the global financial environment tightens, IT budgets shrink. AI investment is not counter-cyclical. Companies do not commit hundreds of billions in capex during a contraction. If the current tightening persists into Q4, the AI infrastructure order book will compress. The bull case is correct that AI earnings are real today. The error is assuming that today’s growth rate is the base rate. Contrarian take: the bulls have one structural point that the hawks are wrong about. GDP growth in the US decelerated, but the composition was resilient. Private final demand remained firm. Consumer spending held. Government spending and inventory noise created most of the drag. This is not a recessionary profile; it is a rebalancing profile. The Bitunix analyst is right to flag the tension between top-line weakness and underlying strength. If the economy is genuinely rebalancing rather than rolling over, then the Fed has room to cut without triggering a hard landing, and the AI cycle can keep compounding. That scenario is possible. The market is pricing a soft landing, and there is real data supporting it. The flaw is not the conclusion. The flaw is the timeframe. Soft landings are judgment calls that only historians get right. In real time, they look like recessions that never arrive until they do. There is a second layer of the contrarian argument that matters more. Central banks are not a permanent alliance. The Bitunix report describes coordinated tightening. That coordination is conditional, not structural. When the financial stability break occurs, as it did in early August 2024 with the carry trade unwinding, the central banks flipped from tightening to smoothing almost overnight. The Bank of Japan stepped back from hawkish messaging. The Fed signaled flexibility. The so-called tightening alliance is one crisis away from a coordinated easing pivot. The risk of that pivot is not that it fails. It is that it arrives too late, after the liquidity damage has already propagated through leveraged balance sheets. The lesson from Terra and from the August unwind is the same: you do not get a warning when the structural trade breaks. You get the consequence. A flat line is more dangerous than a spike. Takeaway: the market will keep trading the Fed’s September decision as if it settles the macro question. It will not. The variable that matters is the collective stance of the Fed, the BOJ, and the currency intervention desks in Asia. The September BOJ meeting is the real event. If they hold, the yen keeps bleeding out, and dollar strength squeezes emerging markets. If they hike, the carry unwind resumes and global volatility reprices upward. The inflation print was a distraction. The quarter ends when Tokyo speaks. Trust the compiler, verify the intent. Set your models to monitor the yen, the intervention data, and the AI capex guidance revisions. Those three inputs will determine whether this market grinds higher or breaks lower. Do not watch the PCE press release. Watch the flows behind it.