Core PCE Runs Hot as Fed's Repo Buffer Hits Zero

Core CPI cooled to 2.47% year over year in July, the kind of number that would normally clear a path for the Federal Reserve to declare victory on inflation. Core PCE, the gauge the Fed actually targets, tells a different story: it held at 3.29% year over year in June, more than 80 basis points hotter than its CPI cousin. That gap is the story this week, and it is widening at the exact moment the Fed's money-market shock absorber has gone quiet. our Fed policy framework

Start with why the CPI-PCE gap matters. The Fed has said for years that PCE, not CPI, drives its reaction function, largely because PCE adjusts for substitution effects and weights housing and healthcare differently. When the two measures diverge by 80 basis points, it usually means shelter and services costs are still running hot inside the PCE basket even as CPI's goods components disinflate.

Related reading: The Week Ahead: PCE, Jackson Hole and NVIDIA Test the Rally, Fed Minutes Show 9-3 Split as Treasury Steps In, Bitcoin Tops $75,000 on $3B Squeeze, Treasury Liquidity.

Anyone reading only the CPI headline this week is underestimating the inflation the Fed is actually watching. Headline CPI sat at 3.30% year over year in July and headline PCE ran even hotter at 3.67% in June, so the divergence is not an isolated core-measure quirk.

Layer onto that a plumbing story that deserves more attention than it is getting. The overnight reverse repo facility once absorbed more than $2 trillion in excess money-market cash at its 2022-23 peak. It has now drained to $0.2 billion, functionally empty. That buffer used to be the market's pressure valve: when the Treasury needed to rebuild its cash balance by issuing bills, dealers could draw down RRP balances to fund the purchases without touching bank reserves directly.

With RRP empty, the Treasury General Account's $124 billion rebuild since late July has nowhere to come from except bank reserves. The Fed's balance sheet has been roughly flat, at $6.746 trillion as of August 19 versus $6.760 trillion a week earlier, so there is no quantitative easing offset adding reserves back into the system.

This is structurally similar to the setup that preceded the September 2019 repo-rate spike, when reserve scarcity caught money markets off guard and forced the Fed to intervene. We do not have repo-spread data confirming stress yet, and the 10s2s curve remains positively sloped at plus 50 basis points, up from plus 48 the week before, which is not the shape markets typically show heading into acute funding stress. Treat the RRP-drain-plus-TGA-rebuild combination as a precondition worth watching, not a confirmed event.

There is a reasonable counterargument. RRP balances near zero can simply reflect money-market funds finding better yield in Treasury bills directly rather than parking cash at the Fed's overnight facility, a benign portfolio shift rather than a scarcity signal. Core CPI's deceleration to 2.47% is real, even if PCE has not caught up.

If next month's PCE print converges toward CPI's cooler pace, the divergence resolves on its own without a liquidity scare. I would want at least one more month of TGA and RRP data, plus confirmation from SOFR spreads, before treating this as more than a watch item.

The cross-asset read is where it gets uncomfortable. M2 reacceleration to 5.53% year over year alongside a fed funds rate holding at 3.63% implies a real policy rate that is deeply negative to flat against 3.29% core PCE inflation, an unusually loose stance for an economy where unemployment just fell to 4.1% from 4.4% in February. Loose real rates, a tightening labor market, and hot core PCE is not a combination the textbook associates with a Fed done hiking.

Equity markets are not pricing any of this as a risk: SPY, the S&P 500 ETF, closed near $765.72 and QQQ, the Nasdaq-100 ETF, near $713.44, both within roughly 2% of their 52-week highs. Risk assets are pricing calm liquidity conditions that the plumbing data does not fully support.

What to watch next: the September TGA and RRP prints will tell us whether the reserve drain is a one-quarter funding technicality or the start of something the Fed has to address with balance-sheet action. SOFR spreads are the earliest tell of actual stress, well before it shows up in equity or credit pricing. And the next core PCE print, due in late September, either narrows the CPI-PCE gap or confirms it, which shapes whether the Fed's September meeting leans toward acknowledging sticky services inflation or waving it off as a lag effect.

This content is for informational purposes only and does not constitute financial advice. Past performance is not indicative of future results. Consult a qualified financial advisor before making investment decisions.