Core CPI Misses at 0.3%: Settling the September CPI Print — October Hike Odds Double Overnight, an MPF Bond Playbook for FOMC Week

Track 2 | The Data Trading Desk

Three days ago (12 October) our CPI preview set three scenarios, with monthly core inflation as the trigger: at or below 0.2% was Scenario A (dovish), at or above 0.3% was Scenario B (hawkish). The Bureau of Labor Statistics answered on 14 October: 0.3% — Scenario B triggered. This is the settlement piece: actuals versus forecasts, the market’s instant repricing, and an execution blueprint for the bond sleeve of your MPF portfolio into FOMC week.

Forecast vs actual: four numbers, three misses

(Forecasts: Dow Jones/Reuters consensus; actuals: BLS release, 14 October.)

The composition matters. Petrol prices surged 9.1% month-on-month, accounting for more than half of the monthly increase, with the energy index up 5.2% — AFP linked it to the US–Iran conflict and the Strait of Hormuz disruption. But the Fed has a long record of looking through energy shocks: the real signal is core. Core ran at 0.3% m/m for two straight months and accelerated to 2.6% y/y, with services prices (airfares, hotels, medical care) broadening out — what one ING economist called broad-based inflation, not just an oil story.

The macro matrix: risk-off tilt, with the bond market hiking ahead of the Fed

The market’s 24-hour reaction was a complete hawkish repricing:

• CME FedWatch odds of a 27–28 October hike: ~22% → 47.1% (Reuters) — doubled overnight. A December hike sits at ~94%, with at least one hike before the 9 December decision fully priced.
• The 10-year Treasury yield touched 5.33% intraday, about 3bp shy of a 24-year high (FactSet), settling near 5.31% (+10bp); the 2-year rose to 4.93–4.95% (+14bp).
• Futures markets officially parted company with the Fed’s dot plot: futures price the policy rate at 4.50–4.75% by April 2027 (three more hikes), while the Fed’s September projections stop at 4.00–4.25%. The market is pricing roughly two more hikes than the Fed itself acknowledges.
• Citigroup capitulated: the last major bank still forecasting no further hikes this year now expects a 25bp December hike, a hold thereafter, and no cuts until June 2027.
• Safe havens sold off too: gold fell ~2% to around US$4,075–4,088 an ounce; the PHLX Semiconductor Index dropped 4.3%, the Nasdaq 1.9%, the S&P 500 1.1% (its first 1%+ down day in over 50 sessions); the dollar index hit an 18-month high of 102.75.

Note the same-day reversal: the tech selloff triggered haven buying that dragged the 10-year from 5.33% back to 5.189%, down on the day. The bond market ran both scripts — rate-hike panic and recession haven bid — within a single session. That is a preview of FOMC-week volatility.

Backtest and win rate: the bond sleeve is back at the 2022 cliff edge

MPF bond funds are already this year’s worst asset class. MPF Ratings data (via the South China Morning Post, 7 October): global bond funds -3%, Hong Kong-dollar bond funds -2.5%, Asia bond funds -2.3% over the first nine months — dead last across the board; the MPFA’s end-September provisional data has bond funds at -1.5% over the past 12 months, also last. September’s best performer was money market funds (+0.2%) — up 3.4% over the past 12 months per MPFA data.

The precedent is 2022: the BEA Global Bond Fund fell 13.57%, HSBC’s Age 65 Plus Fund 13.21%, Manulife’s Hong Kong Bond Fund 9.46%. The Fidelity Hong Kong Bond Fund runs about 3.8 years of duration — every 100bp move higher in yields costs roughly 3.8% of NAV. On the average MPF balance of HK$338,950 (MPF Ratings, September), a 20% bond sleeve of HK$67,790 loses about HK$2,576 on a single 100bp move — roughly six weeks of contributions.

The point is win rate, not prediction: we are not betting on whether 27–28 October delivers a hike (47% is itself the answer — a coin flip). We are managing volatility asymmetry around FOMC week. Holding duration risk as yields accelerate off 24-year highs is negatively asymmetric; money market funds — up 3.4% over 12 months and September’s top category — are the natural tactical parking bay.

The switching blueprint: four moves for Scenario B

The preview’s Scenario B blueprint (core at or above 0.3%) read: “delay switches until after the FOMC, tactically park 20–30% of the bond sleeve.” Settlement confirms the trigger. Execution:

1. Tactically park 20–30% of the bond sleeve: move two to three tenths of global/HKD bond fund holdings into the same scheme’s money market or conservative fund. This is not calling the bottom in bonds — it is swapping duration risk for the +3.4% (past 12 months) certainty of carry, to be reassessed once the FOMC dust settles. On a HK$67,790 bond sleeve, 25% parked is about HK$16,948 — dodging one 100bp shock saves roughly HK$644, and the parked cash still earns interest.
2. Leave the equity sleeve on discipline: a -4.3% day in semiconductors looks scary, but it is the concentration risk our AI-concentration piece already quantified (the TSMC-plus-Samsung double-giant exposure), not new information. The behaviour tax is 1.2 percentage points a year (Morningstar’s Mind the Gap 2026) — don’t pay it on the back of one CPI settlement.
3. Age 65 Plus / DIS derisking members, watch the bond leg: the Age 65 Plus Fund is roughly 80% bonds and suffers in the same yield surge. Members being mechanically pushed ~6.7pp a year into bonds by the statutory derisking schedule should not exit the DIS (never manually override the statutory mechanism) — but should understand their portfolio’s rate sensitivity is rising passively.
4. Don’t sell bond funds at the trough: global bond funds are already -3% this year; liquidating now repeats the 2022 error (selling after -13.57%, missing the coupon repair that followed). Tactical parking reduces sensitivity; it is not liquidation.

Forward-pricing defence: the FOMC-week exclusion calendar

MPF fund switching uses T+1/T+2 forward pricing — you place the order without knowing the dealing price, so FOMC-week volatility becomes your dealing noise. The discipline:

• Exclusion zone: no switching instructions 26–29 October (FOMC meets 27–28 October). CPI day already demonstrated ±10bp intraday yield swings — under forward pricing, that is your bid-ask spread.
• Tonight (15 October): PPI plus retail sales — the last pre-FOMC puzzle pieces; markets will reprice again on the data, so split large switches into two batches on separate days.
• The calendar ahead: October CPI on 10 November → FOMC 8–9 December (Citi’s hike call) → November CPI on 10 December (the morning after the decision). The MSCI semi-annual review effective week (end-November close) is a second exclusion zone for switches.
• A single ±2% timing error on the HK$338,950 average balance is about HK$6,800 — roughly eighteen months of management fees at the market-average FER of 1.36%. Forward pricing is not free.

The one-line takeaway: core inflation at 0.3% moved “will they hike again” from tail risk to base case (94% for December), and the bond market has already hiked on the Fed’s behalf at 5.33%. The best defence for an MPF bond sleeve is not predicting the FOMC outcome — it is parking 20–30% in a money market fund, standing down through FOMC week, and letting volatility finish its run.