The decision cleared. The path did not.
Last week's useful distinction for automated traders was not "hawkish or dovish." It was event risk versus path risk.
This post is educational. It is not financial advice and it does not promise returns.
A priced hike can still leave the book wrong
On 16 September the FOMC raised the federal funds target by 25 basis points to 3-3/4 to 4 percent on a 12–0 vote. The statement said economic activity was expanding at a solid pace, inflation remained elevated, and the move would support a "timelier" return to 2 percent.
That binary was largely in the futures strip before the press conference. What was not finished was the continuous question: how many more times, how soon, and through which data.
For an algo stack, that split matters. Many systems treat FOMC day as a special state — wider stops, reduced size, paused entries — then restore "normal" permissions once the statement is out. If the model of risk was the announcement itself, those permissions come back too early.
Rate vol answered the event. Equity vol did not agree for long.
Saxo's 17 September options brief put the post-decision split cleanly: rate volatility compressed once the hike was delivered, while equity volatility firmed. On the day of the decision, FRED shows the VIX closing at 17.71. By 18 September it had fallen to 14.81.
That sequence is easy to misread as "event passed, risk off the table." Front-month equity vol can cheapen because the known meeting is behind you. The term structure and the rates market can still be pricing a harder path. Cheap near-dated implied move is not the same as low path uncertainty.
If your risk layer uses recent realized vol or a soft VIX threshold as permission to add risk, you can scale up exactly when duration, FX, and equity discount rates are still being rewritten by the policy path.
Wednesday reopened the next meeting. Thursday held the pressure.
The reminder arrived on 23 September.
S&P Global's flash U.S. Composite PMI rose to 58.4 from 56.0, a 62-month high and the fastest business-activity reading since July 2021. Manufacturing jumped to 57.0. Growth was strong, and survey commentary pointed to supply bottlenecks and rising costs alongside the boom.
The same day, Fed Governor Michael Barr told a Chicago housing conference that, in his base case, "further policy adjustments are likely to be needed" to bring inflation down in a timely way. CNBC put CME FedWatch odds of an October 27–28 hike in the low-to-mid 70% range as the session progressed, with the two-year yield climbing more than 13 basis points toward about 4.9%.
CNN and related coverage put the 10-year near 5.10%–5.12%, the highest since 2007. Equities sold off, with rate-sensitive tech leading the decline. Reuters also flagged weak demand at a $70 billion 5-year note auction that cleared at the highest auction yield since 2007 — another sign that the path, not the prior week's vote, was pricing the book.
On 24 September the pressure did not fade just because the PMI print was a day old. CNBC reported the 10-year still near 5.12% early Thursday, the 2-year near 4.90%, and FedWatch still around a 70% chance of another October hike, with oil and the growth narrative keeping the strip firm.
That is path risk across sessions: no new FOMC statement, no surprise vote. A growth print, a policymaker speech, and an auction cleared into a live rate path — and automation that had already flipped back to "post-event normal" was trading that path whether it knew it or not.
Durable lesson
Treat policy events as two clocks.
The event clock is discrete. It ends when the decision and press conference are out. The path clock is continuous. It keeps ticking through PMI, payrolls, energy prices, auctions, and every Fed speak that rewrites fed funds futures.
For operators running automated risk, the practical questions are mechanical:
- After a hike that was already priced, which strategies regain full size, and which stay on path-aware limits until the next meeting odds stop moving violently?
- Does your risk model use cheap front-end equity vol as a green light while the 2-year and FedWatch strip are still repricing?
- When yields gap higher on a data day, do correlated systems (rates, FX, equities, copy paths) hit independent caps, or do they all lean on the same stale "post-event" flag?
- Is October 27–28 already marked as an elevated state in the control plane, or only as a calendar note?
The worst post-FOMC failures are not usually wrong calls on the day of the decision. They are normal-state automation running while the path is still live.
Soft next step
If a rates shock changed how much risk you want mirrored across accounts, the free copy ratio and lot size calculators can help with the arithmetic. They do not replace broker specs, exchange rules, or a kill switch.