I asked myself a simple question this week, if I look at the whole of 2027, which month has the highest risk? It came up while I was building a small calendar tool for TradeInsight (it publishes the US week ahead, Fed speeches, FOMC, CPI, jobs reports and earnings, on calendar.tradeinsight.info), the tool only looks one week ahead, so I wanted to zoom out and see the full year.
First, a disclaimer, nobody can tell you which month will crash, if someone tells you that with confidence, they are selling something. What we can do is look at the things that are already on the calendar, and what history says about each month, and then decide where to be a bit more careful.
The FOMC calendar for 2027
The Fed has already published its 2027 meeting dates, the ones with a star come with the Summary of Economic Projections (the famous dot plot), which is usually where the market gets surprised the most.
| Month | Dates |
|---|---|
| January | 26–27 |
| March | 16–17* |
| April | 27–28 |
| June | 8–9* |
| July | 27–28 |
| September | 14–15* |
| October | 26–27 |
| December | 7–8* |
Each meeting date is tentative until confirmed at the meeting immediately preceding it.
Why September
September 2027 stacks up a few things in a short window,
- FOMC on September 14–15, with projections, so the whole rate path can be repriced in one afternoon
- Quarterly triple witching on Friday September 17, options and futures expire on the same day, liquidity and volatility both get bigger around it
- September 30 is the end of the US federal fiscal year, if the FY2028 funding is not passed, there is a government shutdown risk from October 1
- And the seasonality, since 1950 September has been the worst month of the year for the S&P 500, the average return is negative (around -0.7% depending on who counts), and it is up less than half the time
September Almanac & Vital Stats: Worst Month of the Year 1950-2023
I don’t trust seasonality much on its own (it is an average of many very different years, a few big drops can pull the whole number down), but when it lines up with a dot plot meeting, a big expiry day and a funding deadline in the same two weeks, I think it is worth paying attention.
October is the runner-up
October is the month with the scary history, 1929, 1987, 2008, it is historically the most volatile month even though its average return is not bad. In 2027 it also has the Q3 earnings season and another FOMC (October 26–27, no projections this time), and if the September funding fight is not solved, a shutdown would run straight into October.
What about the rest of the year
March, June and December also have a dot plot meeting plus a triple witching in the same month, so they are not quiet either, but their seasonality is much better than September. And 2027 is the third year of the presidential cycle (the year before an election), which historically has been the strongest of the four years, so the base risk for the year looks lower, not higher. 🤔
There are also risks that don’t have a date, the debt ceiling, tariffs, geopolitics, a surprise in AI capex, none of these fit nicely into a month table, so they are not in my ranking at all, which is the biggest weakness of this kind of exercise.
What I am thinking for now is, it is probably cheaper to look at protection (or lower leverage) before mid September, than to keep paying for hedges the whole year, but I still need to check the CPI and jobs report dates for 2027 when they are out, my calendar tool only fetches one week at a time.