The rules
Lucca and Moench (2015): buy the S&P 500 at 2:00 pm ET on the trading day before each scheduled announcement by the Federal Open Market Committee (FOMC), the Federal Reserve committee that sets US interest rates, and sell fifteen minutes before the announcement; hold cash otherwise. Long only, fully invested in the window, unscheduled meetings excluded. Statements came at about 2:15 pm in their 1994-2011 sample; since March 2013 every scheduled statement is released at 2:00 pm, so fifteen minutes before is now 1:45 pm. The paper also reports windows from the prior day's close to 2 pm.
- Universe
- SPY
- Bars
- 1-minute
- Position size
- 97% of equity in SPY during each window, cash otherwise. Published: fully in the index during the window, cash otherwise.
- Rebalance
- Eight times a year, from 14:00 on each meeting's first day to 13:45 on announcement day. Published: from 14:00 the day before to fifteen minutes before the announcement.
What the research found
Lucca and Moench (Journal of Finance 2015; New York Fed Staff Report 512) studied S&P 500 returns above Treasury bills around 131 scheduled FOMC meetings from September 1994 to March 2011. The drift, the return in the 24 hours to 2 pm on announcement day, averaged a statistically significant 49 basis points (hundredths of a percent), against under 0.5 basis points for all other such windows. Holding from 2 pm until fifteen minutes before the announcement gave an annualized Sharpe ratio (return per unit of volatility) of 1.14. Results were almost identical with the E-mini future, which tracks the S&P 500, and the drift appeared in five European indices at 29 to 52 basis points. Counting each meeting day's return in the years before 1994, when decisions were not announced, the drift averaged 36 basis points over 1980-2011, with none before 1980. No trading costs were modelled. Knox and Vissing-Jorgensen (Federal Reserve working paper 2026) put the drift at about 180 basis points a meeting in the 15 meetings from March 2008 to December 2009. Boguth, Gregoire and Martineau (Journal of Financial and Quantitative Analysis 2019) find that over 2011-2017 it came only before meetings with a press conference; Lucca and Moench's 2018 update agrees: about 40 basis points there from the prior day's open to about lunchtime on announcement day. Kurov, Wolfe and Gilbert (Finance Research Letters 2021; E-mini) find it essentially gone after 2015, with or without a press conference; over all 70 meetings from April 2011 to December 2019 the published window, the 24 hours ending 15 minutes before the statement, averaged 0.047% against 0.038% on other days. Hu, Pan, Wang and Zhu (Journal of Financial Economics 2022), using S&P 500 futures (E-mini from September 1997) from the prior 4 pm close to five minutes before the release, found 27.14 basis points a meeting over 1994-2018 but 3.96 (not significant) in 2012-2018. Song (working paper 2026; E-mini, 2016-2023) finds the cumulative return from the prior day's open insignificant at every minute before 2 pm; the 24 hours to 2 pm earned 0.08% (not significant) from September 2016 to March 2020 and 0.37% (significant) from April 2020 to December 2023, 0.28% of it overnight. The published edge, 49 basis points a meeting over 1994-2011, compares with the library's cost of 10 basis points (0.10%) for a buy and a sale.
Results
| Measure | Orbital test | SPY, same days |
|---|---|---|
| Window | Oct 30, 2023 to Oct 30, 2026 | Oct 30, 2023 to Oct 30, 2026 |
| Total return | −0.87% | +88.60% |
| CAGR | −0.29% | +23.54% |
| Sharpe | −2.94 | 1.22 |
| Discounted Sharpe | −6.98 | — |
| Max drawdown | −3.9% | −18.7% |
| Win rate | 52.0% | — |
| Profit factor | 1.34 | — |
| Trades | 25 | — |
Hypothetical: a backtest with modeled costs, not traded with real money.
Hypothetical growth of $10,000
R13 Pre-FOMC Drift · −0.9% vs SPY +88.6%
Rebuild it in Orbital
Paste this into Orbital Chat. The Strategy Developer builds the workflow and asks before it runs any backtest.
Build a program on 1-minute bars for SPY, using the session clock. On each of these days before a scheduled FOMC announcement, if flat, buy SPY with 97% of equity when the 13:59 bar closes, so it fills at the 14:00 open: 2023-10-31, 2023-12-12, 2024-01-30, 2024-03-19, 2024-04-30, 2024-06-11, 2024-07-30, 2024-09-17, 2024-11-06, 2024-12-17, 2025-01-28, 2025-03-18, 2025-05-06, 2025-06-17, 2025-07-29, 2025-09-16, 2025-10-28, 2025-12-09, 2026-01-27, 2026-03-17, 2026-04-28, 2026-06-16, 2026-07-28, 2026-09-15 and 2026-10-27. On the next session, the announcement day, sell the whole position when the 13:44 bar closes, so it fills at the 13:45 open. None of these days or the announcement days after them closes early. Long only; uninvested capital stays in cash. Backtest from October 30, 2023 to October 30, 2026 with $100,000.
Is it real?
- Strategies tested together
- 10
- Beta to SPY
- 0.02
- Alpha vs SPY
- −4.8% a year
- Costs paid
- $2,436, 156% of the gain before costs
- Data
- 1m decisions on 1-minute bars, 754 sessions
- Fills
- Next open at market, sized at the fill
Read this before you trust it
SPY, a fund that tracks the S&P 500, stands in for the index. The paper reports no SPY returns: it finds the E-mini future's pre-FOMC returns almost identical to the index's, and uses SPY only for liquidity measures such as trading volume. The entry fills at the 14:00 minute's open and the exit at the 13:45 open, which is the published rule of selling fifteen minutes before a statement that has come at 2:00 pm since March 2013. Cash earns nothing between meetings. 97% of equity, leaving room for costs. Sixteen fills a year, each at 0.05% slippage (a fill worse than the quoted price).
Sources