Announcement Days Carry the Premium, and the Jump
Are announcement days responsible for a disproportionate share of observed returns?
Measure returns on predeclared announcement and other dates within the studied market and period, then compare with uncertainty and costs. A historical premium does not override FTMO news restrictions or justify running a prohibited trading window.
The US equity premium and FX event response do not establish a safe or permitted strategy around FTMO news restrictions.
Evidence map
| Aspect | Finding |
|---|---|
| What it is | Two empirical results on different markets and samples. |
| Key result / formula | Savor and Wilson compare excess returns on scheduled inflation, employment and policy announcement days with other trading days. |
| Why it matters for backtesting | The equity study suggests checking whether a proposed strategy's returns depend on a small set of scheduled dates. |
What it is
Savor and Wilson document a concentration of average US equity excess return around scheduled macroeconomic announcement days in their study. Andersen and coauthors examine short-window exchange-rate reactions to the surprise component of releases. Neither result is a general instruction to trade through a news window.
Key result / formula
Their sample shows a higher average premium on announcement days, consistent with compensated macroeconomic risk in that equity setting. The size of that difference belongs to their period and market; it cannot be copied into a prop-firm account. Andersen, Bollerslev, Diebold and Vega compare high-frequency foreign-exchange quotes with the difference between an announced value and the market's survey expectation. Their event study finds a rapid exchange-rate response to this surprise. The speed and asymmetry of the response depend on the release and currency pair. A bar whose timestamp covers the release cannot establish a tradable fill before the observed move unless the signal and execution clocks are separately recorded.
Why it matters for backtesting
It does not show that exposure on those dates is safe or permitted for every account. A prop-firm news restriction, when applicable, governs the trading rule even if avoiding releases has an opportunity cost. The FX study requires event time, release vintage, survey expectation, instrument quote time and order fill time before a claimed news reaction can be reproduced. On coarse bars, the sequence of the move within a bar may be unknown. A manual audit can align a public event calendar with the strategy's bar convention and compare exposure, costs and returns in windows declared before inspecting outcomes. Selecting the largest absolute-return bars afterward is a descriptive stress check, not proof that those bars were scheduled announcements.
Worked check
Write the list of release types and the pre- and post-event window before reading strategy returns. For each event, record whether the strategy held a position before the release, when the signal became available, and the earliest feasible fill. Compare matched exposure outside the window and report costs and excluded dates. If an account rule requires closing or not opening positions near news, enforce that rule first; the empirical equity premium does not override it.
Source
Savor & Wilson, "How Much Do Investors Care About Macroeconomic Risk? Evidence from Scheduled Macroeconomic Announcements", Journal of Financial and Quantitative Analysis 48(2), 2013, 343-375; Andersen, Bollerslev, Diebold & Vega, "Micro Effects of Macro Announcements: Real-Time Price Discovery in Foreign Exchange", American Economic Review 93(1), 2003, 38-62. Primary source