1\section{Economic framework and hypotheses}2\label{sec:theory}34\subsection{Anticipation versus revelation}56Let $X_{i,q}$ denote an accounting flow of firm $i$ for fiscal quarter $q$ (operating income in our primary7specification), revealed to the public at the earnings-release date $A_{i,q}$ and structured in the SEC filing of date8$F_{i,q}\ge A_{i,q}$. A forecaster standing at a trading day $c<A_{i,q}$ holds an information set $\mathcal{I}(c)$ that9contains, at most, prices and volumes up to the close of $c$ and the accounting statements that were \emph{available} at10$c$, i.e., quarters $j$ with $F_{i,j}+1\le c$. The quantity of interest is the conditional distribution of the surprise11\begin{equation}12 S_{i,q} \;=\; \frac{X_{i,q}-\mathbb{E}\!\left[X_{i,q}\mid \mathcal{A}(c)\right]}{TA_{i,j^*(c)}},13 \label{eq:surprise}14\end{equation}15where $\mathcal{A}(c)\subset\mathcal{I}(c)$ is the accounting information alone, $j^*(c)$ is the latest available quarter,16and the expectation is a seasonal random walk with drift \citep{foster1977}, the classic honest benchmark when no17point-in-time consensus exists. Semi-strong efficiency \citep{fama1970} implies that prices at $c$ impound all public18information, including $\mathcal{A}(c)$; the ``prices lead earnings'' literature \citep{beaver1980, kothari1992,19collins1994} documents that daily and annual returns indeed forecast accounting changes. None of this is the question20we ask. Our question is whether the \emph{resolution} at which the market is observed carries incremental information,21holding the information set, the universe, the calendar coverage, the horizon and the model capacity fixed.2223\subsection{Why resolution could matter, and why it might not}2425Three mechanisms make fine-grained activity informative beyond daily aggregates. First, in sequential-trade models26\citep{kyle1985, easley1987}, informed traders split orders and trade when liquidity is deep, so the \emph{composition} of27a day's volume and return (intraday autocorrelation, concentration of variance in a few minutes, volume on down-bars) can28reveal private information that the daily net return hides. Second, realized measures built from fine bars separate the29continuous and jump components of variation \citep{barndorff2004, andersen2003} and their sign30\citep{barndorff2010semivariance, patton2015}; negative semivariance and signed jumps have been linked to the31cross-section of expected returns \citep{bollerslev2020good} and may proxy for the arrival of adverse firm-specific news.32Third, limited attention \citep{hirshleifer2003, hirshleifer2009, dellavigna2009} predicts that the incorporation of33public but costly-to-process information is gradual, so that the time path of trading within the day (opening and closing34concentration, overnight versus intraday returns) carries information about ongoing processing.3536Against these channels stand two forces. The Epps effect and microstructure noise contaminate 1-minute returns37\citep{hansen2006, barndorff2008}, so that the finest sampling is not the most informative for variance measurement38\citep{liu2015}; and whatever information intraday activity contains about fundamentals may be fully summarised by the39daily close, in which case the frequency premium is zero. Algorithmic trading may further have reduced the amount of40pre-announcement information acquisition \citep{weller2018}. The sign and size of the frequency premium are therefore41empirical questions, and a precise null is as informative as a positive estimate.4243\subsection{Hypotheses}4445All hypotheses, endpoints and decision rules were frozen before the test period was opened (\texttt{docs/RESEARCH\_DESIGN.md}46in the replication package). Let $\mathcal{L}(M)$ denote the out-of-sample log loss of model $M$ for the binary adverse47outcome $D_{i,q}=\mathbb{1}[S_{i,q}\le -0.5\%]$, and let B1--B5 be the nested information sets of Table~\ref{tab:infosets}.4849\begin{description}[leftmargin=1.5em]50\item[H1 (incremental information).] $\mathcal{L}(\mathrm{B5})<\mathcal{L}(\mathrm{B2})$ at the 20-trading-day horizon:51 one-minute features improve forecasts beyond point-in-time fundamentals and daily OHLCV. Null: no improvement.52\item[H2 (temporal resolution).] The gain of B5 over B2 exceeds the gains of B4 (5-minute) and B3 (30-minute) over B2,53 with the same feature families and the same number of inputs. Null: the gains are equal.54\item[H3 (heterogeneity).] The B5$-$B2 gain is larger for small, illiquid and highly levered firms, the groups in which55 information asymmetry and processing costs are highest. Null: homogeneous gain.56\item[H4 (timing).] The gain increases as the cutoff approaches the release ($h=40\to1$) and is absent at placebo cutoffs.57 Null: no gradient.58\item[H5 (pricing).] The ex-ante deterioration probability predicts the announcement-window abnormal return and59 post-cutoff factor-adjusted returns. Null: no return predictability after risk adjustment and costs.60\end{description}6162A positive H1 does not identify \emph{who} trades on what: observational prediction is not causal evidence of informed63trading, and we do not have quotes, trades or trader identities \citep{gregoire2022, hendershott2015}. The paper's claims64are confined to the forecasting content of anonymous bars and to the frequency at which that content appears.65