\section{Related literature}
\label{sec:literature}

\paragraph{Prices lead earnings.} Since \citet{ball1968}, accounting research has documented that stock prices
anticipate accounting income: annual returns forecast earnings changes \citep{beaver1980}, leading returns raise
earnings-response coefficients \citep{kothari1992}, and the low contemporaneous return--earnings association reflects
prices' timeliness rather than noise \citep{collins1994}. The post-earnings-announcement drift \citep{bernard1989,
bernard1990} and the accrual anomaly \citep{sloan1996, richardson2005} show that the incorporation of accounting news
is incomplete, while fundamental analysis \citep{ou1989, piotroski2000} extracts forecasts of future earnings from
statements alone. Recent work replaces hand-crafted ratios by machine learning on detailed statement items
\citep{chen2022, easton2024, cao2024}. All of these studies operate at daily or lower frequency and treat the market side
as a return; our contribution is to fix the accounting side and vary the \emph{resolution} of the market side.

\paragraph{Information in intraday activity.} Market microstructure theory ties the composition of trading to private
information \citep{kyle1985, easley1987}, and realized-measure econometrics separates continuous variation, jumps and
their signs from high-frequency bars \citep{andersen1998, andersen2003, barndorff2004, barndorff2010semivariance,
patton2015, lee2008}. Signed jump variation and downside semivariance price the cross-section
\citep{bollerslev2020good, bollerslev2016beta}, jumps cluster around news \citep{lee2012, bajgrowicz2016}, and
announcement days change betas and the volume--volatility relation \citep{patton2012, bollerslev2018volume,
savor2016}. The question of the best sampling frequency has been settled for \emph{variance measurement} -- five-minute
realized variance is hard to beat \citep{liu2015, bollerslev2016errors, bollerslev2018risk} -- but not for the
\emph{forecasting of fundamentals}, which is our object. \citet{zhang2023volatility} show that pooling intraday
commonality improves machine-learning volatility forecasts; we include volatility families at every frequency and use a
symmetric placebo target to separate a volatility channel from a signed-information channel.

\paragraph{Pre-announcement trading and price discovery.} Short sellers \citep{christophe2004}, individual investors
\citep{kaniel2012}, institutions \citep{hendershott2015} and option traders trade informatively before earnings
announcements; differences of opinion shape announcement returns \citep{berkman2009}; algorithmic trading may reduce
pre-announcement information acquisition \citep{weller2018}; and the announcement reaction itself occurs largely in
quotes rather than trades \citep{gregoire2022, brogaard2019}. These studies use trader-type or quote data that bars do
not contain. We ask a complementary question with anonymous OHLCV only: how much of the forthcoming accounting outcome
is visible in the shape of pre-release trading, and does that visibility depend on resolution?

\paragraph{Attention and processing.} Limited attention \citep{hirshleifer2003, hirshleifer2009, dellavigna2009,
aboody2010} and media coverage \citep{fang2009} govern how fast public information is processed; the timing and
heterogeneity hypotheses H3--H4 are motivated by these channels and by distress-risk measurement \citep{campbell2008,
beneish1999}.

\paragraph{Forecast evaluation and machine learning in finance.} We follow the evaluation standards of
\citet{diebold1995}, \citet{giacomini2006}, \citet{hansen2005} and \citet{hansen2011mcs}, with multiple-testing control
\citep{holm1979, benjamini1995}, cluster-robust inference \citep{petersen2009, cameron2011}, block bootstrap
\citep{kunsch1989, politis1994} and purged, embargoed time splits \citep{lopezdeprado2018}. Equal-budget comparisons of
information sets are inspired by the asset-pricing machine-learning literature \citep{gu2020, freyberger2020, kozak2020,
chen2024deep, avramov2023} and its warnings about false discoveries \citep{harvey2016, chordia2020}. Gradient-boosted
trees \citep{ke2017lightgbm} provide the nonlinear benchmark. The companion UQO working paper on intraday shock
propagation \citep{boucher2026wp15} uses the same vendor data for a different question; no result is reused here.