1\begin{table}[htbp]\centering\small2\caption{Final: QLIKE relative to Random walk (trailing RMS)}3\label{tab:final_cmp_qlike}4\begin{adjustbox}{max width=\textwidth}5\begin{tabular}{lcccc}6\toprule7Model & $H=5$ & $H=15$ & $H=30$ & $H=60$ \\8\midrule9Seasonal volatility & -6363806176097.87\% (-7.9) & -7197644638305.72\% (-16.3) & -6261095459335.80\% (-8.4) & -4990594413186.81\% (-5.3) \\10EWMA-$t$ & -13401604177657.55\% (-34.5) & -11820625959028.91\% (-37.0) & -9792149701524.73\% (-19.7) & -7096098905937.95\% (-10.4) \\11HAR-RV & +3913236955668.53\% (2.9) & -10997388927703.89\% (-12.4) & -11101382476059.22\% (-16.4) & -11022768398499.26\% (-14.8) \\12GARCH(1,1)-$t$ & -6502045053666.21\% (-15.8) & -5672608189934.02\% (-17.9) & -3686696984534.77\% (-5.9) & -662150490172.45\% (-0.8) \\13Ridge & +5421776570677.55\% (4.5) & -12163950482648.31\% (-27.4) & -12503369384123.90\% (-21.0) & -11996364901816.19\% (-15.1) \\14Elastic net & +5473828257824.70\% (4.5) & -12144402667402.54\% (-27.4) & -12486397655586.67\% (-20.8) & -11981337790529.30\% (-15.1) \\15LightGBM & +33452663275996332.00\% (1.6) & -7986186304322.45\% (-5.4) & -11651629022629.18\% (-10.6) & -12817878377422.13\% (-16.0) \\16MLP (1m) & -14720898574375.21\% (-18.4) & -13512237008252.89\% (-17.6) & -12538500544410.84\% (-18.1) & -11631845999717.30\% (-15.2) \\17TCN (1m) & -16919024556944.02\% (-34.2) & -15473009084858.32\% (-48.8) & -14195580190418.18\% (-31.3) & -12904396803622.03\% (-22.0) \\18MRPT-S (1m+5m+30m) & -17440754813666.25\% (-31.0) & -16305890333133.57\% (-47.8) & -15169985549993.04\% (-31.2) & -14264208670641.35\% (-22.8) \\19MRPT-S (1m only) & -17163562658657.87\% (-31.4) & -15931251452952.67\% (-45.8) & -14662674862923.33\% (-27.6) & -13519829970214.44\% (-19.2) \\20MRPT-B (1m+5m+30m) & -17239229400549.80\% (-40.5) & -16046379760611.44\% (-53.4) & -14931279197542.62\% (-36.3) & -13974559954094.02\% (-26.5) \\21\bottomrule22\end{tabular}23\end{adjustbox}24\begin{minipage}{0.95\textwidth}\scriptsize Relative change in mean QLIKE versus the reference (negative = better); Diebold--Mariano $t$-statistic with HAC variance in parentheses.\end{minipage}25\end{table}