Shanghai　crude　oil　futures　contracts　(SC)　has　been　the　third　largest　trading　volume　crude　oil　futures　worldwide.　From　wavelet　denoising-higher　moment　perspective,　we　investigate　the　relationships　and　portfolios　between　oil　and　Chinese　stock　sectors,　based　on　the　daily　data　of　WTI,　SC　and　Chinese　sectors.　Given　the　existence　of　noise　and　time-varying　conditional　higher　moments　in　financial　data,　the　Wavelet　denoising-GARCHSK　and　Wavelet　denoising-GARCHSK-Copula　portfolio　methods　are　proposed　to　model　the　margin　of　variables　and　construct　oil-stock　portfolios,　respectively.　The　empirical　results　indicate　that　the　Wavelet　denoising-GARCHSK　outperforms　the　other　methods　in　terms　of　log　likelihood,　AIC　and　BIC.　Then,　these　results　also　denote　that　there　are　the　notable　industry　characteristics　in　the　oil-stock　relationships,　depending　on　whether　the　sector　is　petroleum-dependent　or　petroleum-independent　as　well　as　locates　upstream　or　downstream　of　petrochemical　industry.　In　addition,　most　of　the　Chinese　stock　sectors　have　stronger　upper　tail　dependence　with　SC　than　WTI,　while　the　lower　tail　is　the　opposite.　Finally,　it　is　discovered　that　the　oil-stock　portfolios　obtained　by　Wavelet　denoising-GARCHSK-Copula　portfolio　model　outperform　the　control　groups　in　terms　of　return　and　ASR　and　that　WTI　is　a　better　diversifier　for　investors　in　Chinese　stock　market　in　most　cases.　©　2020　Elsevier　Ltd