sql server - Better Practice: Delete&Insert vs. Update -

i need parse data (~60000 rows) csv file , write them mssql table (the data date/time , value decimal number). each day 1 such csv file. catch is, in csv file each day have data last 5 days, meaning have data dates wrote in past days, need replace data file.

i trying decide between 2 methods: bulk delete of old data need re-write when new csv file, , insert, vs finding each record based on date&time , id , update it.

1.what better practice cause less fragmentation , maintenance issues on database?

2. which be less expensive performance point of view?

if comes choosing between two, prefer keeping db in shape on high performance because file written during night anyway.

edit: in case i'll add maintenance plan rebuilds indexes on daily basis after bulk deletion , insertion of new data, enough avoid fragmentation issues, or there missing?

faster & better way delete old data, import data ssis or bulk insert in case if don't have ssis , rebuild fragmented indexes. example script.