Why did Estimated Subtree Cost increase with nonclustered

I have following table with six rows

CREATETABLE TimePeriod
(
        TimePeriodId intnotnullidentity(1,1)primarykey,
        TimePeriodName nvarchar(30)notnull,
        OrderId intnotnull,
        sDelete bitnotnulldefault'False'
)
When I run this query
        SELECT*FROM TimePeriod where timeperiodname='quarter'

Then Estimated subtree cost comes 0.0032886

But when defined non clustered index on  column TimePeriodName

CREATEUNIQUENONCLUSTEREDINDEX IX_TimePeriod_TimePeriodName
ON TimePeriod
(
        TimePeriodName ASC
)
 WITH (PAD_INDEX  =OFF,
  STATISTICS_NORECOMPUTE  =OFF,
  SORT_IN_TEMPDB=OFF,
  IGNORE_DUP_KEY=OFF,
  DROP_EXISTING=OFF,
  ONLINE=OFF,
  ALLOW_ROW_LOCKS  =ON,
  ALLOW_PAGE_LOCKS  =ON)

Then run query
        SELECT*FROM TimePeriod where timeperiodname='quarter'

I get Estimated subtree cost is  0.0065704

So why is cost incresing when non clustered index is defined on the table?
Where I am doing mistake?