Hi,
It will automatically failover to anther link and failback to loadsharing if link is up again If you configure using a config example #4 from http://www.cisco.com/en/US/tech/tk365/technologies_configuration_example09186a00800945bf.shtml#conf4
All routing decision is based on BGP path selection decision and you will have load-sharing and redundancy at same time with below config example.
interface Serial 0
ip address 160.20.20.1 255.255.255.0
no ip route-cache
interface Serial 1
ip address 150.10.10.1 255.255.255.0
no ip route-cache
router bgp 11
neighbor 160.20.20.2 remote-as 10
neighbor 160.20.20.2 route-map UPDATES-1 in
!--- This allows only the networks up to 128.0.0.0.
neighbor 150.10.10.2 remote-as 12
neighbor 150.10.10.2 route-map UPDATES-2 in
!--- This allows anything above the 128.0.0.0 network.
auto-summary
route-map UPDATES-1 permit 10
match ip address 1
set weight 100
route-map UPDATES-1 permit 20
match ip address 2
route-map UPDATES-2 permit 10
match ip address 1
route-map UPDATES-2 permit 20
match ip address 2
set weight 100
access-list 1 permit 0.0.0.0 127.255.255.255
access-list 2 deny 0.0.0.0 127.255.255.255
access-list 2 permit any