Dynamic programming
- Now we move to time periodT 1 .By Bellmanís principle one should
solve the problem
VT 1 (w) = max
c 2 [ 0 ,w]
p
c+VT(k(w c))
=
= max
c 2 [ 0 ,w]
p
c+
q
k(w c)
.
Obviously the goal function is concave as
p
cis concave,k(w c)is
linear hence concave and thex7!
p
xis concave and increasing. So it is a
convex KT-problem.
d
dc
p
c+
q
k(w c)
=
1
2
p
c
+
1
2
p
k(w c)
( k)= 0
p^1
c
=
p
p k
w c
,ck=w c
σT 1 (w)=c=
w
1 +k.