Humans exchange labor or goods for money, which they later exchange for
labor or goods they want. The money is an abstraction, and we carry
that abstraction a step further by putting the money in a bank,
exchanging the money for a piece of paper acknowledging the bank has custody
of it.
Dogs readily learn to perform for a reward: food or frequently just
approval. I suspect it would be far more difficult to train a dog to do a
trick for a token which could later be exchanged for food or a pat on the
head, and even more difficult to train the dog to exchange the token for a
piece of paper that offers repayment with interest.
So I would say that money and its abstractions, contracts and other
promises of future performance, and their abstractions, are kinds of
behavior that exist in the human world but not in the animal world.
Someone will point out that there is evidence certain animals deal in
tokens -- sticks or shells that are exchanged for food or favors, but I
doubt that abstractions of these tokens will be found in the animal world.