Transaction isolation level in financial applications

Question

I read about transaction isolation levels (serializable, repeatable read, read, read unfixed), and I'm a little confused about their use. Do I understand correctly that, for example, in banking applications (money transfers, etc.), the isolation level used is probably serialized, since this rule is not allowed to be relaxed?

Answer 1

You can choose the level of isolation depending on how you want your application to work. Depending on what you are doing with the data, you may need different isolation levels for different db calls.

There are no hard and fast rules for any particular industry, such as banking, but some data (such as financial transaction data) is obviously important and you want to do everything possible to avoid data loss and data inconsistency.

The isolation levels are better explained than I can here.
http://en.wikipedia.org/wiki/Isolation_%28database_systems%29