I was reading a paper by Mehmet Asutay on the difference between how Islamic finacne diverges from the original conception of the "Islamic moral economy" and I was struck by something which I have not spent much time thinking about. A lot of the focus with Islamic banks is on why they are using murabaha and ijara rather than mudaraba and musharaka, and within the sukuk markets there has been some controversy about whether mudaraba sukuk are structured in a way that is or is not Shari'ah-compliant.
As a result, there has been a lot of attention paid to the structure of mudaraba sukuk, and relatively little paid to the particular structures used by Islamic banks when they make mudaraba investments (in long-term assets). Why is this? For the most part, it is due to convenience. Sukuk offering documents are publicly available when the sukuk are listed while bank financing documents are private documents between the bank and the customer.
However, it would be likely that in most countries (perhaps excepting those with separate Islamic banking regulations, but not necessarily), the regulatory rules around banks offering mudaraba would be more strict than similar rules around sukuk. The reason being that Islamic banks, like other banks, are regulated to avoid failures, and the main risk is that their assets decline in value (or even are perceived as having declined in value, triggering a run).
As such, the use of mudaraba would be scrutinized by regulators to assess their probability of generating a loss for the bank, which would deplete its capital. Since the bank (as rabb al-mal) would in a traditional mudaraba be responsible for bearing the losses itself, and it would be more of an equity than a debt investment, there would be more pressure, other things equal, for the bank to push the boundaries on what is permissible while still remaining Shari'ah-compliant in order to reassure regulators and to limit the capital charge on the mudaraba assets it holds on its balance sheet than would a sukuk issuer.
I cannot really say much more about what actually goes on in practice, because as I mentioned the mudaraba contracts are conducted on a private basis. However, given the tendency of banks to be more financially risk-averse than the overall investment market (since by their nature, banks are leveraged and rely on short-term financing in the form of current accounts from depositors), it would be interesting to see a study of bank mudaraba contracts done by Shari'ah scholars independent of the banks being analyzed that looks at the degree to which the products are structured to replicate bank loans, or the degree to which they maintain the traditional profit and loss sharing structure that is typically associated with mudaraba contracts.