I have been having some interesting discussions with a few people about my recent posts about the Saudi Hollandi Bank mudaraba sukuk and ijara sukuk in general and it has led me to think more about the overall direction for the sukuk market now that the prospects for new issuance seems to be improving.
Should sukuk mirror conventional bonds?
One of the overarching themes is whether sukuk should mirror conventional bonds or whether there can be a different risk-return profile while retaining investor interest. Until now the sukuk market is primarily focused on producing Shari’ah-compliant bond-equivalent financial products. This trend of product development was criticized for creating products which used Islamic financial contracts in theory while creating an identical economic outcome with protections for investors that mirror conventional bonds. This criticism was partially resolved with the AAOIFI ruling on sukuk which limited, among other things, the ability of mudaraba and musharaka sukuk to have repurchase agreements where the issuer would repurchase the sukuk at par upon maturity or in cases of default.
This ruling, coming in the midst of the early stages of the Great Recession, was a contributing factor in the steep decline in new sukuk issuance. This was due to the heavy use of mudaraba and musharaka sukuk by Islamic banks who represent a significant share of sukuk issuance. The AAOIFI resolution occurred several months after Sheikh Taqi Usmani made comments that 85% of all GCC-based sukuk were not structured in a Shari’ah-compliant manner. Until the Saudi Hollandi Bank sukuk, few issuers brought new mudaraba and musharaka sukuk to the market. Instead, most new issuers used the ijara structure, which was not subject to the limitations on purchase undertakings at par to redeem sukuk.
The use of repurchase agreements in ijara sukuk
The reliance on ijara sukuk with repurchase agreements after similar structures were prohibited for mudaraba and musharaka sukuk highlighted a discrepancy in the Shari’ah standards between different types of sukuk. There are differences between the two structures and there are fewer restrictions on how profits and losses are shared between the two parties in the ijara structure. However, the ijara sukuk structure includes a purchase and sale transaction at issue and maturity, which is not part of the actual leasing transaction.
The ijara transaction is relatively uncontroversial. The SPV, as beneficial owner of the assets, leases those assets to the issuer company for a set period with the rents benchmarked to an interest rate and sometimes a variable rental rate that resets periodically to incorporate changes to the benchmark interest rate. The other part of the transaction involves a sale of the asset by the issuer company to the SPV using the proceeds from the issue of sukuk certificates and a sale back to the issuer company at maturity in default or at maturity to redeem the sukuk.
The sale and repurchase (at par) transactions do seem to be more controversial because they essentially strip the asset out of the transaction for the purpose of investors looking at the transaction. From the view of the economic outcome, an ijara sukuk with an repurchase clause and where the sale of the assets to the SPV are not a ‘true sale’ but instead transfer beneficial ownership to the SPV is an unsecured obligation of the issuer company. As a comment to an earlier post pointed out, there may be fewer legal rights for sukuk holders (although I do not have enough legal expertise to say this with certainty) than for a conventional unsecured bond. The purchase undertaking may create only damages that the investors can claim, rather than a debt that may rank higher in order of priority in a liquidation. In a ‘true sale’, the legal ownership of the asset would be transferred to the SPV and in a default, the SPV will be able to take possession of the assets if the issuer is not able to produce the funds needed to fulfill the purchase undertaking (again, I do not have legal expertise enough to make this statement conclusively).
The impact for the Islamic finance industry of the sale-lease-repurchase structure of the ijara sukuk is important. Frequent claims are made that Islamic finance is more resilient, provides a superior outcome, and is more stable than conventional finance based on its connection with real assets. However, if the transaction creates unsecured obligations that may or may not be equal in a liquidation to the claims of unsecured bondholders, then it would be wholly reasonable to dispute this claim and point to secured financing as a more stable form of financing. In a secured financing (and I think this would be true if the assets were sold in a ‘true sale’ to the SPV), the investors have recourse to the underlying assets if there is a default. Both parties share the risk of failure: the issuer loses the use of the assets unless the investors decide to continue leasing the assets to the issuer and the investors lose the difference between par value and the market value of the assets in the sukuk.
This idea leads to whether sukuk should be equivalent to bonds in the investor protection and the model of an ijara with a ‘true sale’ may provide this (it could potentially be overcollateralized to provide additional protections). At the same time, it would provide a path by which investors could receive protection while perhaps setting up a more satisfactory resolution of a default for both investors and the issuer. The SPV would retain ownership in the default (rather than giving it up for an unsecured claim through the purchase undertaking) and the issuer would lose the use of the asset. The asset could continue to be leased to the issuer while a price was negotiated for the purchase of the asset at a value above market value and below the par value, which would lead to a sharing of risk between issuer and investors.
Mudaraba versus musharaka: Are there incentive problems that could hurt investors?
For mudaraba and musharaka sukuk, there seems to be a different trend emerging following the Saudi Hollandi Bank sukuk. Instead of having the issuer advance interest-free loans to make regular payments, the mudaraba would pay out regular coupons (fixed-rate or floating-rate) during the term of the sukuk, while also tracking profits (and profit-sharing) from the mudaraba. At the redemption, the redemption value would be equal to the value of the mudaraba assets plus any amount in the reserve account (which accumulated profits in excess of the periodic payment amounts) until the par value is reached. If the assets plus reserve account was insufficient to make repayment at par, the investors would receive less than par.
This structure includes more risk for investors than an ijara sukuk because the value of the sukuk assets falling below the par value would likely lead to a reduced payout at maturity. This could limit the appeal of these types of sukuk, although this was not evident in the Saudi Hollandi Bank sukuk which was four times oversubscribed. This demand may not be present if a mudaraba sukuk structured in a similar way were to default with investors paid less than the par value. However, if Islamic finance wants sukuk to become an alternative asset class for investors, this is not a bad thing.
This structure seems particularly advantageous for Islamic banks, which often lack the physical assets to structure other forms of sukuk. Their business model, which relies upon making a spread between cost of financing and the returns they receive from extending financing (offset by losses in financing), which should be sufficient to redeem the sukuk at par in most times. However, as we have seen, not all times are “most times” and the loan losses may be accentuated if the banks are incentivized to take more risk than is prudent for the investors.
The risk of this excessive risk incentive is present in the sukuk structure used in the Saudi Hollandi Bank sukuk, which could have been mitigated if it were a musharaka sukuk, as I’ll describe in a second. The way it was structured, the bank commingles its own investments with the investments accounted for in the mudaraba sukuk to develop its Islamic banking business. The mudaraba assets are then used to extend financing to produce a return greater than the floating-rate payments to the sukuk certificateholders. Any losses are borne by the investors. The excess amount remaining in the reserve account beyond that required to redeem the sukuk at par are paid to the bank as an ‘incentive payment’.
This structure creates a perverse incentive for the bank. If they are able to generate sufficient returns to cover the periodic payments plus redemption at par, they receive the excess profits. If their investments fail to produce sufficient returns to make periodic payments and redeem the sukuk at par, the investors bear that loss. If the returns are insufficient to cover the periodic payments and redeem the sukuk at par, but the sukuk assets are valued at par, then the bank pays out the difference between the actual returns and the variable-rate profits, which are then an additional cost to the bank.
The incentive created is for the bank to take excessive risks with investors’ capital to generate a return greater than the cost of capital because every dollar generated above the profit rate will flow to the bank, while any losses from adopting the riskier approach will be borne by the investors (although the bank will likely be hurt as well through the impact on its capital and ability to raise money in the future). This seems disturbingly close to the ‘heads I win, tails you lose’ incentive created in the banks that failed during the financial crisis.
One possible solution would be for the bank to adopt a musharaka structure instead, especially since the bank commingled assets with the mudaraba assets. In the musharaka, in contrast with the mudaraba, the losses are shared between both parties based on the assets contributed to the joint venture. This would align the incentives of each party better and make it less likely that the risk taken on by the bank was closer to optimal from the perspective of the sukuk investors.
This post is far from complete in the discussion of the issues of ijara, mudaraba and musharaka sukuk, but it is my hope that it will be a starting point for further discussion about the future of sukuk structures. The sooner this discussion occurs within the industry, and most importantly with the Shari’ah scholars who can provide the Shari’ah-compliance perspective that I cannot, the better suited the Islamic finance industry will be for the recovery in financial markets and the flood of new issuance sure to come when the recession ends. If this introspection makes its way into new issuance, there should be greater certainty about the rights of sukuk investors should another crisis occur (which it most certainly will at some point).