December 2012

December 2, 2012

When Abu Dhabi Islamic Bank issued its latest sukuk, the demand was large (the $500 million sukuk was upsized to $1 billion after it received orders of $15 billion).  However, the structure used–a perpetual, non-cumulative mudaraba with a call after 6 years—demonstrates how a sukuk can be structured as a mudaraba that will be called at its par value, even though an AAOIFI resolution in 2008 said that a purchase undertaking should  not specify ex ante the price at which the purchase is completed.

Specifically, AAOIFI prohibited: “the Mudarib (investment manager) […] to undertake {now} to re-purchase the assets from Sukuk holders or from one who holds them, for its nominal value, when the Sukuk are extinguished, at the end of its maturity.”  In a literal sense, ADIB is not undertaking to purchase the assets from Sukuk holders “at the end of its maturity”, since the it carries a perpetual maturity, and will only be redeemed if ADIB chooses to redeem it.

The sukuk documents, however, specify that if the sukuk are called, they will be called at their “outstanding face amount together with any Outstanding Payments”.  However, from my perspective as neither a lawyer nor a Shari’ah scholar, this looks like it should fall afoul of the AAOIFI resolution unless the argument is made that the mudaraba sukuk does not fall under the AAOIFI resolution because there is no fixed “end of its maturity” when ADIB is committing to repurchase the assets.

There are economic justifications for why the option for ADIB to call the sukuk at par is different than a similar purchase undertaking in a mudaraba sukuk with fixed maturity.  For example, if the value of the assets of the mudaraba declined, the bank would likely not be in the position to exercise its call option (or would be prohibited by its regulators from doing so). So, the value of the assets in which the mudaraba capital was invested will be at least equal to the face value of the sukuk.

But, why not provide some sharing of that upside between the bank and its mudaraba sukuk holders.  If there were potential upside to the call value of the sukuk, it could lower the cost to the bank in terms of periodic distribution amounts by providing upside if the bank’s assets rise in value.  This would allow the sukuk to retain its debt characteristics (fixed periodic distribution amounts paid from the profits generated by the mudaraba assets) but also introduce some profit-sharing in the redemption of the sukuk if they are called by the originator.

It’s unlikely that this would attract more investors (since the ADIB sukuk was already heavily subscribed) and also because bond investors are primarily concerned with the potential for losses, not gains, but neither would it scare off any potential investors and it would become more fully aligned with my own reading of AAOIFI’s intent with their mudaraba sukuk resolution.

December 9, 2012

Several years ago, the idea of takaful for sukuk was considered, but as far as I know the discussion never moved forwards towards implementation and I think this is a good thing.  The development of takaful coverage for sukuk (insurance against sukuk default) introduces a tail risk to Islamic finance that doesn’t exist today, and would be increased by the incentive structure of takaful operators.

A principal-agent problem is created where the takaful operator is paid a fee based on the volume of premiums paid, but where the takaful participants bear the risk of loss in the investments made with the premium income, as well as facing loss if the premiums are insufficient to meet all claims (see Chapter 7 of this book).

There are well established actuarial formulas for providing family takaful (life insurance), auto and health takaful, but it would be much more difficult for a takaful provider to accurately price a sukuk takaful, and they would be incentivized to underprice it to attract more business.  In addition to this potential problem, there is the challenge in accurately assessing the probability of tail events that would lead losses on the insured sukuk to become highly correlated with one another, and inversely correlated with the value of the invested assets in the takaful pool.

To explain further, consider a takaful pool that is formed to insure 10 sukuk, issued by a variety of GCC corporates from different countries and different industries.  The takaful operator will price the premiums based on expected default rates for the different sukuk.  This will be difficult on its own because there have been few defaults in sukuk.  But, assume that the takaful operator can find a large enough pool of both bonds and sukuk to assign a probability of default for each, as well as the recovery value if each one defaults within a defined time frame.

The premiums will be collected based on this calculation (which, again, may be underpriced since there will be an incentive for the takaful operator to underprice the default risks or overstate the likely recovery in the the case of default in order to attract more issuers).  The takaful provider will invest the premiums in sukuk, equities, real estate, and structured products, in a way to meet its expected needs for any defaults.

What happens in a traditional takaful provider if a tail event, like a global financial crisis happens?  The value of the assets in the takaful pool will probably decline in value and there might be a shortfall in the assets of the takaful pool to meet claims.  Typically, takaful operators would respond in this case by providing a qard loan from the operator’s capital to the takaful pool to cover the shortfall, and when things recover, any future surpluses go to repaying the qard from the takaful operator.

The decline in the value of the assets of the takaful pool led to a temporary shortfall.  There will not be an effect on the number of claims from takaful providers  since mortality rates are not highly—if at all—correlated with financial market performance.  Therefore, so long as the takaful operator is adequately capitalized, it can withstand the impact and move on.

In a sukuk takaful arrangement, this will not be true.  The takaful operator, in normal times, will be able to identify enough Shari’ah-compliant assets that are uncorrelated with the likelihood of the issuers defaulting to make management of the takaful pool work, where the investments are made to provide high enough returns to withstand the expected rate of defaults.  There may be temporary shortfall if the takaful coverage was drawn on, which can be met with a qard loan from the takaful operator.

But, what happens if there is another global financial crisis?  Unlike family takaful where claims are based on factors independent of the value of the assets in the takaful pool (so decline in value of the assets can be met with qard loans from the operator), the claims on a sukuk takaful pool would be inversely correlated with the value of the assets in the takaful pool.

As the effects of the financial market crisis spread, the assets decline in value at the same time as the probability of default by the sukuk issuers rises because, for example, the financial crisis has led to a recession and an inability of the issuers to refinance their sukuk when they come due, or an inability to receive working capital financing as banks avoid taking any additional risks.

Furthermore, a new risk emerges which will have an endogenous effect on the likelihood of it occuring: sukuk investors become concerned that the takaful pool will have sufficient assets to meet the claims, and become concerned about the takaful operator’s ability to provide a qard loan to meet claims.  A rational response by sukuk holders to this risk would be to sell the sukuk, driving down their price and driving up their yield, which will make the issuers more likely to have difficulty refinancing their sukuk (and thus increasing the probability of default).

And this type of tail risk has occurred in the conventional market.   Remember AIG, the global insurer that nearly collapsed because its liabilities swelled as the historical correlation between the assets it insured (bonds insured through CDS) broke down?  There is not necessarily anything unique within takaful or sukuk that would prevent a reoccurrence of a tail event leading to significant declines in the value of the takaful pool’s assets at the same time as the correlation between the insured sukuk rises and default rates jump exceed historical averages.

As a result, I think it is better if the idea of takaful for sukuk (would that be called sukuk insurance, or Islamic CDS?) remained an idea considered and rejected as impractical and overly risky.

December 16, 2012

It is commonly understood that at least part of the reason for the decline in sukuk issuance in late 2007 and 2008 was due to Sheikh Taqi Usmani’s critical comments about the structures used by sukuk issuers.  In November 2007 said 85% of all sukuk were not Shari’ah-compliant because they included repurchase clauses that specified the price being par (a number decided at the outset and not based on the underlying value of the asset being bought and sold).  While I think there is some doubt about the impact his statements had in terms of issuance (this was around the same time that the financial crisis was hitting global debt markets).

There was likely more of an impact from the uncertainty around what was allowed and what was prohibited that lasted until February 2008 when AAOIFI issued a resolution (pdf) that clarified that the restriction on purchase undertakings was specific to mudaraba and musharaka sukuk, while ijara sukuk were exempted.  By this point, the financial crisis was intensifying with Bear Stearns being sold to JP Morgan, so it would be hard to disentangle the effect of the clarification provided by AAOIFI from the macroeconomic events of the day.

Since the end of the financial crisis, the Shari’ah-compliance risk of purchase undertakings has diminished significantly in the minds of investors as sukuk issuance is setting another record high in 2012, estimated at $121 billion for the year.  However, the risks remain and issuers that want to differentiate themselves could turn to sukuk structures that are less exposed to future changes in Shari’ah opinion.

Now, what is the risk?  In the recently released Thomson Reuters Zawya Sukuk Perceptions and Forecast Study 2013, Tariqullah Khan defines the risk as being ‘structure risk’.  He writes, “I would suggest that structure risk is the risk of losing investment value because of ambiguity between the exposure to price risk of the sukuk asset as an equity stake, and the credit risk of the originator because of the expectations for performance of the investment within the maturity time of the sukuk as a conventional bond.”  He further elaborates that if asset price risk, credit risk and rate of return risk of an asset are bundled and not separated, structure risk exists in the design of the asset.

From my reading, this can be described as the risk that the change in the value of the underlying asset will adversely affect investors return if the repurchase agreement clause is repudiated by a Shari’ah ruling, and the repurchase has to be done at market value upon maturity (rather than at par).

It would make for an interesting struggle between investors and the issuer about whether the terms of the contract can be changed based on subsequent rulings.  In general, I would think they could not, since the sukuk documentation usually does not allow for ex post changes to the legal arrangement based on subsequent Shari’ah rulings.  There are typically repurchase clauses in sukuk for regulatory and tax changes, but these may be viewed differently from a Shari’ah perspective because they amount to an unwinding of a previous sale, rather than a commitment to repurchase assets.

However, if the sukuk were subject to a dispute in an arbitration process that allowed Shari’ah input (such as is being introduced through new arbitration rules in Malaysia) then perhaps the structure risk would materialize into potential loss if the repurchase agreement were deemed non-compliant.  Perhaps the bigger risk would be the risk that if the repurchase agreement became non-Shari’ah-compliant, any Islamic financial institution holding the sukuk would be forced to sell, creating a temporary depression in the market price of the sukuk (and generating mark-to-market losses for other holders).

And what can be done about the structure risk without undoing the entire rationale for creating sukuk as Shari’ah-compliant alternatives to fixed income instruments?  After all, if a sukuk moves away from the idea of redemption at par upon maturity and instead is based on market valuations of the underlying assets, then it will more or less cease to function as a fixed income instrument and will become equity.  Investors want the assurance that in all but extreme situations their sukuk will be redeemed at par while Shari’ah scholars want some asset price risk shifted to the investors.

The post-AAOIFI resolution period is instructive here and can offer specific ways that the needs of each can be incorporated in a way that moves away from repurchase at par being specified in the offering documents, but still shifts some risk to investors.  Note, the discussion of what follows represents my opinion from reviewing offering documents, so please forgive any errors.

One sukuk that I have reviewed that was issued post-AAOIFI resolution approached the problem of making the repurchase amount dependent upon market factors, rather than specifying a par repurchase.  During the term of the sukuk, profits were split between the originator and the investors to pay out a return benchmarked to an interest-rate benchmark.

Any profits in excess were placed into a reserve account, which could be used to make future periodic payments if actual profits were below the rate on the sukuk (the AAOIFI resolution also limited the ability for originators to advance qard loans to make periodic payments).  If the profits were insufficient to cover the periodic payments and the reserve account was empty, the payments were made, but the value of the mudaraba was diminished by an equal amount.

At maturity, the originator repurchased the sukuk assets for the value of the mudaraba (i.e. less any shortfalls) with any balance in the reserve account applied against any shortfall between ‘market’ value and par.  The originator kept any excess above par as an ‘incentive fee’.

The other was a perpetual, callable mudaraba, where certain requirements had to be fulfilled in order for the sukuk to be called that limited the likelihood that it could be called at par if the market value of the underlying mudaraba assets were below par.  If there were losses, the sukuk remains outstanding and the originator can miss periodic payments without triggering a default.

Both of these structures incorporates a greater likelihood that the investors could experience losses if the underlying business saw losses or less than anticipated profits, but retain the fixed income return, while at the same time minimizing the likelihood that the investors would receive less than par at maturity.  A similar level of creativity should be taken for other types of sukuk to more explicitly incorporates loss-sharing that is based on the actual value of the underlying assets, rather than just the risk of bankruptcy of the originator (i.e. credit risk).  It will provide lower structure risk, which should in the end be a selling point for investors.

December 23, 2012

No newsletter.

December 30, 2012

Now that it is the end of the year, I think it’s time for a few predictions about next year.  With Yogi Berra’s caveat in mind (“Prediction is very hard, especially about the future”) here are a few predictions.

1) The focus within sukuk markets will continue to shift eastward: During the past two years, the Malaysian sukuk market has continued to maintain a leading position in terms of both primary market issuance  and secondary market issuance.  There is little reason for this to change, absent a dramatic rise in the volatility of the Malaysian Ringgit, the currency in which the onshore sukuk markets are denominated.  The coming year will see a rise in the number of sukuk issued in non-Ringgit currencies but maintain their primary listing in Labuan (with secondary listings in either London or Luxembourg).

2) The debate about Shari’ah standards will return to the forefront of discussion during the year: Whether we revisit the controversy from 2007/08 around repurchase agreements in sukuk documentation, or if the focus intensifies around the use of tawarruq in interbank financing after Oman prohibits Islamic banks from using the product, 2013 will see a return of “Shari’ah risk”.

3) The need for Islamic microfinance will continue to be emphasized, but will not translate into practice: Islamic microfinance has been raised as an important issue for Islamic finance for years, and amidst all the talk, there has been very little action on the parts of Islamic financial institutions to dedicate a share of their resources to promoting Islamic microfinance.  While Islamic financial institutions will continue to promote their commitment to Corporate Social Responsibility, 2013 will not see a significant shift from words to actions on the part of Islamic financial institutions

4) A company from the Americas will issue a sukuk, the first from this region since the 2009 GE Capital sukuk:  There has been very little take off in sukuk issuance in the US and Canada, as well as throughout Central and South America.  In 2013, I expect a company from the region to issue a sukuk.  On the top of my list (in terms of likelihood) would be a follow on sukuk from GE Capital or a company in the oil & gas industry in Western Canada.

5) Islamic financing of renewable energy projects will become more common: The sukuk issued in Malaysia from Australian firm The Solar Guys to finance a solar generation project in Indonesia will serve as a ‘proof of concept’ for a ‘green sukuk’ and with the huge plans for renewable energy generation in the GCC, other ‘green sukuk’ will be issued by GCC-based issuers.

We’ll see at the end of 2013 whether these predictions were right or wrong (and I sincerely hope that number 3 will be wrong).