Dubai World Restructuring: Implications of Low Interest Rate

One of my regular readers (or at least I think so), The Real Nick, raised a question to an earlier post of mine on this topic. "Please Sir, get to the point - what's the bottomline? How much less is this "gracious offer" from a de facto defaulter than what the original yield would have been for the lenders?"

A good question. And one that might of interest to others out there among SAM's vast readership. Heck, if Gulf Finance House can claim to have a proven business strategy, then I for sure am on much sounder ground claiming a "vast" readership. I have as the saying in Karachi goes a veritable "war chest" of readers out there.

This post addresses TRN's question and throws in a few other observations.

First, to his question what is the bottom line? How big will the discount be? 

It will depend on two factors. 

1. The original interest rate on the original loan. 
2. The repayment pattern of the rescheduled obligation.  

As to the first, the original interest rate, its effect will depend first on whether the loan or bond was priced on a floating or a fixed basis. For those on a floating rate, the question will be the margin. For any of the technically inclined out there I am ignoring basis risk.

As to repayment pattern, the longer the average life of the rescheduled debt the higher the haircut. Or in other words, If larger payments occur later in the repayment schedule the average life will be longer than if there were equal payments.

ASSUMPTIONS

To do the math we need several inputs.

First interest rates. Not all DW obligations are at the same rate. Not all margins are public. And I'm not inclined to try and estimate an average rate on US$22 billion of debt. So I've arbitrarily picked two rates: (a) 2.0% as a "margin" over floating rate LIBOR and (b) 5.5% as a fixed rate (which is the rate on the Nakheel Sukuk #2). These should be representative enough to give some ideas.

On the repayment pattern, I've come up with three scenarios. To keep things simple I've assumed annual payments of interest and principal in arrears, that is, at the end of the year. It's likely that the banks will want semi-annual payments. At least of interest. 

1. Scenario A: 6 equal annual installments. 
2. Scenario B: Staggered Installments of 0%, 10%, 15%, 20%, 25% and 30%. 
3. Scenario C: Staggered 0%, 0%, 10%, 15%, 20%, and 55%.
   

One more variable required to calculate the cash flow received for interest: LIBOR. Let's assume 1.00% for our "Base" Case. One year LIBOR is around .87%.

For the floating rate instruments, the discount rate will be the margin plus LIBOR. That's 3%.

For the fixed rate instruments, the discount rate is an invariable 5.5%.

A lot of assumptions.  Many if not all of which will be different when the deal is struck.  But, at this point, all that's really required are directional results to get a sense for likely impacts. These working assumptions will also serve to illustrate how the variables work and interact.

Here are the results.

Base Case: 1% LIBOR, Floating Rate Discount 3% Fixed Rate Discount 5.5%

Original Rate	Scenario A	Scenario B	Scenario C
Floating 2.0%	6%	8%	9%
Fixed 5.5%	14%	17%	20%

 

Alternative Case: 2% LIBOR, Floating Rate Discount 4%, Fixed Rate Discount 5.5%

Original Rate	Scenario A	Scenario B	Scenario C
Floating 2.0%	6%	8%	9%
Fixed 5.5%	11%	13%	15%

 

COMMENTS

1. The worst case under the 100% repayment is a bullet at the end of year six. With a 1% LIBOR and the 5.5% and 3% discount rates, that translates into a 22% and an 11% discount. 
2. For all scenarios, as LIBOR approaches 5.5%, the haircut on the Fixed 5.5% instrument reduces dramatically. That's because the discount rate remains fixed. With the Floating Rate, each time the rate goes up we add the 2% margin to it to determine the discount rate. What this means is that to get to a 1% haircut on the floating rate obligations, LIBOR needs to be around 200%. 
3. On an economic basis, the discount rate on both the fixed and the floating should be the same given the same repayment schedule and assuming that recovery is the same, that is, both instruments have the same probability of default ("PD") and same loss given default ("LGD"). 
4. Through the magic of accounting, one instrument (the fixed) is favored over the other. Mathematically, one can construct other scenarios where this favoritism reverses, though the issue with alternatives is the likelihood of their occurrence. As well, in an environment of 200% LIBOR, collecting DW's debt is going to be a rather minor issue among much larger problems. 
5. What this analysis suggests is that the repayment schedule is likely to be a prime driver of the haircut. A schedule with principal repayments weighted to the back-end will result in larger impairments under IAS #39 and thus larger haircuts.
   

IMPAIRMENT TESTS - PLURAL

Under generally accepted accounting principles like IFRS impairment is not a one time event during the life of a financial instrument or asset. Whenever there are signs of potential impairment, the holder must re-evaluate the asset. If further impairment has occurred, an additional provision must be booked. For non equity investments IAS #39 allows the write-up of impaired assets. Thus, provisions no longer needed can be reversed through the income statement.

If indeed DW reschedules on a floating rate basis, then whenever LIBOR changes, the future estimated cash flows change. When the cash flows change, a new impairment test is required.  Since banks may  reverse provisions, if all that changes are interest rates increasing, then some writebacks may be possible.. Over the proposed six-year tenor, fixed rate instrument holders will have the potential for larger writebacks than those holding floating rate paper given the structural factors mentioned above including the current low level of interest rates.

ACCOUNTING HAIRCUTS & ECONOMIC LOSSES

As noted above there is a discrepancy between the IAS #39 mandated haircut for DW based solely on the nature of the pricing on the instrument. That is, based upon its original rate and whether that rate was floating or not.

In the real, non accounting world, economic loss is what matters. Economic loss is not dependent on the pricing convention on the instrument. It is dependent on the instrument's cash flows and risk adjusted discount rate. The latter a function of the holder's WACC adjusted for any additional risk posed by this instrument over its general WACC.

PRICING AND FUNDING

It's also important to understand that on loans or bond rates like LIBOR are pricing references. The holder is not required to obtain his financing at LIBOR or whatever the benchmark is. Nor is it required to match fund the loan or bond. If interest payments are based on six month LIBOR, the holder can fund with shorter date money (daily, weekly, monthly, quarterly) or longer dated money (nine months, twelve months). The funding or "gapping" pattern chosen would depend on the shape of the yield curve.

Nor does making a loan at LIBOR plus a margin guarantee the lender will be able to secure funding at LIBOR. If it has to pay over the benchmark, that is not the borrower or issuer's problem. It is the holder's Similarly, in the case where the holder can source funds for less, it gets to "keep" the difference.

If a bank has a very large base of US Dollar retail customers, its cost of funds may be below LIBOR. On the other hand, small banks and most GCC banks probably can't borrow at LIBOR – but have to pay a premium over that rate. 

What this means for DW lenders is that those whose cost of funds is above LIBOR the "accounting" haircut is going to be higher than those whose cost is lower.   Though it won't be visible as the funding cost difference will be part of the undifferentiated amount of interest expense.