Monday 19 April 2010

Dubai World Restructuring – Why Interest Rates Matter


You may have seen the news reports that DW and its creditors were haggling over the interest rate on the rescheduling.  DW reportedly having offered 1%.  And its creditors (or some of them) insisting on 5%. From the context it seems that this is a fixed rate and not a margin over a benchmark like Libor. I can't imagine that Reuters would confuse "margin" with "rate".

Of course, both parties are negotiating so it's probably safe to bet that neither side is expecting to get what it is asking for.

What do these levels mean in terms of a haircut? What's the appropriate discount rate to use for a net present value calculation?

Let's turn to the second question first.

There are two discount rates we can use. 

The creditors' proposed rate of 5%.  This rate is most likely designed to minimize/eliminate the haircut that will arise from an IAS #39 impairment test due to the new rate being lower than the contractual rate.  (Recall the assumption is that lenders are carrying their DW exposure at historical cost.  If they're not, then IAS mandates they discount estimated cashflows based on a market rate.  Also we're making a critical assumption here that the banks are estimating they will receive all their original principal back plus interset).  Why this focus on an accounting haircut as opposed to looking an economic value?  Banks will be most concerned about accounting losses as these are the most visible. 

How do we estimate the economic (real) haircut?  We can use a recent market rate. DEWA just raised a US$1 billion five-year bond at 8.5%.  Some caveats.  It's important to note it's highly likely that the DEWA bond was deliberately priced above market. Why? First of all, Dubai cannot afford to have a major transaction fail in the market. So the bond has to be priced to ensure success. Second, the issue was reportedly 11x oversubscribed. That in itself is a pretty good indication that the price could have been tighter. Of course, Dubai would not only like the deal succeed but also to have a very strong display of market appetite for its paper for this highly visible post crisis issue.  So another reason to overprice.  While the rate might not necessarily be the right market rate for DEWA credit,  the Dubai World rescheduling is not a BBB- credit like DEWA. So we can argue that 8.5%  is probably a more reflective market rate than the 5% rate and thus a better measure of economic loss.

It's important to note that neither rate is completely accurate.  But we're not after what is fundamentally elusive precision here as much as we are interested in a directional measure of the haircut.

Now that we have our discount rates, we need cashflow scenarios.

The actual proposed cashflows on the rescheduling aren't known so we'll use two proxy "worst" and "best" cases to set bounds - much as we did with the interest rates.

First, a bullet repayment at the end of the repayment term with interest at 1% being paid annually. This represents our worst case from a calculation of discount.   While this would be the ideal debt restructuring from DW's perspective, it's highly unlikely this could be sold to creditors.

Second, equal amortization of principal with interest at 1% being paid annually. This represents our best case. The ideal case from the lenders' perspective (absent of course an even better case of a generous neighboring Shaykh paying off the entire debt today).  But not very realistic and certainly not what DW would want.  It is highly likely that restructuring principal amortization will be back ended as is typical in debt restructurings. 

And finally to conform to the details of the restructuring – we'll model a 5 year and 8 year tranche.

To summarize, these two scenarios give us a reasonable approximation of the best and worst discounts. Results are rounded to the nearest percentage. The discount can be figured by subtracting the NPVs shown below from 100%.

Bullet Principal Repayment 1% Interest Rate Paid Annually in Arrears

Discount Rate5 Years8 Years
5%83%74%
8.5%70%58%

Equal Principal Amortizations 1% Interest Rate Paid Annually in Arrears

Discount Rate5 Years8 Years
5%89%85%
8.5%81%74%


On the 5 year tenor, the banks are trying to avoid an accounting "haircut" of between 11 and 17%. And on the 8 year tranche between 15% and 26%. 

From an economic perspective the haircut is even more 19% to 30% on the 5 years and 26% to 42% on the 8 years. As stated above, the banks are going to be more concerned about the visible accounting haircut. 

These are fairly crude approximations - the compound effect of uncertain cashflow patterns and inexact discount rates - but they give an idea of what is at stake here.

No comments: