Thursday, December 10, 2009
A Closer Look at Dubai's Debt
December 10, 2009 By Mohamed Jaber Dubai & Paolo Batori London This is a summary of UAE: A Closer Look at Dubai's Debt, December 7, 2009. The Dubai government's announcement on November 25 that it intended to restructure the debt of Dubai World and its subsidiary Nakheel took investors by surprise. Although the markets' initial sharp reaction has since subsided, significant uncertainty remains due to the dearth of information regarding the scale and scope of the emirate's debt problems. This event has particularly shone the light on Dubai's fiscal standing against the background of its significant public sector debt. We attempt to assess the extent of this debt and its potential burden on the emirate's finances. We also evaluate the current pricing of Dubai's credit in relation to its peers based on current market valuations. Government Debt The direct non-bank liabilities of the Dubai government currently amount to about US$18.7 billion. Prior to 2008, these debts were not economically significant as they were limited to a US$0.41 billion dollar bond. However, since April 2008, the government has raised an additional US$18.3 billion worth of non-bank debt. Little is known about the uses of these funds, other than that about US$15 billion of the newly raised debt has been earmarked to the Financial Support Fund (FSF). One of the FSF's central roles is to provide financial support and liquidity on a commercial basis to the government and the government-related entities (GREs) that are deemed to be of strategic and developmental importance to Dubai. In fact, its establishment back in July had served to further blur the line between the government's explicit versus implicit support for GREs. Overall, Dubai's direct and contingent debt liabilities are estimated to stand at around 50% of GDP. In both absolute and relative terms, this level of debt is not necessarily a cause for concern and is lower than the regional average. However, one also needs to consider the potential contingent liabilities of the Dubai government. Given the strategic significance of some of its GREs, it is difficult to see the government leaving them to fend for themselves. As such, we estimate that the government's implicit guarantee of these institutions would add contingent liabilities of about US$20 billion to its existing debt burden, thereby raising its debt (both direct and contingent) to GDP ratio to about 50%. Again, this level, although high, is not alarming by regional or even international standards. The market's appetite for Dubai's sovereign debt very much depends on how the current situation is handled. Going forward, we believe that the Dubai government will need to tap credit markets for additional funding. This may be necessary in order to finance development expenditures and shore up the standing of some of the emirate's most strategic GREs. As a result, we expect the government's direct, non-contingent debt to increase by about 27pp to 51% of GDP by 2011. However, the ability of the Dubai government to raise additional debt on somewhat favorable terms will depend on how successful it is in managing the current debt restructuring process. It is difficult to over-emphasize the need for a timely resolution of Dubai World's debt problems. A protracted negotiation process that leaves creditors with a significant loss would not be in the emirate's interest. Neither would be the lack of timely disclosure of other potential debt challenges affecting Dubai's GREs. We believe that a clear and well-communicated government strategy to deal with the debt situation is imperative at this stage in order to quell rampant market speculation and limit the long-term reputational damage to the emirate's credit standing. Moreover, ring-fencing the troubled debt of various GREs would help to limit the impact of the current crisis on entities that have a solid business model and are able to shoulder additional debt. Quasi-Government Debt There are no official estimates of the debt owed by Dubai's GREs. Calculating the overall size of their debt liabilities is further complicated by a number of factors, including: (i) the complex organizational structure of Dubai's quasi-sovereign holdings; (ii) the lack of transparency regarding operations of its mostly unlisted companies; and (iii) the lack of data on bilateral, non-disclosed loans extended to these entities. Nevertheless, we rely on publicly available information to derive an estimate of the value of the GRE's outstanding bonds and loans. We then adjusted these estimates in an attempt to account for non-reported debt. In total, we believe that the debt of Dubai's GREs currently stands at around US$89 billion, or 116% of the emirate's GDP. The debt is split among the emirate's three largest holding companies, with Dubai World and the Investment Corporation of Dubai accounting for the largest shares, although the latter arguably holds a number of companies with high franchise values. The GRE's disclosed debt is also held by a diverse group of investors, with about one-third held in bonds and the rest in loans. International banks hold about 44% of total GRE debt, versus 13% for UAE banks. Moreover, about 24% of this debt is set to mature by end-2010, with another 24% maturing in 2011. In sum, Dubai's public sector debt - which includes that of the government and its related entities - is estimated at around US$108 billion, or 140% of GDP. However, these estimates may overstate the real debt burden on the public sector, since they don't take into consideration the assets held by these GREs, on which there is unfortunately little information. It is likely that other GREs will also announce debt restructuring plans over the near term. The high leverage of many of Dubai's government-related entities and their significant exposure to real estate and financial assets that have underperformed since 1H08 make it likely that further restructuring of GRE debt may be needed. The absence of detailed financial statements for these companies makes it difficult to accurately assess their financial soundness. Nevertheless, we tried to derive a guesstimate of the value of the debt that may need to be restructured by first examining the GRE's outstanding obligations and then subjectively assigning a probability of restructuring to them based on anecdotal evidence. In the process, we developed three different scenarios based on the increasing likelihood of debt restructuring of the GREs: Scenario 1: Assumes that only the recently announced Dubai World debt will be rescheduled. Scenario 2: Adds to scenario 1 the assumed rescheduling of the debt of Istithmar, Drydocks, Dubai Financial Group, Dubai Holding Investment Group (holding company level) and Dubai International Capital (holding company level). Scenario 3: Adds to scenario 2 the assumed rescheduling of the debt of Dubai Holding (holding company level), Dubai Holding Commercial Operations, Bourse Dubai and Dubai Sukuk Center Limited (DIFC). Dubai's Fiscal Position Given the attention afforded to Dubai's debt overhang, it may be useful to examine its fiscal standing and assess the impact of this debt burden on its finances. Dubai's fiscal revenues are dependent on fee and oil income. As of 2008, tax income - mainly related to customs fees and limited corporate taxes (e.g., tax on foreign bank income) - accounted for no more than 23% of overall government revenues. Conversely, proceeds from oil exports made up around 26% of government revenues. Non-customs fees - including those levied on road use and real estate - accounted for about 45% of total fiscal revenues. The latter are estimated to have dropped by about 15% in 2009, mainly due to the: (i) significant decline in Dubai's real estate sector in 2009; (ii) slowdown in domestic spending; (iii) negative population growth; and (iv) lower oil prices. We expect fiscal expenditures to decline in 2010. Government expenditures, which had expanded at an average annual rate of about 48% during 2007-08, are estimated to have increased by about 10% in 2009. We believe that the government's finances will likely be strained next year due to tighter funding channels and a continued need for government support for some of the emirate's most strategic GREs. As such, we expect public spending to decline by about 6% next year. Current expenditures, which make up about 60% of government spending, should drop slightly in 2010, while development expenditures will likely experience a sharper decline on the back of tighter financing. On balance, we expect the fiscal accounts to register a deficit of about 4.5% of GDP in 2009 and 2.9% in 2010. The burden of debt-servicing will likely increase over the medium term. We also attempt to assess the impact of servicing the debt on the emirate's finances. In order to perform this exercise, we assume that: (i) the government will only be responsible for servicing its direct debt obligations, not those of its GREs; (ii) the government will need to raise an extra US$20 billion during 2009-11 to finance expenditures and shore up the finances of highly strategic GREs; and (iii) the average cost of financing will remain below market levels, mainly thanks to subsidized financing from the federal government and the emirate of Abu Dhabi. Assuming a full roll-over of debt at maturity, debt servicing expenses are expected to rise from about 7% of total expenditures in 2009 to around 22% in 2011, somewhat in-line with other oil-importing Middle Eastern countries. If we were to take into consideration the government's contingent liabilities vis-à-vis its most strategic GREs, the toll of servicing the debt on government finances would necessarily be higher. Estimating the Fair Value of Dubai's Credit Goals of Our Scenario Analysis The goal of this analysis is to give a quantitative frame of reference, to reconcile different markets (corporate, sovereign and global EM markets) in a context of high volatility and uncertainty. Therefore, investors should be aware that our conclusions have to be considered as indications of fair value, based on the below-listed assumptions, which could reveal themselves as strong in some peculiar market situation. Given the anticipated debt restructuring and the recent volatility in Dubai fixed income markets, we undertook a macro/debt scenario analysis in order to: 1) assess if Dubai 5Y CDS is in line with a diversified sample of other emerging markets countries based on straightforward credit metrics (in this case external debt/GDP); 2) assess if Dubai CDS and the Nakheel bond are pricing in a homogenous market/debt restructuring scenario; 3) determine a short-term fair value range for Dubai CDS; and 4) determine a medium-to-long-term target range for Dubai CDS. Methodology and Assumptions First, we regressed the total external debt/GDP ratio and the 5Y CDS level for 33 emerging countries to, broadly, estimate fair value levels for Dubai implied by our macro scenarios. Second, we used the three different debt-restructuring scenarios outlined above to estimate the magnitude of external debt burden going forward. Scenario 1 assumes a debt restructuring of US$26 billion. Scenario 2 assumes US$34.7 billion and Scenario 3 US$46.7 billion. Third, given concerns among investors relative to the effective value of the implicit sovereign protection of GREs, we have analyzed a debt scenario which sees Dubai ring-fence its debt and the debt of its most strategic entities - what we term the Last Line of Debt restructuring scenario (LLD). Fourth, we have assumed several haircuts on any potential restructuring. Assumptions: 1) The relationship between total external debt and GDP remains stable over time. The R-squared of our binomial regression is 0.56. 2) The sample of emerging countries is representative of the EM world as a whole. 3) External debt/GDP is one of the main drivers of spreads and can be considered as a good proxy of credit risk. 4) Gross external debt/GDP, we find, is a good proxy of perceived credit risk. Net external debt/GDP may have been a better gauge, had it not been for the lack of data. Conclusions and Strategy Implications Assuming a scenario of 50% of haircut (which is also what seems to be priced in for the Nakheel 2009 bond), we draw the following conclusions for the 5Y Dubai CDS: • The current level looks in line with the haircut projected by the Nakheel 2009 bond and the rest of the EM sovereign 5-year CDS spectrum; • The CDS spread is currently positioned above the upper limit of the short-term interval of confidence (based on our macro and restructuring assumptions). Therefore, it offers value if our baseline scenario (scenario 2) and our assumptions were to materialize. In which case, it could converge towards 300-350bp in the short term; and • If we assume that, at the extreme scenario, Dubai will take a tough line with investors exposed to the quasi-sovereign assets (LLD scenario), the CDS spread should converge towards 175-225bp in the medium-to-long term. However, if this situation were to materialize, we cannot rule out an increase in short-term volatility, as investors' willingness to finance that particular emirate/region could temporarily fade.