Its all about supply
Credit and swap spreads over benchmark government bonds are influenced by various factors, but one of the most crucial is the relative supply of government and corporate debt. Throughout Europe and the US, Barclays has been warning of a significant widening of spreads this year largely based on our forecasts for better than expected public finances and thus lower government borrowing requirements, and far higher than expected levels of corporate bond issuance.
Government supply is dwindling...
In the US for instance, estimates of the governments fiscal surplus and the likely size of future debt buy-back operations have been revised higher all year, and the lack of a strong mandate by the eventual US president will limit the degree to which future budget surpluses can be whittled away by fiscal largesse. (If he wins the presidency, we would expect Bushs planned 2% of GDP annual fiscal easing to be reduced in size to a 0.5-1.0% easing due to the balanced nature of the Senate). Europes strong cyclical growth recovery has helped boost previously weak budgetary positions: the budget deficit in the Eurozone is forecast at -0.3% of GDP this year, down from -1.3% in 1999 and -2.0% in 1998.
In the UK, we expect the government to post a GBP35 billion budget surplus this year, followed by surpluses of GBP15 billion and GBP10 billion over the two subsequent fiscal years. Improving public finances has been a factor behind the continued inversions of the US and UK benchmark yield curves throughout the year, while the Eurozone curve is flatter than would otherwise be expected given the current combination of above trend GDP growth, above target inflation, and an ECB that has a credibility problem.
..but corporate supply is surging - blame Schumpeter and shareholders
In contrast, the finances of the US and European corporate sector are in many cases looking stretched as leverage levels grow. Throughout 2000, corporate debt issuance has increased sharply. Two primary factors explain the changing supply dynamics of corporate Europe and corporate US. Firstly, event risk has emerged as a word that strikes fear into the holders of credit product. This is the growing tendency for corporates to increase leverage to finance M&A activity or share buy-backs if their equity performance is laggardly.
In essence, shareholder value is being pursued with a growing disregard for bondholders. The problem has been compounded by the growing threat of technological obsolescence, whereby the rapid advance in new technologies, notably internet-related, is forcing some companies to leverage themselves and expend into new businesses in order to prevent a rapid erosion of core businesses. We are witnessing the credit market implications of Shumperterian creative destruction.
3G technology is also a culprit
The shifting supply dynamics between government and corporate debt has been exacerbated by the series of auctions in the Eurozone of licenses for third generation (3G) mobile telephones. The license sales have essentially created a transfer of wealth from the corporate to the public sector. Successful bidders have been forced to increase leverage to finance their license and the subsequent upfront investment in new technology, while the government has gratefully received the payments. The UK governments budget surplus this year has been boosted by the EUR38 billion it raised from auctioning off 3G licenses, while the German government received a windfall of EUR50.8 billion. The widening of telecom spreads in Europe has therefore led a broader rise in credit spreads.
Asia is different: shareholder value?
Asia has been immune from many of these trends in 2000. Since the region is still in the midst of a reform programme aimed at deleveraging corporate sectors and creating a greater emphasis on profitability and efficiency, the pursuit of shareholder value has not been a driving force behind debt issuance in most countries in the region.
Event risk in Asia remains the collapse of companies that have failed to deleverage quickly enough - Daewoo, Hyundai - rather than companies wishing to issue debt in order to create M&A activity and enrich shareholders. Similarly, Asia has not yet to begun its round of 3G license auctions.
Other factors have driven spread widening in Asia
The absence of G7-style event risk and 3G-related issuance shocks has meant that the trend towards wider credit spreads (primarily on USD debt given the immaturity of many countries' local bond and swap markets - has been driven by other factors: a contagious reaction to wider credit and swap spreads in the US and Europe, reflecting how emerging markets are risk assets and thus tend to trade in tandem with the trend on the bellweather US risk markets; a reassessment of previously over-optimistic views towards the stability of Asias cyclical growth upswing in 1999 and into 2000.
Divergent spread trends - Singapore and Japan contrast US and Europe
The widening of spreads in Europe and the US this year has been pronounced, reflecting the double-whammy of diminishing government supply of debt and increasing corporate supply. The 10 year USD swap spread, for instance, has widened from 76 bps at the start of 2000 to 115 bps. The contrast with Singapore and Japan - countries which have among Asias most developed credit markets - is stark. In Japan, a budget deficit of around 10% of GDP and weak demand for capital in the corporate sector - a function of a large output gap measuring around 4% of GDP - means that government issuance swamps corporate bond issuance. Ten-year JPY swap spreads for instance measure just 24 bps. Meanwhile, in Singapore, debt issuance trends have favoured the government and overseas companies (which, due to FX regulations, are forced to swap SGD proceeds of debt issuance back into USDs and this bias swap spreads tighter) rather than local corporate bond issuance. Ten-year SGD swaps are currently just 22 bps wider than the equivalent Singapore Government Securities (SGS), down from over 40 bps at the start of the year and the interest rate swap curve is flat from five to 10 years.
Times, they are a-changing - in Singapore at least
In the US and Eurozone, we do not foresee the basis for a marked recovery in swap and credit spreads into 2001. Public finances remains healthy, and in the US the lack of a strong popular mandate for whichever candidate wins the presidency will preclude aggressive fiscal easing. (As mentioned above, we are only assuming a 0.5-1.0% of GDP easing of fiscal policy in the US over the coming years, which poses far fewer risks to the economic upswing or the medium-term health of public fiances than the original expectations of Bushs 2% of GDP easing.)
Meanwhile, event risk remains acute and has led to deteriorating credit quality even before the effects of a slowdown are fully felt. The US has also yet to auction its 3G licenses. In Japan, it is difficult to forecast change in current issuance dynamics. The long-awaited and increasingly vital period of fiscal reform is being delayed by the LDP until such time as a sustainable economic recovery is underway, and while current signs of growth are encouraging, there are clearly risks that this will be yet another false dawn. However, trends are likely to change in Singapore, whereby the balance in issuance will swing clearly towards domestic corporate issuance in 2001.
Singapores 3G license auction - and growing competition for capital
The changing supply dynamics in Singapore into 2001 will hinge on three trends.
- We expect Singapores February auction of four 3G license sales to increase corporate debt issuance.
- Net government debt issuance will fall.
- Even aside from any 3G-related debt issuance, we forecast growing competition for capital among local firms and thus growing local corporate debt issuance.
These trends will reinforce our forecast of a significant rise in SGD swap spreads over the coming months, which we expect to push the 10 year SGD swap rate widen to 50 bps. Spread trends in Singapore will therefore grow more reminiscent of Europe and the US rather than Japan.
Shifting issuance dynamics #1 - a 3G auction shock?
Singapore kicks-off the 3G auction process
The Singapore government has announced that it will sell four 20-year 3G licenses next February. A minimum price of SGD150 million per licence has been set. This prompts the question of whether we will see the raft of corporate debt issuance seen in Europe and which has been so instrumental to a sharp widening of Telecom credit spreads this year. In determining the likely success of Singapores auction, the experience of Europe offers insight.
Europes experience - expect the unexpected
For some rough-and-ready estimates of the likely auction results, Singapore can be compared to those countries which sold their 3G licenses via auction. (France decided on a beauty contest, which reduced the price received by the government for each license and effectively saw the public sector subsidise Telecom companies to the tune of EUR10.8 billion.) However, the experience of Europe is highly variable suggesting that Singapores auction will be an uncertain and unpredictable affair.
In Germany and the UK, for instance, the license auction significantly overshot private sector and government estimates. The German 3G auction raised EUR50.8 billion while the UK raised EUR38 billion. In contrast, the Italian auctions were something of a damp squid, with the EUR12.16 billion raised around half of the governments expected EUR21 billion. Also, worryingly for Singapore, smaller European countries have generally seen disappointing auctions, with the Netherlands raising just EUR2.5 billion and Austria an even lower EUR706 million.
Europe's 3G auction results
|Source: Barclays Capital|
Europes experience deflated for Singapore
Cross country comparisons are by nature highly suspect. However, to get an idea of what Europes experience means for Singapore we have decided to deflate the auction results by the relative size of Singapores GDP. We favour this approach to other analysis such as comparing auction results on a license fee/per capita basis since we wish to get a sense of what value the market will be willing to place on a countrys 3G market, and hence the comparison is not influenced by the number of licenses placed for auction.
For instance, Singapores economy is just 5.2% the size of Germanys, implying that Singapores auction result will be 5.2% of Germanys EUR50.8 billion or EUR2.64 billion, which when converted into SGDs is SGD3.94 billion or SGD985 million for each of the four licenses. The table below uses this methodology to deflate Europes auction results for Singapore.
The European auctions deflated for Singapore:
Deflated by GDP
SGD per license
|Source: Barclays Capital|
* This is below the reserve price of SGD150mn a license
Europe therefore creates some widely differing examples. The UK and German experience would suggest an auction process that will draw out bids of around SGD1bn per license, while the Netherlands suggests a figure of SGD247mn. Austria implies a figure below the reserve price of SGD150mn. The question then becomes which of the examples is more valid for Singapore?
A Teutonic Singapore...
There are some factors that suggest that Singapores license may be closer to the SD1bn mark than the reserve price. Auguries of a successful auction include:
- Singapore is likely to encounter a fair degree of competition for the four licenses. Having been fully de-regulated ahead of schedule this April, Singapores telecom market is already highly competitive. The government-linked Singtel now faces keen competition from two Telecom consortiums, MobileOne and Starhub, both of which technically enjoy the support of partners with deep pockets. MobileOne includes Telstra, Cable and Wireless and PCCW, while Starhub includes BT and NTT. Incumbent Telecom companies have tended to be compelled to participate in the auctions given the fear that existing businesses will become obsolete (that man Schumpeter again) if they fail to take the gamble on 3G. (Leverage or die as it were.) As for non-incumbent participation, companies with pretensions of developing a pan-Asian strategy are likely to require a Singapore operation. Hence, Deutsche Telekom, Vodafone, and albeit less likely France Telecom, may participate, while Hutchison could also have a dabble following its success in the UK auction and participation in Italy. The Hong Kong mobile phone company, Sunday, has already said it will participate in the auction. In short, a fair degree of competition is likely during the auction process.
- Singapore is likely to be characterised by a market with low roll-out costs. Establishing a 3G network will be reduced by Singapores population density, whereby 4 million people live in an area barely 30 miles by 20 miles. Moreover, average per capita income is among the highest in the world, while the current penetration rate for mobile telephones is 66% compared to Europes average of around 43%. Moreover, 72% of current 2G Telecom services is post-paid, a system that lends itself more readily to 3G services than the pre-paid system.
- Telecom companies will feel more confident in committing funds to finance the investment of 3G technology in Singapore by the governments recent history of compensating firms for any unexpected loss of a monopoly position.
...or an auction with Austrian characteristics?
Naturally enough, there are some competing factors that suggest a smaller auction result is in the offing:
- Funding constraints. While on paper Starhub and MobileOne include backers with strong financial resources, in reality the financial backing may be less certain. Starhub may suffer from the fact that BT may be looking to exit the consortium as part of its debt reduction plans, selling this and other Asian assets (although NTT would still be expected to provide strong backing). MobileOne may suffer from C&Ws policy of exiting all mobile businesses globally, (although again, Telstra is likely to remain a strong backer). More significantly, if the major consortium partners are not willing to guarantee the debt of Starhub and MobileOne and thus force the companies towards non-recourse lending, they may be placed at such a large financial disadvantage that they cannot compete with Singtel. After all, with a cash pile of SGD5.8 billion Singtel can afford to pay for a license without raising debt, but without recourse lending the cost of capital for MobileOne and Starhub will be high if not astronomic, meaning that they could only bid low for a license. Of course, the UK example saw rival companies ratchet up the bidding process before dropping out to place rival Telecom companies at a competitive disadvantage in subsequent auctions. However, the financial constraints may limit the ability for Singtels rivals to bid at high levels.
- Purely and simply, the experience of Europe suggests that small markets such as Austria and the Netherlands, have failed to deliver large bids. With an economy of around USD100 billion and a population of 4 million, Singapore is obviously a small country.
What is the likely issuance stream?
The experience of Europe, allied to the conflicting auguries from Singapore, shows how projecting an auction result is a difficult task. While on balance, the arguments for a successful auction appear stronger, it is difficult to predict a UK or German outcome of SGD1 billion a license simply based on the experience of smaller European countries. Indeed, the UK and German auctions appear to have been the exception rather than the rule in Europe. By the same token, an auction result in excess of the reserve price is likely given that in many respects, Singapores market is more attractive than those smaller European countries through dint of low roll-out costs, and unusually high penetration rate for 2G technology. Putting a guestimate on price per license, we would say that a level around SGD500 million or close to Italys example, is likely, with the risk being to a higher number.
Using a working estimate of SGD500 million a license, we would estimate that maximum SGD debt issuance would total SGD1.5 billion given that Singtel can easily afford to finance its license from cash. (Whether SGD500 million is right or wrong, at least we believe that the value of these licenses will be right. Unlike the early bidding for the UK license, by February 2001, Singtel will have trialed 2.5G and NTT DoCoMo will have accumulated significant experience from i-mode, so product and revenue potential from 3G should be clearer.) Of course, not all of the cost of licenses would be financed via SGD bond issuance or indeed raised in SGDs at all. However, given that we would anticipate an upside risk to the final auction result, we will assume - remember, this is by nature a rough-and-ready estimate! - SGD1.5 billion worth of fresh SGD corporate bond issuance.
Obviously, compared to the major bond markets of Europe or the US, an increase in issuance of this magnitude would not register. However, it is worth remembering the size of Singapores bond market. SGD1.5 billion in fresh issuance would amount to around 5% of total outstanding corporate bonds and 3.7% of total SGS issuance. A proportional increase in corporate bond issuance in the US would be a USD141 billion increase in issuance. (A model of US swap spreads created by our global fixed income research team would suggest that an increase in corpoate bond issuance of this magnitude would widen 10 year USD spreads by over 7 bps.) Before we conclude later on in this article the likely impact on spreads of changing issuance dynamics, we can at least say that the 3G auction process will provide one element of a switch towards corporate rather than SGS issuance into 2001.
Shifting issuance dynamics #2 - government issuance
Creating debt for debts sake
One feature of the Singapore bond market is that in many respects, the entire SGS market is an artifice. Public finances in Singapore are among the strongest in the world, with the government enjoying an average budget surplus of 2.3% of GDP over the past five years, despite the regional financial crisis. The decision to issue debt has been largely a function of the governments desire to deepen Singapores capital markets by creating a benchmark yield curve. Hence, SGS issuance is a matter of policy rather than a function of need.
One clear implication is that the 3G license auction in Europe will not have the two-way effect on issuance it has in Europe: the transfer of wealth from the private to the public sector associated with the 3G auctions will increase corporate debt issuance, but is unlikely to influence existing SGS issuance plans. 3G will at most be a single whammy for SGD spreads.
Net SGS issuance to slow - a switch to qualitative market development
Determining SGS issuance trends therefore involves double-guessing MAS policy. On this score, we are confident that net SGS issuance into 2001 will decline. Our view is based on the assumption that MAS has already achieved its goal of developing a liquid and deep benchmark yield curve and is instead likely to focus more on improving the depth and efficiency of the credit markets and of trading activity in the SGS market.
The issuance of SGS has indeed been rapid in recent years as the government has tried to create a liquid yield curve. In the first nine months of 2000, net SGS issuance measured around SGD7 billion, which pushed total outstanding SGS to SGD40.7 billion. The market is extremely liquid, and an established system of market makers has been developed. However, MAS is not able to rest on its laurels since Singapores bond market is still in need of a substantial degree of development.
Corporate bond issuance has been below government issuance, to the extent that as of September, the stock of corporate SGD debt measured SGD28 billion, or less than 70% of the capitalization of SGS. One clear impact of the large issuance of SGS relative to corporate debt, is that credit and swap spreads have been biased tighter, towards levels which are impeding the development of the market. At present, the SGD swap curve is flat between five and 10 years, demonstrating no trade-off between duration and credit risk, while SGD issuers are generally believed to be able to raise funds at spreads too tight for their credit risk, much to the chagrin of the local buyers of SGD debt looking for higher yield.
Credit arbitrage adds to the woes
The problem can be better expressed in numbers. Between January-September, total net SGS issuance measured around SGD7 billion. Corporate bond issuance over this period amounted to SGD10.3 billion. However, of this amount SGD2.2 billion was SGD issuance by overseas companies taking advantage of Singapores low yields to undertake credit arbitrage activities. MAS regulation 757 compels the proceeds of SGS issuance to be swapped back into foreign currency, and hence this SGD2.2 billion actually creates receiving interest in the SGD swap market and tightens spreads in the same way that SGS issuance does.
What next - reduce SGS issuance, tackle the spread problem
Singapore clearly needs to improve the efficiency of the local bond market. At the end of the day, all pretensions the government may have towards creating an internationally significant bond must be discounted as long as there is no trade-off between duration and credit risk on the SGD swap curve since this questions the ability of the market to price credit.
It is the development of the sophistication of the market rather than increasing its size that we feel will dominate policy going forward now that the SGS benchmark curve is liquid. (We remain extremely optimistic on the long-term prospect of the SGD credit market, and hence assume that the current anomalies of the market will be addressed.)
Continuing the rapid increase in SGS issuance will actually hamper the development of the market by biasing spreads tighter, and so we would expect a slowdown in the pace of fresh issuance. In this vein, the government has announced plans to buy-back part of the 30% of SGS which are off-the-run, in order to maintain liquidity in the benchmark issues.
A debt buy-back programme will start on 15th November, which will reduce SGD1 billion in off-the-runs from the market. The buy-back programme is in our view indicative of a shift away from rapid expansion of the market towards policies aimed at improving its quality. In total, we estimate that net SGS issuance next year will be closer to the SGD6 billion to 7 billion mark than the SGD9 billion to 10 billion originally slated for 2000 (excluding buy-back operations).
Other policy approaches
On a related point, our expectations that the government will focus on developing the quality of the Singapore bond market rather the size of the SGS market, suggests that other policy shifts may in store, many of which will bias swap and credit spreads wider.
- To prevent SGD bond issuance by overseas companies from automatically biasing spreads tighter, and indeed, to allow greater trading of Singapore interest rates, one obvious area for reform is to gradually increase the internationalization of the SGD via easing restrictions on MAS regulation 757.
- To create a more efficient SGS benchmark curve - the current curve is too steep considering Singapores record of having one of the lowest trend rates of inflation in the world, with the 10 year average rise in the CPI measuring just 1.9% - the government may support changes to the trading activities of the market. The creation of a strengthened repo market would facilitate short-SGS positions and thus curve flattening trades, while hedging activity - and curve trades could also be supported by creating a futures contract based on SGS. By supporting a flatter curve, the prospects for wider long-term credit and swap spreads would be supported.
- The current 18% reserve requirements in Singapore may be reduced or adjusted in such a manner that will rescue the heavy and automatic buying of SGS at the front-end of the curve and which biases front-end yields lower. As Singapores financial system is exposed to greater foreign competition, an incremental process of lower reserve requirements will be required to allow for higher returns on equity and profitability from local banks.
Wish-list contents vs. issuance dynamics
Naturally, we do not forecast any of these changes. They remain part of a wish-list rather than any projection of future policy changes. However, we have listed some of the changes we would like to see just in the sense that if we are correct in our assumption that the government will focus on qualitative development of the SGD bond market, then changes of this nature cannot be fully ruled-out. As it stands, however, the one firm conclusion we feel able to draw is that we expect net SGS issuance to be around SGD4 billion lower in 2001 compared to 2000.
Shifting issuance dynamics #3 - competition for capital
Singapore corporate issuance to pick-up
One area where the government may be supported in its efforts to develop the SGD bond market, and widen back-end swap spreads, is that we expect a cyclical pick-up in local bond issuance in isolation of any 3G license related issuance.
Growth without demand for capital
To date, a rapid pick-up in economic growth - real GDP is forecast to expand by 9.0% in 2000 - has not been associated with a significant increase in the demand for capital. Bank lending, for instance, rose just 3.4% y/y in September (and fell 0.4% m/m), and in value terms is still 1.1% below the recent peak in late-1998. The fact that a growth recovery has been combined with such limited demand for capital has been one of the key factors behind our long-standing buy recommendation for the back-end of the SGS curve. Several factors explain the limited demand for capital in a rapidly expanding economy.
Exports are driven by FDI
Firstly, and crucially, Singapores economic upswing has been driven by the export-oriented manufacturing sector, which is in turn dominated by foreign companies. (Non-oil domestic exports rose 18.8% y/y in September on a three month moving average basis.) At present, foreign companies account for 77% of all fresh investment in the manufacturing sector and as a rule do not fund capital formation from SGDs.
Subdued investment outside of manufacturing
Secondly, there has been little in the way of fresh investment in the more domestically focused industries. The retail sector has emerged from a multi-year downturn, for instance, with retail sales rising 20.6% y/y in August, but the sector remains dominated by excess capacity. This explains why high retail sales numbers have not created upwards pressure on prices - retail sector deflation remains the order of the day. Similarly, excess capacity continues to undermine the construction sector. In Q2, Singapores GDP rose 8.0% y/y, while the construction sector experienced an output contraction of -9.8% y/y. The weakness of SGD issuance by local corporates relative to SGS issuance is a function of this limited demand for capital.
But domestic capital formation starts to rear its head
However, there are encouraging signs that the demand for capital among local companies is starting to pick-up, as strong levels of economic growth lends support to capital formation, albeit with a lag. For instance, with Singapores crisis-related output gap set to close this year, a pick-up in fresh investment activity among local firms would be expected, even accounting for excess capacity in some sectors of the property market and retail sector.
For instance, while bank lending is still far below levels normally associated with 10+% nominal GDP growth in Singapore, there is a clear trend towards a gradual recovery which we expect to persist into 2001. The 3.4% y/y rise in bank credit in September, while sluggish, was at least up from 2.9% in August, and continues a recovery from a 20 month period of y/y declines stretching to this April.
Indeed, while SGD corporate debt issuance has been weak considering the strength of aggregate demand in Singapore, there has nevertheless been obvious evidence of rising levels of growth influencing bond issuance. For instance, 1998 saw just SGD4.2 billion worth of SGD corporate debt issuance, at a time when real GDP rose by just 0.3%. In 1999, economic growth recovered to 6.5%, and we saw a commensurate rise in SGD corporate debt issuance to SGD9.2 billion. In 2000, with real GDP growth forecast at 9%, SGD corporate bond issuance measured SGD10.3 billion between January-September, of which SGD8.1 billion is local company SGD debt issuance.
Moreover, Q4 looks set to maintain the uptrend with SGD1.7 billion worth of local corporate debt issuance seen in October alone. Additionally, investment commitments by local manufacturing sector firms rose 6.6% y/y in the 12 months to June, an improvement from -27.8% y/y a year earlier. As such, the recovery in bank lending has been mirrored by higher corporate debt issuance, as one would expect it to.
An uptrend rather than a surge in corporate borrowing is expected
Of course, we do not expect a dramatic pick-up in domestic investment. We believe that the capital intensity of Singapores current upswing will be lower than pre-crisis given the overhang of excess capacity in some sectors, and by the fact that Singapores current focus on economic liberalization and most notably the gradual easing of government influence in major local corporates (or Government Linked Companies) will increase efficiency in the economy and thus lower the capital intensity of growth. (Given that levels of savings are expected to remain reasonably constant, one manifestation of this trend will be a higher than usual current account surplus relative to GDP - 20.7% of GDP in Q2 - than would be expected following the closure of the output gap.)
However, a steady uptrend in bank lending towards the expected 7% to 8% growth rate of nominal GDP next year is likely, and with it increased SGD corporate debt issuance. At present, local corporate SGD bond issuance is on course to measure SGD12 billion, and simply assuming the trend seen in H2 2000 is maintained, into 2001 we would conservatively estimate this to rise to around SGD17 billion, excluding any 3G related issuance.
Put it all together and what have you got - wider SGD swap spreads
Putting together the three factors - 3G license sales, lower SGS issuance, and higher local corporate debt issuance - and we have the prospects for a quite pronounced shift in the relative supply of debt. Using data between January-September (because Q4 is already showing some of the trends we are forecasting, such as higher corporate debt issuance and MAS debt buy-backs) we showed that SGD6.7 billion worth of SGS issuance was counterpoised with SGD8.1 billion in local corporate debt issuance and SGD2.2 billion in foreign SGD debt issuance, which due to regulation 757 acts to tighten spreads. It is therefore no surprise that spreads tightened over this period.
Into next year, however, issuance will bias spreads wider. We are assuming net SGS issuance of SGD6 billion. Against this we would expect SGD17 billion worth of local corporate debt issuance and SGD1.5 billion of 3G-related issuance. This therefore leads to a supply imbalance of SGD12.5 billion in favour of corporate debt.
Of course, if we do not see the repeal or softening of regulation 757 then SGD bond issuance by overseas entities will again create receiving interest and thus tighten spreads. If we conservatively that foreign SGD bond issuance will amount to 25% of total corporate issuance, then the net issuance of local corporate debt over SGS and foreign SGD issuance will remain SGD8 billion. (This difference amounts to 27% of the entire size of the corporate bond market as of October.)
Will this move spreads?
Quantifying the impact of this degree of issuance on spreads in Singapore is where the analysis gets woolly. The market is too underdeveloped and lacks a sufficient time series for us to create the type of models used to look as US and European markets. Our global fixed income strategy team has a model of US swap spreads that shows that a USD20 billion rise in the stock of corporate bonds relative to government debt widens USD swap spreads by 1 bps.
Only as a purely indicative exercise can we apply SGD bond trends to major market models adjusting for the relative size of the markets. For instance, the small size of the SGD bond market means that an SGD8 billion increase in local corporate debt relative to SGS is in relative terms equivalent to USD754 billion increase in the stock of corporate bonds relative to US Treasuries, enough to widen spreads by 38 bps according to Barclays USD swap spread model mentioned above!
Naturally enough, the less developed SGD bond market means that we cannot simply extrapolate the results of the US model to the local markets and look for a SGD8 billion increase in the stock of Singapore corporate bonds relative to SGS and foreign SGD bonds to widen 10 year SGD spreads by 38 bps. However, the conclusions do serve an important purpose in terms of highlighting how the shifting issuance dynamics into 2001 will act to significantly widen SGD swap and credit spreads, in a complete contrast to the trend seen throughout most of this year.
10 year SGD swap at 50 bps
In our eyes, the shifting issuance dynamics in Singapore add fresh validity to our forecast for wider SGD swap and thus credit spreads. The SGD swap curve remains quite simply too flat and too tight to SGS, and as for the 10 year SGD swap spread, we continue to advise paying the swap and buying 10 year SGS, looking for the spread to widen to 50 bps over the coming months. The changing issuance dynamics form another in a growing list of factors that we expect to widen spreads and steepen the SGD swap curve:
- We are willing to bet on MAS developing an internationally significant bond market - or else we would not be writing research on the market. This means we would expect the current anomalies in the market such as a flat SGD swap curve in spread terms to be a temporary event that the authorities will address, whether it be via debt buy-backs, encouraging higher levels of corporate bond issuance, adjustments to regulation 757, or the development of an SGS repo market that facilitates curve flattening trades and thus help widen spreads. To assume that the spread between two and 10 year SGDs swaps will remain flat over the medium-term is effectively stating that Singapores credit market will not develop.
- International trends. While a current account of over 20% of GDP insulates Singapore against international yield trends - MAS does not need to be overly concerned about maintaining a constant yield differential with the USD in order to prevent the SGD from weakening - an international trend towards steeper swap curves will however provide support to the uptrend we anticipate for Singapore. Our forecast of wider swap spreads and steeper curves in Europe and the US into 2001 reflects our view that the year will see continued corporate bond issuance outpace government issuance. We envisage the 10 year USD swap rate, for instance, rising from current levels of 115 bps towards 130-140 bps over the next three months.
We therefore continue to look for a significant widening of SGD swap spreads in Singapore, which will feed through to wider credit spreads, a trend that will ironically help the market develop by creating a better pricing of credit risk. The shifting issuance dynamics in Singapore are therefore likely to prove one of the spurs to this move. Thus, we reiterate one of our favoured strategy recommendations in Asia: pay the 10 year SGD swap and buy the 10 year SGS looking for the spread to widen from 22 bps to 50 bps. Even if the trade is not viewed as compelling as a macro-strategy in its own right, it is worth looking at by holders of SGD credit as an effective way to hedge a credit portfolio going into a climate of widening spreads amid a pronounced shift in issuance.
Desmond Supple is Head of Research, Asia at Barclays Capital. Email: [email protected]