BIS Quarterly Review, December 2003

The BIS Quarterly Review released today is divided into two parts. The first analyses recent developments in financial markets, financing flows in banking and debt securities markets, and activity on derivatives markets. The second part presents four articles on topics related to debt securities markets. The first of these proposes an explanation for why corporate credit spreads are much wider than expected losses from default; the second analyses common factors in emerging market spreads; the third characterises the market for sovereign credit default swaps; and the final article suggests a way to unify government bond markets within individual countries in East Asia.

International banking and financial market developments

Signs of growth boost investor confidence

So strong were the signs of a global economic recovery in October 2003 that they surprised even already sanguine investors. Yields rose, equities resumed their upward climb, and corporate and emerging market spreads narrowed. Sentiment towards emerging markets was further boosted by an unprecedented series of credit rating upgrades. In October alone, 10 sovereigns were upgraded, mostly in East Asia. Such confidence about the outlook had not been so widely shared in September, when investors in government bond markets had seemed disappointed by the macroeconomic data.

In contrast to other markets, foreign exchange markets saw unusually sharp movements, with the US dollar coming under intense downward pressure after a G7 meeting in late September. Significantly, the fall in the dollar did not lead to a rise in US bond yields, a scenario some investors had feared.

Although signs of potential concern emerged in some markets, these were isolated events. The downgrading of two automobile companies highlighted vulnerabilities in this volatile sector of the corporate bond market. The arrest of one of Russia's well known business leaders raised questions in investors' minds about the country's new status as an investment grade borrower. And allegations of improprieties in the mutual fund industry threatened for a time to undermine the optimism of equity investors.

The international debt securities market

A sharp fall in net borrowing by euro area entities was responsible for reduced fund-raising through the international debt securities market in the third quarter of 2003. In spite of narrower spreads, aggregate net issuance fell by 14% to $299 billion. Turbulent market conditions appear to have played an important role. Gross issuance was particularly weak in August, a month that saw heightened interest rate volatility. However, the decline proved to be transitory, with such issuance rebounding in September to its second highest level ever.

The decline in fund-raising by euro area entities was partially offset by borrowing by others. Emerging market borrowers came back to the market, notably Turkey and Venezuela, two countries which received rating upgrades in July. Borrowing by US financial institutions was also relatively strong, helping to push net issuance of US dollar-denominated securities to $138 billion. This is the first time in two years that net dollar issuance has been at least as large as net euro issuance.

Derivatives markets

The aggregate turnover of exchange-traded financial derivatives decreased in the third quarter of 2003. The aggregate nominal value of trading in contracts monitored by the BIS amounted to $223 trillion, a 9% decline from the second quarter. Activity was uneven among the major market risk groups, with turnover in interest rate contracts falling substantially and that in stock index contracts growing modestly. A pronounced drop in money market contracts accounted for the overall decline in interest rate contracts.

In the over-the-counter (OTC) derivatives markets, activity accelerated in the first half of the year. The latest BIS semiannual data on aggregate positions in these markets place the total notional amount of outstanding contracts at the end of June at $170 trillion, a 20% rise from the end of 2002. The previous increase had been 11%. During the first half of the year, gross market values rose by 24% to $7.9 trillion, compared with a 43% increase in the earlier period. Gross market values have been expanding at a more rapid pace than notional amounts since 2001.

The international banking market

In the second quarter of 2003, flows between banks again overshadowed flows to corporate and other non-bank borrowers. This rise in interbank lending was largely the result of increased repo activity and inter-office claims. While credit granted to non-bank borrowers was also robust, it reflected investments in international debt securities, primarily public sector debt. As in the previous four quarters, credit in the form of loans to corporate and other non-bank borrowers remained subdued.

Movements in deposits drove the net flow of funds into emerging markets, which remained positive for the second consecutive quarter. However, differences across regions were substantial. A large net inflow to the Asia-Pacific region was mainly due to deposit repatriations by banks in China. At the same time, repatriations by banks in Russia were behind a net inflow to the emerging European countries, the second largest in five years. Conversely, increases in deposits in reporting area banks, especially by banks in Brazil, contributed to the fifth consecutive net outflow from Latin America.

Special features

The credit spread puzzle

A major puzzle is why spreads on corporate bonds tend to be much wider than would be implied by expected losses from default. For example, the average spread on BBB-rated bonds in the last five years was about eight times the expected loss. Efforts to explain the puzzle have suggested that taxes, risk premia and liquidity premia can account for the spreads. Jeffery Amato and Eli Remolona, both of the BIS, review these explanations and offer a novel alternative.

The authors point out that actual losses in a corporate bond portfolio often exceed expected losses. Previous explanations of the credit spread puzzle have assumed that investors can diversify away the risk of such unexpected losses. However, Amato and Remolona show that the skewness in the distribution of corporate bond returns calls for an extraordinarily large portfolio to achieve such diversification. They present evidence from the market for collateralised debt obligations suggesting that such large portfolios are unattainable. Hence, investors always face the risk that actual losses will exceed expectations. The authors argue that spreads are wide because they compensate investors for such risk.

Common factors in emerging market spreads

Emerging market bonds have become an important asset class for portfolio managers. A striking feature of the spreads on these bonds for different countries is that they tend to move in tandem over time. Patrick McGuire of the BIS and Martijn Schrijvers of the Netherlands Bank investigate the extent to which these spreads can be explained by just one or two factors that are common across issuers.

Focusing on a sample of 15 emerging market issuers, McGuire and Schrijvers find that common factors account for an average of one third of the total variation in the daily movements of the various spreads. A single common factor explains approximately 80% of the common variation. This factor seems to reflect primarily changes in investors' attitudes towards risk.

Sovereign credit default swaps

Contracts on sovereign debt constitute a growing part of the market for credit default swaps (CDSs). Frank Packer of the BIS and Chamaree Suthiphongchai of the Bank of Thailand are able to characterise the sovereign segment of the market by examining data covering several years of quotes and trades from an important CDS inter-dealer broker.

Packer and Suthiphongchai find that trading in sovereign CDSs is concentrated in fewer names than is the case for corporate or bank CDSs. They also uncover a concentration of contracts in reference assets of relatively short maturity, apparently a result of the fairly high proportion of low-rated sovereigns that tend to issue at short maturities. Finally, the authors discover that premia on CDSs written on very low-rated sovereigns tend to be significantly wider than those written on correspondingly low-rated corporates. The authors suggest that this phenomenon is consistent with the market being less sure about the consequences of sovereign default.

Unifying government bond markets in East Asia

The Asian crisis of 1997 taught policymakers in the region two main lessons. First, to build up foreign reserves. Second, to develop local bond markets. Robert McCauley of the BIS highlights an interesting link between the two goals.

When central banks in the region accumulate foreign exchange reserves from capital inflows, they also sterilise the inflows; that is, they sell debt in the domestic market to prevent the bank reserves from expanding. McCauley shows that central banks with large reserves end up issuing their own debt for this purpose after selling off government debt from their portfolios. He argues, "The fact that these obligations have commonly taken the form of central bank debt, however, has meant that much of the opportunity - to develop the bond market has been missed".

McCauley proposes the conversion of central bank debt into government debt even if it means "overfunding" by the government. With a single big borrower, the debt would trade in a larger and more liquid market. The author then explains how this could be achieved by Indonesia, Korea, Malaysia, Taiwan (China) and Thailand, and what the implications would be.