Srejber: Low interest rates and narrow credit spreads – a conflicting story?
I would like to begin by thanking you for inviting me to talk to you here today. I shall use this opportunity to talk about risks in the international financial system and I shall do so by discussing a very specific aspect, namely the simultaneous existence of low long-term interest rates and narrow credit spreads. The reason I have chosen this particular aspect, is that the way you look upon this combination leads to totally different conclusions concerning the risk currently inherent in the financial system. But let me start by saying something about how credit spreads normally tend to develop over time.
Credit spreads reflect the premium demanded for buying risky assets instead of risk-free governmental bonds. The lower the quality of a security is, the higher should the compensation be in the form of higher yield. These spreads change with the outlook of the economy. In good times, when companies’ balance sheets are strong and the perceived overall financial risk is low, credit spreads tend to narrow. During bad times, with slumping economic performance and possibly financial turbulence, yield spreads widen out as the real as well as the perceived risk is higher. Wider spreads simply mirror investors demand for a higher premium in order to assume risk. There seems to be what could be called a spread cycle. What has this to do with the current low long-term interest rates?
Falling long-term interest rates are often seen as the financial markets signalling the coming of a slow down in global economic growth. The thought is built on the notion that the risk of inflation due to strained resources diminish in an economic slowdown. That means, in turn, that the premium that investors demand for investing in longer term bonds in order to cover the risk of inflation gets lower. However, if this hypothesis is right the very same markets seem to be signalling a conflicting story through historically narrow credit spreads. As I said, spreads tend to get wider when the economic outlook gets gloomier. Evidently, that is not what has happened. Instead spreads seem to be anchored firmly on historically low levels. The downgrading of Ford and General Motors in May apparently did not have any profound impact on investors risk appetite since the junk bond market has recovered since then. Using credit spreads as a guide to the future; the economy seems to be as healthy as ever. So how should we interpret these conflicting signals?
Lasting effects
One interpretation is that the low long-term interest rates are not signalling a slow down at all (1). Instead, the low rates are the outcome of effects related to central banks and globalization.
During the last 15 years or so we have seen the coming of independent central banks with a clear mandate to achieve low and stable inflation. The Swedish central bank is such an example. There is convincing empirical evidence that these banks have been successful. That definitely supports the argument that the equilibrium level of long-term interest rates could have become lower than they used to be.
The arguments concerning globalization concentrate on the opening of markets to fiercer competition. This has, together with the integration of China and India into the global markets of goods, lead to a downward pressure on consumer prices. We have also seen how technology, deregulations and institutional changes have had profound effects on the financial market. They have all helped to lower transactions costs and vastly expand the possibilities to share financial risks, which has led to lower costs for investments. These changes have most likely influenced the inflation-related risk-premium demanded by investors and thereby led to lower nominal and real long-term interest rates. The inclusion of countries like China and India into the market economy may also have had the effect that global growth has become steadier. Global growth used to be running on three engines - US, Europe and Japan. With the entry of new countries into the world markets, the risk for a global slowdown from engine failure probably has become smaller. That means in turn that the risk for a slump in the demand for goods and services also has become smaller, and hence that lending to the corporate sector in general may have become less risky. On the other hand, the heightened competition may very well imply that the risk that a specific company will run into problems has not diminished. However, on the whole this development could have contributed to lower the risk-premium and thereby to lower real interest rates.
Transitory effects
No doubt, there is a lot of substance in these arguments. On the other hand, low interest rates could also, at least to some degree, be the result of effects that are of a more short-term nature. For example, the low rates in the US have partly been explained by high demand for Treasuries from central banks in Asia and oil-producing countries. Should that demand lessen it would naturally have the opposite effect on interest rates. Furthermore, it has been noted that pension funds have become more concerned with liability management than asset management. The reason behind this change is said to be the under-funding of retirement plans that has forced pension funds and insurance companies to raise the share of long-term bonds in their portfolios. When these portfolio shifts are finished, the effect will subside.
Yet another possible short-term effect has to do with the hard cost-cutting programs in the corporate sector that followed in the wake of the stock market’s sharp fall from the heights in the year 2000. The fall in corporate investments meant that one important source of competition for capital disappeared. With fewer securities to invest in, but an unchanged demand for portfolio investments, interest rates will fall. But sooner or later the need for new real investments in the corporate sector should surface. Competition for capital will rise again, leading to higher interest rates.
Anyway, if the effect of the structural changes following from inflation targeting central banks and globalization are strong enough, there is no problem to reconcile low long-term interest rates with narrow credit spreads. But there is also another possible explanation, and it is based on what Alan Greenspan, among others, has called the search for yield (2). The term is somewhat confusing since you have to take it for granted that market participants constantly are searching for yield. What it intends to capture, though, is the notion that investors for some reason have come to under price financial risks.
A global saving glut
The search for yield-interpretation normally starts with the rise in private savings in developing and emerging market countries in the wake of the financial crisis during the 1990’s, such as those in Mexico and East Asia. Since capital outflows had been devastating for many of the countries that were exposed to the crises, some of them started to build up huge currency reserves as a buffer against future outflows. Foreign exchange interventions in order to promote export-led growth added to the fast accumulation of reserves in some countries. The change was reflected in fast-growing current account surpluses. The surge in oil prices resulted in growing surpluses in the oil producing countries. These changes effectively turned many developing and emerging-market countries from being net importers to become net exporters of financial capital. The result was what the former Federal Reserve employee Ben Bernanke has called a global saving glut.
These savings were accommodated by the US and some other industrialized countries. The rising demand for these countries sovereign debt drove down the risk-free interest rates. But the most prominent feature, according to this story, was the rise of the stock markets. From the mid 1990’s unto the bursting of the tech bubble in 2000, stock markets played the key equilibrating role in interna-tional financial markets. As the holders of these stocks in the developed countries saw their wealth rise they lowered there savings with growing current account deficits as a result.
The stock market crash and the concomitant fall in investment opportunities could have put an end to this, but it did not. When companies started to cut down on costs in order to restore their balance sheets, their demand for capital to invest fell. In order to equilibrate the market for global saving, real interest rates fell with the effect of a surge in housing prices on the one hand and a search for yield on the other.
Low compensation for risk
With interest rates on government bonds already low, the story continues, investors tried to reach for higher returns, by moving into riskier assets like commodities and complex credit products (3), without demanding full compensation for that. This development may have been facilitated by the success of quantitative risk models like Value-at-Risk, which build on historical volatility in returns and correlations between different asset classes. Volatility on corporate bonds has been low for a number of years and the correlation between these bonds and risk-free assets is less than perfect, which makes them a logical alternative for inclusion in a diversified portfolio.
Accordingly, there seem to be at least two possible ways to explain the simultaneous existence of the low long-term interest rates and narrow credit spreads we see today. The interesting difference between these two explanations, at least from a central banker’s perspective, is that they lead to totally different conclusions concerning the current systemic risks on the global financial markets. The first interpretation, which sees low long-term interest rates mainly as the outcome of structural changes leading to a lower inflation premium, is one of a stable situation. In the second interpretation low long-term interest rates could indeed be a signal of a coming economic slowdown. That spreads do not widen out, is explained by an intense search for yield leading investors to under-price financial risks. The proponents for the second explanation see the activities of pension funds, insurance companies, private equity investment companies and hedge funds as a symptom. I will dwell a little on the latter.
Expansion of the hedge fund industry
Hedge funds are normally valuable vehicles on the financial markets. When they seek out arbitrage opportunities, they increase the efficiency of the markets by helping to correct any mispricing of financial assets. By helping other financial institutions to offload risks, hedge funds do provide an important source of risk transfer and diversification. But this time, at least according to the search for yield-interpretation, hedge funds contribution the efficiency of the market may have been partly off set. The massive expansion of the hedge fund industry could actually have reinforced the mispricing manifested in historically-low credit spreads.
According to the Hennessee Group, a global adviser to hedge fund investors, the number of hedge funds has doubled, to 8,000, since the year 2000. The amount these funds control has tripled, to $1 trillion, during the same time. And as Bank of England has noted in its Stability Review, inflows into hedge funds have continued at record levels, despite relatively modest returns recently shown. In order to generate returns in the face of the low interest rates offered, the fiercely competing funds may have increased their involvement in less liquid markets and in new asset classes, for example commodities and energy. Furthermore, it seems that many of them are financing an increasing part of their investments with short-term loans. That is worrisome. Low liquidity assets and high leverage could prove to be a dangerous combination if interest rates rise or spreads widens out.
Potential triggers
There are currently some potential candidates that could trigger the broadening of spreads. Higher interest rates following on from a large adjustment of the dollar exchange rate caused by the unwinding of the US current deficit is one such candidate (4). If that should materialize, investors will demand a compensation for the currency risk. That compensation will come in the form of higher interest rates on US bonds. So far, there is no apparent sign of a sustained decline in capital flows into the US. But it seems clear that the financing of the country’s current account deficit partly hinges on the willingness of central banks to accumulate further dollar assets (5). A serious hint that these central banks may lower their dollar holdings is naturally a strong incentive for other investors to reduce their dollar holdings to. The effects of a sharp dollar depreciation would not only affect the search for yield, but will most likely also have a negative impact on global economic growth through trade adjustments.
Similarly, a hike in investment rates around the world will probably lead to higher interest rates by shrinking the current “saving glut”. That seems highly possible. After years of cost cutting, the need to invest may be rising in the private as well as the public sector.
Another potential trigger could be inflation data that is worse than expected. That may not only lead to higher long-term rates, but also higher monetary policy rates. It seems that financial markets have largely priced in a moderate and gradual monetary tightening in the US, while the ECB is expected to stay put for some time. Should higher prices on oil and other commodities spill over to consumer prices, this fairly calm scenario may change quickly.
However, the negative impact on credit spreads might not stem solely from rising interest rates. If rising prices on oil and commodities do not spill over into inflation, they will feed into companies cost of production and weaken their financial base. Likewise, a rise in the American saving rates, maybe caused by a fall in housing-prices or a markedly slow down in the Chinese economy, would have a negative impact on global growth and hence on corporate earnings. Earnings disappointments relative to market expectations would probably occur and lead to wider credit spreads. Should this in turn lead to a downgrading of companies, spreads would widen further.
Higher long-term interest rates and wider spreads will also affect the numbers obtained from quantitative risk models. If this led to the unwinding of investment positions based on expectations of low or gently rising rates, we will probably see corrections in many asset markets. This is especially problematic in a high leverage market.
Conclusion
The current combination of historically low long-term interest rates and narrow credit spreads may well be explained by successful inflation fighting by central banks and a benign macro economic environment. Unfortunately, there may also be a less comfortable explanation causing concerns about the financial risks that are currently being accumulated. This explanation focuses on the existence of an intense search for yield in the meaning that market participants are under-pricing financial risks and thus driving down interest rates and credit spreads to unsustainable levels.
The rapid expansion of the hedge fund industry in recent years could be seen as a symptom of this search. The heightened competition among funds combined with few high-yield opportunities may have weakened the funds’ function as gardeners weeding out mispricing on the financial markets. Instead, there are some signs that hedge funds may have contributed to such mispricing by indulging themselves in less liquid markets. Likewise, it seems that many globallyactive banks have developed new lines of business in response to the low margins in more established financial markets. The expanding market for structured credit products along with a rising interest in prime brokerage activities, as well as energy and commodity trading, may therefore be the result of an escalating search for yield. These activities could have enhanced the vulnerability to unexpected market developments, a problem that may be further enhanced by the wide-spread use of quantitative risk models like Value-at-Risk models.
VaR-models normally try to compute the likely market risk based on the assumption that historical volatility and correlation patterns will hold in the future as well. Less volatile assets or assets that are not closely correlated will reduce a portfolio’s risk. Recently, volatility in many financial markets has been below their historical averages. These numbers are directly fed into the models and may thus influence the perceived risk borne by the company, giving it a green light to take on more financial risk. The problem is that in a financial crisis volatility rises while correlations tend to go to one. The latter means that prices on different assets change in the same direction more or less simultaneously. In other words, when you really need the risk-lowering effect of diversification it may not be there. That calls for serious stress testing of the models, like for instance to see what happens to the risk of a financial portfolio when the correlation between the assets are set to one.
One should also bear in mind that these models originally where designed to limit trading risks, which means that they should keep track on risks for very short time horizons, up to a few days. Today, they are widely used by the management of companies in order to estimate risks during a whole financial year. This is done by scaling up the values you get for the very short time horizons, which may not always be relevant for longer periods. Furthermore, most VaR-models are built on simplifying statistical assumptions. The essence of that is that they do not take into account the risk of very bad outcomes. One can only hope that such weaknesses are clearly understood by decision makers on different levels, should the current combination of low long-term interest rates and narrow credit spreads turns out to be the riskier alternative of the two explanations discussed today.
Notes
(1) For a general discussion about the business cycle see IMF World Economic Outlook, April 2002.
(2) See for instance Alan Greenspans remarks to the International Monetary Conference in China in June this year (published on the Federal Reserve’s home page).
(3) The development of the structured credit market is discussed in BIS Quarterly Review, June 2005.
(4) For a discussion of the magnitude of such a change see “Obstfeld, M & Rogoff, K, “ Global Current Account Imbalances and Exchange Rate Adjustments”, May 16, 2005.
(5) This issue is discussed in Roubini, N & Setser, B: “Will the Bretton Woods System 2 System Unravel Soon? The Risk of a Hard Landing in 2005 – 2006”, February 2005.
