Financial Times – September 25 2009
Reading bond yield curves is to some investors a high form of financial art. To others, it is akin to reading tea leaves for clues about market direction.
Whichever group you subscribe to, there appears to be a new force in town with the power to change the market dynamics round the world: the banks. They appear to be buying a lot of bonds for their own balance sheets and are likely to continue.
For the tea-leaf dismissives, this still matters because heavy bank buying will affect the absolute level of yields and therefore borrowing costs throughout the economy. For the financial-art believers, it means they have a new factor to get to grips with.
The shape of government bond yield curves is often used as a proxy for the wider market’s subconscious thinking. They provide a convenient horizon-encompassing picture covering both short and long-term views, based on constant and heavy trading among some of the biggest market players out there. If the economic outlook changes, it should be reflected in some point along the yield curve even if the wider market has not yet joined the dots.
Different shapes mean different things. The best known is the inverted curve, where longer-dated bonds yield less (in spite of their increased risk) than near-term borrowing, and which often precedes recessions. The most recent examples include the US in 2000 and, if prematurely, in 2006.
Bull steepeners, where short-term yields fall faster than long-term ones (as now), indicate potential economic growth, whereas bear steepeners, which are caused by rises in long-term rates, suggest the market is pricing in trouble ahead.
Change the long-term buying behaviour of bond market players such as banks, and you potentially change the signals the yield curve is sending.
One of the current puzzles in the markets is how bonds have been rallying in tandem with stocks. Equity rallies based on expectations of improving economic conditions usually equate with predictions of rising interest rates and therefore weakening bonds as yields track rate expectations.
Yet, in spite of the fact that, short of a fresh crisis, the next bond trend will be falling prices and higher yields, investors still seem keen on these low-yielding assets. (Two-year notes yield 0.97 per cent in the US, 1.23 per cent in Germany and just 0.73 per cent in the UK.)
Banks could be one of the reasons for this. With less opportunity to lend, no appetite for risky investments and plenty of liquidity, they appear to be parking the funds in government bonds.
According to research by Steven Major at HSBC, this buying is one of the factors keeping short-dated yields so low and different countries’ yield curves so similar. All G7 curves (except Japan) are very alike in terms of levels and shapes just now. If the low levels were solely caused by the quantitative easing employed by the US and the UK, then the curves of other countries, which haven’t used the same techniques, should be different.
Bank bond buying is unlikely to be a short-term phenomenon either. While new bank capital rules are not yet final, the one thing that seems certain is a general requirement to hold higher reserves, and for much of these to be in high-quality liquid assets such as top-rated government bonds.
The question then is what these new buyers might do to the shape of the yield curve. Currently, curves are historically steep (see chart) because the short-end is held down by official reassurances of low interest rates.
From here, the expectation is that the curve flattens. For curve followers, the key question is whether this will be a bull flattening (short-term rates rising faster than long-term ones, which is what a normalising economy would produce) or a bear, when long-term rates come down because of expectations of lower growth (and consequently lower rates) ahead.
Banks are likely to be buying ultra-short-term paper at the moment. But if their presence is a long-term phenomenon, they would be expected to extend “up the curve” by buying longer-term paper to match their deposits and earn better returns. According to Mr Major, this could lead to an “ironing” of yield curves as bank buying of bonds up to five-year maturities keeps those yields lower and helps hold down even longer-dated yields, potentially damping the entire curve.
Tea-leaf dismissives will class bank buying as a more technical factor, not a genuine economic signal from the curve. Before banks, the 2006 inversion of the US yield curve was often explained by “technical” buying by Asian central banks with excess dollars. But whoever the buyers were, the signal was correct; a recession has since taken place.
Yield curves have a Cassandra quality. But their enviable signalling record suggests that investors ignore them at their peril.