Global: The Term Premium - A Puzzle Inside a Riddle Wrapped in an Enigma

Sir Winston Churchill may well have been mistaken for a market strategist talking about the term premium rather than the foreign policy of the former Soviet Union. Bond yields and term premiums have stubbornly refused to come off their lows in spite of rising policy rates, strong growth and rising inflation.

The enigma: Why are bond yields so low? Will they stay low? Our proprietary model MS-FAYRE suggests that US 10-year Treasury yields should be at 5.5% — a whole 100 bps above their current levels (Fels and Pradhan, Fairy Tales of the US Bond Market, July 2006). If historical relationships are still valid, then the 17 increases in the fed funds rate and rising inflation expectations should have led 10-year rates higher. In this sense, 10-year rates have stayed too ‘low’, well below the fair value predicted by MS FAYRE and models of other researchers. The fact that the 10-year rate has stayed flat despite the tightening of monetary policy led ex Fed Chairman Greenspan to dub the enigma of low bond yields a “conundrum”.

A direct implication of the conundrum is that some factor(s) must be exerting downward pressure on the 10-year rate equal in magnitude to the upward pull from the factors included in fair value models.

Yield curves in the US and euro area have flattened dramatically since mid-2004, with the 10y-30y segment flattening more than the 2y-10y flattening would warrant. This is about the time that pension funds started buying long-dated bonds to cover the shortfall in duration from their liabilities. As the bonds with the highest duration at the end of the curve get richer, demand shifts to earlier points on the curve, with yields falling in inverse proportion to the duration of the bond.

Asian central banks have accumulated reserves far in excess of import requirements. Given the risk profile of these institutions, government securities in the US and Europe have been obvious destinations without much sensitivity to the price. Their sustained presence in bond markets is likely to have kept bond yields low (see Mutkin, Guzzo, Pradhan, Dec. 2006).

The impact of these factors also has implications for the estimates of term premiums from quantitative models.

The Riddle — Why has the term premium fallen? Will it stay low? The term premium estimated from the FAYRE model and from the model of Fed researchers Kim and Wright (2005) are highly correlated even though they use very different methodologies (see Fels and Pradhan, July 2006). These robust estimates suggest that the term premium has fallen since the 1980s as part of the Great Moderation, and has stayed close to zero during the 'conundrum' period. What accounts for such low term premiums, and will they continue to stay so low?

The potential solutions to the 'conundrum' may partly explain why the estimated term premium has been pushed so low. These recent forces have pushed bond yields lower than they would have been otherwise. What remains beyond the effect of the standard factors, i.e., the term premium, will be a much lower number than would have been the case.

However, the term premium riddle is not so easy to solve. The term premium can itself fall for fundamental reasons, taking yields lower with it. Thus, we are left with a ‘signal extraction’ problem: did yields fall because of special factors, or did they fall because of a decrease in the term premium? The answer probably is a bit of both! The term premium on the 10-year Treasury rate has been very well correlated with the MOVE index, an index of implied volatility on options on Treasuries. Intuitively, lower volatility in the market for Treasury securities implies investors will receive lower compensation for the reduced uncertainty, pushing term premiums and yields lower. The decline in volatility is at least part of the answer to the riddle of the low term premium. Finally, if interest rate volatility and the term premium are related, then volatility coming off its lows could mean a similar movement in the term premium and yields.

The Puzzle — Why has volatility fallen? Will it stay low? The correlation between interest rate volatility and the term premium leaves us with a final puzzle: What could have moved interest rate volatility to its current lows, and what could a trigger a rebound? Implied volatility on interest rate options tends to reflect periods of uncertainty regarding interest rate decisions, macroeconomic events or technical factors.

There are indications, however, that the current bout of low volatility can be attributed to a much-talked-about and less easily defined force — excess liquidity coupled with increasing integration of financial markets. The intuition behind excess liquidity can be extended to refer not just to the easy availability of credit and loose financial conditions, but also to the willingness to use these conditions to enter into leveraged positions seeking returns. This 'search for yield’ has led investors from asset class to asset class, bidding up prices and sustaining investor interest in spite of adverse news and events. The willingness to take on and maintain positions has made prices less pervious to shocks and news, i.e., it has lowered volatility across asset classes. As a result, spreads have compressed and implied volatility measures across interest rates, equities and currencies have moved to near-historic lows together.

However, if excess liquidity is reason enough for volatility to plunge, then its withdrawal should be reason for it to surge. Even though some measures of excess liquidity suggest a withdrawal of liquidity is at an advanced stage, no tell-tale surge in volatility has taken place as yet to confirm this link.

Excess liquidity is difficult to define and equally difficult to measure. Metrics that view excess liquidity via its price (i.e., interest rates) or its quantity (measures of money stock) may provide different answers. Quantity measures of excess narrow money in the G5 set of countries suggest that liquidity is no longer excessive (see Fels, Turn of the Liquidity Cycle, May 2006) . However, measures for broader definitions of money don't give the same answer, suggesting that financial conditions may still be easy. While our proprietary natural rate of interest models suggest that interest rates are at neutral in the euro area and above neutral in the US, this is definitely not the case in Japan or even China. Another proprietary measure of G-10 interest rates relative to their long-term mean (Stephen Jen, Charles St-Arnaud, Aug 2006) suggests that interest rates still have a distance to go to reach ‘normal’ levels. The upshot is that there are indications of a broad-based withdrawal of liquidity from the world economy, but this withdrawal has not yet become ubiquitous.

With US rates unlikely to push below neutral, and most other central banks looking to normalize interest rates (i.e., bring them in line with neutral rates), it seems but a matter of time before the withdrawal of excess liquidity shows up in most metrics, and most importantly in the availability of credit and the attitude of investors. The end of easy financial conditions in 2007 could put the puzzle together, sending volatility higher. Term premiums may follow, solving part of the riddle. Finally, the increase in term premiums would take yields higher, providing at least partial relief from the 'conundrum' of low bond yields.