Fixed Income? Part II.
Jarrod Wilcox, April 4, 2004

Tinder for an Interest Rate Conflagration

Preface

A recent column (March 12, 2004, Part I of a three-part series on Interest Rate Risks in 2004) described a scenario of economic imbalances that may have been created by a) the shock of increasing productivity within emerging markets that are resulting in increased low-priced imports of goods and services into the US together with b) the stimulative efforts of the Federal Reserve to combat unemployment and prevent deflation.  The most obvious symptoms of this scenario have been low short-term interest rates, a weakening dollar, low inflation despite strong corporate profit recovery and a huge government budget deficit, and lagging employment recovery.

Recently, Federal Reserve Chairman Alan Greenspan has sounded optimistic about the imbalances created.  For example, he notes that consumer balance sheets have enjoyed increased asset valuations to go along with their increased consumer and mortgage debt.  Nevertheless, for reasons given in more detail below, I believe investors should not be lulled by interest rates that continue low for longer than the consensus has expected, but should prepare for the real possibility that there will be a sudden and substantial increase in bond interest rates even greater than now forecast by a forward rate curve that is upward sloping within the 3 to 10 year range.  Late last week a favorable US jobs report triggered a sharp jump in interest rates, which may prove the opening round in a scenario of much more extensive increases.

In this column I will describe in more detail the imbalances that I believe have stored up the potential for such a crisis.  As a forest fire is made possible not only by a careless camper or a random lightning strike but by the previous storing up of deadwood and undergrowth followed by a period of dry weather, so, too, an interest rate crisis requires preconditions.  We don’t need to know the trigger that will set off an interest rate surge – it could be a rather small event.  But if we understand the imbalances that could cause vicious circles that result in a bigger movement, we may be able to take more intelligent precautions.

We won’t consider such long-term and wide-ranging sources of instability as 1) the build-up in Social Security liabilities relative to sources of support or 2) international security tensions potentially requiring major defense and foreign aid funding.  We have to draw the line somewhere.  However, there are more immediate possible sources of fuel for the interest rate fire.  Here are three key examples – one each from the domains of real interest rate risk, inflation risk, and credit/liquidity risk.

1. Foreign Creditor Buildup – Real Interest Rate Risk

One imbalance that could fundamentally shift the supply of funds for debt investment that has been discussed on and off for the last decade is the growing amount of US debt held by foreigners.  This is a consequence of low US savings and an excess of imports over exports that provides foreigners with dollars to invest.  The big current investors are the Japanese, who are fighting to keep their export-led possible economic recovery alive.  The chart below shows only Treasury debt; the totals would be much bigger if agencies and corporates were included. 

Just because no significant interruptions of willingness to invest further have occurred doesn’t mean they can’t do so.  Note that the Chinese and British, though large holders, seem to have begun to decrease the share they take of increases in US debt.  The preconditions seem to be a dollar so weak that it discourages further purchases of US bonds, especially for the holding of currency reserves, combined with the cessation of strong exports to the US by trade surplus nations.  The first condition may be close to being satisfied; the second only requires the success of increasing protectionist sentiments in the US Congress and the lessening of Japan’s need to buy US bonds.

2. Energy --Inflation-Risk

Energy costs are one of the prime drivers of US inflation rates, both directly and indirectly, and they are importantly dependent on external supplies whose prices in dollars go up, albeit in a smoothed way, as the dollar declines versus other currencies.  If that weren’t enough, consider what will happen if China and similarly less-developed but rapidly industrializing countries consume an additional 10 million daily barrels of oil from the world’s current 78 million. If effect, there would be less supply available for the US.  The following chart shows what can happen to energy prices at the consumer level when there is a major supply disruption, such as in the 1974 and 1980 Mid-Eastern energy crises.  The slopes of each of the two lines, one for consumer energy prices and the other for all other components of the consumer price index, reflect the rate of inflation.  Not only did energy prices go up by a factor of three during that period, but they seem to have had a long-run effect on prices of everything else in the US. 

Note that the slope of the non-energy prices accelerated following the energy price increase accelerations, and has gradually decelerated over the last ten years as energy prices followed a flatter line.  The evidence is circumstantial, and surely there have been many other influences on the rate of inflation, but it is suggestive.  Now note that energy price levels has increased far less than other prices since the mid-1980’s, and we are facing a) rapidly growing petroleum usage by China and others, b) declining rates of oil field discoveries, and c) increases in Middle-Eastern political instability.

3. Excessive Leverage Creates Liquidity and Credit Risks

Low short-term interest rates make the business of borrowing short and lending (or investing) long more profitable.  The additional profitability encourages a variety of decisions that result in greater debt leverage.  This can result in both liquidity and credit risk.  Perhaps the clearest example in 2004 is associated with the financing of US housing.  It exists at two levels: 1) credit risk due to a housing construction and price bubble, 2) communicable liquidity risk from Fannie Mae and Freddie Mac, which are the focal point of most mortgage securitization and a large degree of mortgage holding.

Is there a housing bubble?  The answer, clearly apparent in the statistics available from the Office of Federal Housing Enterprise Oversight, the agency that watches Freddie Mac and Fannie Mae, is yes.  The chart indicates that it is already considerably bigger than that experienced in the late 1980’s before the recession of 1990-1991.

Suppose we assume that increasing government oversight of Fannie Mae and Freddie Mac will prevent them from overextending themselves.  So far, I believe this has mainly consisted of requiring them to maintain constant duration as changing interest rates interact with the rapid changes in effective duration in underlying mortgages as pre-payment risks are approached and diminished.  But this merely passes the problem on to those who are on the other side of the rapid trading involved.  We saw the kind of liquidity risk that can spike interest rates as a result in July 2003.

What happened?  Consider liquidity risk.  Small earlier yield increases had decreased expectations for mortgage prepayments, sharply increasing the effective duration of many mortgage-backed securities.  Fannie Mae and Freddie Mac were forced to sell long MBS paper to maintain their duration at target levels.  In the same way that dynamic hedging as portfolio insurance can destabilize the stock market, the enormous size of their positions destabilized the bond markets.  Much of the illiquidity-induced downdraft was reversed in ensuing weeks, but the volatility clearly demonstrated the potential fragility caused by high leverage in mortgage-backed securities.

So What Do We Do?

In Part III of this series, I will identify strategies for bond investors that might be especially appropriate in mitigating the interest rate risks identified here.