“Neither a borrower nor a lender be” wrote the great bard.
Nobody listened to Polonius because he didn’t speak Japanese.
September 4, 2006.
Commentators offered many reasons for the recent bond market rally. The housing meltdown, Fed policy now credible, inflation risk under control, oil prices falling, hurricane risk diminished, Middle East cease fire---these are some of the causes we heard about. They may be right, in part; however, we think there is an additional and powerful explanation. Part of the recent decline in high-grade US bond yields was caused by Japan.
Some dates are key. Japan warned the markets that it would start raising interest rates on March 9, 2006. The first actual transactions to lift the Zero Interest Rate Policy (ZIRP) caused the LIBOR yen rate to rise on May 8th. By mid-June Japanese monetary authorities realized they may be impacting world markets so severely that the risk could rebound to their country. The world’s stock markets had lost nearly $5 trillion in market value in the six weeks from early May to mid June. Simultaneously, nearly all global bond yields rose.
Japan backed off by late June. Since then, Japan has signaled that they will move more slowly. In August, they suggested that rate increases will be coming even more slowly because of data revisions. Japan tightened in May, halted the tightening process in June and further slowed its pace in August. Japan has raised rates only once this year from zero to ¼ of 1%. As this is written, markets expect, at most, only another ¼ of 1% before yearend. For a full description of events and all key dates see the English translation of an August 2, 2006 speech by Policy Board member Atsushi Mizuno; it is now posted on the Bank of Japan's website: http://www.boj.or.jp/en.
Japan changed its monetary policy course three times this year. While it focused on its domestic agenda, it impacted the entire financial world each time. A soggy Labor Day weekend allowed us to elaborate on this theme of Japan and global bond prices (yields).
The rest of this longer-than-usual commentary is divided into four parts. It takes about 8 more minutes to read. We invite email in agreement or disagreement.
Part 1: Setting the stage for the bond market’s 2006 theatrical performance.
The entire global bond market measures nearly $60 trillion. The issuing countries and their respective corporations divide into 41% in US dollar, 26% in euro; 15% Japanese yen, 5% British pound. About 13% totals all the rest of the world.
This global market breaks into two categories. Internationally traded bonds are about $15 trillion or 25%. The 75% or $45 trillion of domestic bonds either trade or are held within each country by their respective domestic institutions or individuals. The $15 trillion of international bonds are denominated in the following currencies: 46% euro, 38% US dollar, 8% British pound, 3% Japanese yen and 5% for all the other currencies of the world. Note that smaller countries often issue debt in the currency of one of the four big guys.
Essentially, these four currencies drive the world’s bond markets. Three of them have been raising rates or are still raising rates. Those three central banks are the European Central Bank (ECB), The US Federal Reserve (FED) and the Bank of England (BOE). The Bank of Japan (BOJ) has held its policy-setting interest rate at zero for years (ZIRP) and has just started to wean that country from this interest free borrowing source. In so doing it is also weaning the rest of the world from the “carry trade.”
Carry trade means borrowing yen at zero interest; then converting the yen into another currency and lending it in euro or dollar or pound or something else. The loan is made at a higher interest rate. The hedge fund or investor pockets the difference known as the “spread.” Trillions of global stock and bond positions directly or indirectly set their prices from this carry trade spread.
We can measure the result of the carry trade in a number of ways. One is to watch risk premia; they have been shrinking for years. This narrowing of credit spreads and other risk premium indicators accelerated as Japan introduced and maintained the ZIRP. This narrowing is something we would expect when the cost of money is zero.
Another item to watch is the pricing of various assets. Stocks, houses, gold, commodities, collectibles----virtually all asset classes have had strategic bull markets. We would expect this outcome as well since the root of the financing cost has been a zero interest rate policy (ZIRP).
The asset class that has had a bear market over the last three years is high-grade bonds. Their yields have been rising (prices falling) when measured strategically. In the US the low yield on the 10-year Treasury note was about 3% in mid-June of 2003. Japan’s lowest yield on its benchmark 10-year note was well below 1%. British yields have been under 4% and the euro-denominated benchmark German bond traded around 3%. The carry trade rooted in Japan’s ZIRP was partially responsible for these extremely low yields. Other factors like fear of deflation certainly contributed to these generational low interes- rate levels just as it stimulated Japan’s implementation of ZIRP.
Part 2: A time line of what happened this year as Japan ended ZIRP and when Japan changed policy this summer.
If our thesis is correct, we would expect to see change reflected in the Japanese bond market earlier than in other global bond markets. That is exactly what happened. Japanese yields had been climbing for months as the end of ZIRP became an embedded expectation. Japanese 10-year benchmark bond yields peaked on May 15, 2006 at 2.025%. This was one week after ZIRP was lifted. It coincided with the severe May-June sell off in stocks and bonds throughout the world. Japanese yields fell from that point. They bounced back briefly in July and then fell again. That yield was 1.61% this past Friday morning. In sum, Japan’s benchmark bond yield moved 41 basis points (a basis point is one one-hundredth of 1%) from its peak to Friday morning’s price. This is a decline in yield of 20% from peak to trough.
The US Treasury note market followed a similar pattern. Yields rose to about 5% in mid-May and climbed to a peak of 5.25% on June 26, 2006. They have been falling since and followed the Japanese yield decline in almost lock step fashion. The Friday morning yield was 4.75%. The US yield decline was 50 basis points in nine weeks or a decline of 9% from peak to trough. Many believe the US would be the most effected by Japanese policy change because the Japan-US spread is the widest among the big four currencies. Therefore US yields should be the most sensitive to changes in Japanese yields. This seems to be supported by the market action we saw this summer.
Similar patterns also appeared in the market for euro denominated debt. The German benchmark bond peaked at a yield of 4.1% on July 5, 2006. Its market movement is almost perfectly correlated with the US Treasury note. We measured this with a moving average on a daily basis and the R squared is in the high 90s. From peak to Friday morning’s 3.8% yield we see a drop of 33 basis points or a decline of nearly 9%.
Even the Bank of England’s independent policy moves only partially impacted their bond market. That market was also influenced by the Japanese. The British 10-year benchmark followed a pattern similar to the others in May and in June. BOE policy factors led it back to a slightly higher yield of 4.8% on August 14, 2006. On Friday morning it was yielding 4.6% or a decline of 20 basis points or almost 5%.
Examination of the smaller currencies like Australia or New Zealand will show similar patterns. In fact, it is hard to find any tradable market in the world that did not follow the movements we have described above. That is particularly true of the May-June period.
During that period, risk premia widened when measured by credit quality. This was also a worldwide phenomenon. The spreads between higher-grade countries and emerging market debt widened as did the spread between US junk bonds and Treasuries. Also, the US based Dow Jones CDX Index of credit-default swaps widened from 34 to 46 basis points in this May-June period. The CDX measures the spread to short term rates for 125 companies in the S&P 500. It is a proxy for investment grade credit pricing.
Part 3: What we believe lies ahead.
We believe that the global influences on bond markets are more interconnected than ever before. We also believe that the world’s central banks remain domestically centric in their policymaking. This is completely understandable. Central banks are creatures of political decision-making. The FED was created by Congress. The ECB derives its power from a treaty; eleven countries ceded sovereignty in monetary affairs to a central authority at its inception.
We are not quarreling with the political construction of the world’s central banks. We accept their creation and purpose. However, the interdependency and interconnectedness of the financial world has galloped forward while the domestic currency-centric monetary authorities have lagged in their processes. This notion is the secret to understanding global bond market volatility.
Let’s explain. At full complement our US Federal Open Market Committee (FOMC) consists of 19 Americans. Their data compilation is highly US centric. The 12 regional Federal Reserve Banks devote the giant portion of their efforts to their regional and local economies (beige book) and to the influence on policy within the US. Sure there are global views articulated by policy makers. Certainly there is some research in this area. But the overwhelming mass of information gathering and policy focus of our Federal Reserve is internally targeted on US economic factors and the outlook for the US economy---it is dollar centric.
The same situation is found at the ECB, the BOJ and the BOE. Each bank is making policy as we would expect. Each is nearly totally focused on its respective domestic economy, inflation rate, employment rate, etc. When Japan decides on a ZIRP, it places little value on how that will impact the rest of the world’s bond markets. It doesn’t worry about US speculators or Cayman Island based hedge funds. With ZIRP, the Japanese monetary authority was looking within.
This is true when the BOE raises rates to fight a domestic inflation. Or when the ECB does the same. Or when other countries in Europe maintain their currency boards or pegs to the euro. Or when the Chinese manage the currency ratio between the Yuan and the US dollar. We cannot find a single monetary authority in the world that elevates the global impact of their decisions to a level which influences the outcome. Central banks are practicing nationalism while the world is globalizing.
If central banks remain domestically centric while the world’s financial interconnectedness expands, the result can only be rising world volatility in financial markets. We see this when any one of those central banks makes a major change. That is what we saw in May-June when the BOJ moved away from ZIRP. That is also what we saw in August when the BOJ implementation of the policy shift slowed as the data revisions were revealed.
Such is the power of a single currency and large central bank when the world is interconnected and where the four major currencies trade relatively freely. The globe’s financial markets quickly re-price currency ratios and can rapidly adjust asset prices. To follow those markets we must track the currency exchange rates just as closely as we follow each country’s monetary authority.
Most bonds are denominated in single currencies. They are claims on that currency only. In addition, they are subject to credit risk, inflation risk, liquidity risk, taxation peculiarities and other factors which determine their relative prices (yields). Yet bonds are reactive to events beyond the borders of the currency in which they are denominated. That is one of the reasons the US bond market has rallied (prices rose, yields declined) in August along with most bonds in the rest of the world.
Part 4: How we incorporate Japan’s policy when we are managing bond accounts for our clients.
Our outlook assumes that the present Japanese policy of a very slow extraction from ZIRP will continue for a while. The Japanese still fear deflation. That means the carry trade will also continue. More and more leverage will be applied on a global basis.
Increasing leverage means that there is higher risk. It also means that a higher profit occurs before the risk is realized. During that time the applied leverage is operating positively. This is where we find ourselves today. Most of the profit is behind us; most of the risk lies ahead. We got a taste of that risk in the May-June worldwide bond and stock market sell off. Yet global markets appear to have priced in the most favorable outcomes. Should developments prove to be more adverse, there could be some sharp reverberations.
Recall again the world’s bonds are worth about $60 trillion. World stock markets are worth over $40 trillion. The four key currencies define most of this asset value. The monetary growth rates and application of leverage are in the hands of these four banks’ decision making officers. While they are talking to each other more than ever before, their international discourse is about preventing systemic failure and not aimed to prioritize financial asset prices. They do not (and should not) guarantee a put to the financial markets. But they are aware that the combined $100 trillion size of these asset markets is more than 2 times the world’s GDP and is a record ratio to that GDP.
In addition, it is clear to us that the politics of the world’s countries do not permit deflation. Central banks will do whatever it takes to avoid it. Because of the Japanese experience, the world has reset the zero inflation line to about 2% as a target. 2% inflation is now acceptable instead of zero change in the level of prices. We must think of 2% as our inflation baseline.
There is evidence that inflation is increasingly globalized. It also appears that those countries and central banks which have stated inflation targets are able to achieve a lower inflation expectation than those which do not state an explicit target. The US is currently in the latter category. The Greenspan FED did not allow a stated inflation target. The Bernanke FED is involved in internal debate on this issue.
In June of 2003 we saw the generational lows in US interest rates. To get there again or go lower we will need to experience a severe recession or depression. The FED will not permit deflation and will print what ever amount of money is needed to avoid it. The same is true for the other central banks around the globe.
The present level of interest rates in the US is a product of many factors. One of them is the slowing of the restoration to normalcy in Japan. Any evidence that Japan will re-accelerate its movement away from ZIRP will cause global interest rates to rise and that will include in the US. This will be true even if the US economy will be slowing and even if the inflation rate in the US is well behaved.
At Cumberland we have been moving our bond duration to neutral slowly. We will not chase this bond market rally from present levels. We think the risks and volatility characteristics of the US bond market are high in a strategic sense.
Caution is warranted because some of the pricing of US bonds is dependent on the policy agenda of Japan’s central bank. For the moment, Japan is weaning itself from ZIRP at a very slow pace because it expects “a sustained period of expansion under price stability” according to Mr. Atsushi Mizumo.
This sounds like perfection to us. We do not want to base our clients' portfolio strategy on a perfect outcome. In the history of world’s monetary affairs, we have never seen perfection preformed by any central bank. We do not believe that this time is different.
We continue to emphasize the highest-quality credit. In many markets we believe the investor is not being adequately compensated for taking additional credit risk by lowering credit quality standards. Measures of credit risk premia suggest that taking on credit risk to obtain higher yield is a bad deal for a US investor. While Cumberland benchmarks against traditional bond indices, we note that our accounts are presently of higher-credit quality than the credit distribution within those indices.
We need to acknowledge the Bank for International Settlements (BIS), Barron’s and Bloomberg for data. Special thanks to Bianco Research for conceptual work on Japan and ZIRP. My colleagues Bill Witherell, John Mousseau, Peter Demirali and Matt Forester each contributed input to this commentary. I thank them for their inputs. All errors are mine.
Please note that the bond aggregates used above do not include the debt of financial intermediaries and do not calculate the additional influence of derivatives. Intermediary debt aggregates are not easily available for the entire world. In the US financial intermediary debt now exceeds 100% of our GDP for the first time in our nation’s history. We estimate that inclusion of global intermediary debt would raise the debt-GDP ratio above 3. We also believe that amount of leverage applied because of derivatives is extraordinary and now nearly beyond estimation.
Please note that all calculations are to the September 1, Friday morning prices before the 8:30 a.m. release of the US employment report. We purposefully used the pre-report prices so that the bond market reaction to it would not be included in these discussions.
We also note that Japan’s stock market was up 1.4% in Monday’s trading and the yen continued weak against the euro. Disclosure: Cumberland’s international ETF accounts are over-weighted Asia and over-weighted Japan within Asia. Our Global Multi Asset Class ETF style holds a long position in the euro ETF. In addition, a cash reserve is in place in all ETF styles whether domestic or international.
|