Bonds & Zombies
Anatole Kaletsky

 

Two weeks ago we noted that the biggest factor which was holding down the yields in US and European bond markets was the price-insensitive buying by three investors groups: Asian central banks, Western pensions and insurance funds and, most importantly, Japanese private investors. This paper will explore in greater detail the bizarre behaviour of these seemingly brain-dead "zombie investors", who gobble up whatever paper the US government may throw towards them, regardless of value or price.

In the rest of this paper we present some stylised facts, which may not be literally true of any particular investor, but can help in the aggregate to explain the indiscriminate greed of the zombie buyers for overvalued bonds. We then draw some conclusions about the impact of price insensitive bond-buying by the most important group of zombie investors, the Japanese private savers driven to desperation by zero rates.

Starting with the Asian central banks, the reasons for zombie buying are obvious. They are buying foreign currencies as an exercise in macroeconomic management (some would call it manipulation) on behalf of their government. As long as maintaining exchange-rate pegs remains the over-riding national goal, Asian governments will continue buying dollars (and euros) and investing their reserves in the corresponding bond markets, regardless of the threat of suffering a capital loss.

Western pension and insurance funds have become insensitive to price and valuation for different reasons. Many of the long-term investment institutions, which used to dominate the US and European markets, are actually not in the investment business any more. Under pressure from regulators and accountants, the pension and insurance funds are effectively pulling out of asset management, defining their core business as liability management instead.

Pension funds have moved from the asset to the liability business because most corporate treasurers no longer see managing the pension fund as a function designed to generate profits or even to minimise pensions costs. Instead, they see running the pension fund primarily as an accounting exercise, whose main purpose is to "immunise" the capitalised pension liabilities shown on the company balance sheet. Because pension liabilities are capitalised by using longterm bond yields, the company can minimise its apparent risk by investing the pension fund entirely in bonds, regardless of yield. If bond yields go up, the pension fund will of course suffer a loss; but this loss will be exactly matched by an improvement on the liability side of the balance sheet. In reality, the idea that bonds are a perfect match for pension liabilities is an illusion, since pension payments will increase much faster than implied by the bond yield if inflation turns out to be higher than the market expects. But accountants have decreed that whatever forecast of inflation appears to be embedded in today's bond yields is, by definition, the best available. And who is a mere economist to disagree?

Life companies are pulling out of asset management for a different but related reason. From the 1950s until the late 1990s, insurance companies competed (especially in Britain) mainly on the basis of their investment performance, selling "with-profit" life policies and annuities, which were designed to allow investors to benefit from equity investment returns. In the past decade, however, regulators have limited the life insurers' investment freedom and severely restricted their use of past investment performance in marketing. As a result, the life insurance industry is moving back to its 18th century roots: most life companies now see their main business as providing mortality protection on the basis of statistical demographics, rather than maximising investment returns. Under this business model, the life company's core competence is not asset management but marketing and distribution. If this is so, then buying equities or even trying to time bond purchases by using either fundamental analysis or technical models is a risky diversion from the insurance company's main business. And since all insurance companies are under the same regulatory pressure not to market their policies on the basis of differential investment performance, there is no competitive benefit in taking investment risks.

In this new business model, the fund manager's job is not to maximise investment returns, but simply to direct premium payments into the bond market as quickly as the salesmen can collect them - and, of course, to achieve the closest possible match between the expected duration of assets and policy risks. Any attempt to market-time bond purchases, or even to improve returns by varying portfolio duration, is seen by management, shareholders, auditors and regulators as reckless speculation, needlessly exposing the company's balance sheet to risk.

Of course, we have grossly oversimplified our descriptions of insurance and pensions fund behaviour. Many of these businesses still believe in managing their assets in an aggressive and imaginative way. At the margin, however, it is clear that the long-term saving institutions are moving in the direction suggested. This can be seen in the Fed's flow of funds statistics, which show that in 2003, bonds accounted for 76% of the insurance industry's net acquisition of financial assets, compared with 41% in 1998. While pension funds were not big bond investors until 2003, according to the Fed statistics, by the first half of 2004, they were investing more than their entire cash flow in bonds. With both insurance companies and pensions funds increasingly focused on liability matching, rather than asset management, it is hardly surprising that more and more insurance companies, like pension funds, are willing to buy bonds robotically, regardless of price.

Now let us turn to the most interesting and important cohort of zombie investors: the private savers and retail investment institutions of Japan. We suggested last month that probably the best explanation for today's "conundrum" of influences global bond yields was the asset-liability matching of Japanese private savers. The argument could be summarized like this:

The foreign private sector is a much bigger buyer of US bonds than either the central banks or the domestic investment institutions. Foreign private investors bought $655 bn of US-issued bonds in 2004, almost three times the official bond purchases of $235 bn (see charts below). By far the most important group of foreign private investors have been the Japanese, either buying directly through foreign bond funds or indirectly through Japanese life insurers and trust banks.

Why have the Japanese been buying so many US bonds? Let us try to go beyond the two easy explanations: that Japanese banks and insurers act on manipulative "guidance" from the Ministry of Finance or that they are simply mad. The alternative explanation is that yields of 4% plus on foreign bonds are irresistibly attractive to savers who live in a financial system permanently distorted by zero rates. comparison with Japan's Of course the yield differential between a JGB at 1.5% and a 4.5% Treasury can be wiped out by a currency loss on the principal in a single morning. But what if the Japanese saver doesn't care about the principal value of his investment because he is only interested in maximizing his income return? Begging the indulgence of regular readers, let us repeat the theoretical example we offered in our Five Corners publication last month:

Imagine an elderly Japanese retiree, with cash savings of Y100m and suppose for simplicity that the exchange-rate is $=Y100. Suppose, further, that our Japanese saver wants only to secure the best possible pension for his remaining 20 years of life, leaving nothing as a legacy to his children (maybe because he did not have any as is increasingly the case). He can do this by buying an annuity, backed by a bond investment either in yen or in dollars. Looking up the annuity tables we find that a yen annuity, based on a 20-year yen bond yield of 2 per cent, would provide an annual income of 6.1% or Y6.1m. A dollar annuity, based on a 20- year dollar yield of 4.8 per cent, would generate 7.9% or $79,000 currently equivalent to Y7.9m.

Thus the Japanese pensioner can increase his initial income by 30% if he buys an annuity in dollars instead of yen. Of course he faces a currency risk, but given the 30% uplift in his annuity payments, he will remain better off as long as the dollar exchange-rate over the next 20 years averages above Y77 (= 61 divided by 79). Assuming a straight-line depreciation of the dollar, that would imply a break-even exchange rate in 2025 of $=Y54. Taking account the time-value of money and the inherent uncertainty about the saver's lifespan, the true breakeven exchange rate is even lower, probably well below Y50.

Will the dollar fall below Y50 by 2025? It could happen. The dollar-yen exchange rate did almost exactly halve between 1985 and 2005. But given the demographic and structural characteristics of the US and Japanese economies, betting against another halving of the dollar does not seem completely stupid. So maybe we should not be so contemptuous of Japanese investors who buy overvalued bonds denominated in dollar, provided they intend to hold them to maturity and to leave them unhedged.

In reality, of course, this is not be what the "zombie" investors are doing when they fight for every new Treasury issue - and for every offering of European governments, or corporate and mortgage bonds. In reality, some of the Japanese insurance companies hedge some of these purchases, in which case they are simply playing the relative steepness of the Japanese and foreign yield curves. Others try to manage their currency overlays to generate extra profits (or suffer even bigger losses). Still others use the arbitrage between US and Japanese bond yields, to sell retail savers structured products which offer better terms than could be funded directly in the JGB markets, but rather worse than could be obtained by bearing the whole of the dollar-yen risk. Exploiting the annuity arbitrage which we have described above by creating such structured products, is a very profitable business for retail savings institutions and investment banks in Japan. They do this in different ways - some of the life companies hedge their entire foreign bond holdings, other have reduced their hedge ratios to as low as 15%.

Supposing we are even roughly right in identifying these groups of priceinsensitive investors, what are the implications of all this zombie buying of bonds?

The zombies' enthusiasm obviously does not mean that bonds are good value, still less that the bear market which began in June 2003 is about to go into reverse. We believe the upward trend in long bond yield will continue at least until they reach 5.5%. The zombie buyers cannot change the basic cycles of macroeconomics any more than King Canute could command the tides. What they can do is slow the trend, extend it over a longer period, and inflict losses on rational investors who quite sensibly want to trade against the zombies, but must mark to market every day. As Keynes said, the market can remain irrational much longer than the rational investor can remain solvent.

Our first conclusion, therefore, is that the crash in bond markets which we expected to see on the basis of the past two economic cycles - 1994 and 1983/4 - will be slower and less disruptive than we thought. From a long-term perspective, however, a slow rise in bond yields is probably more ominous than a sharp bounce. Technically, bond yields are creating a huge multi-year base. In terms of economic fundamentals, the bond market's refusal to push long-term interest rates higher will offset the restrictive effects of monetary tightening. As a result, long-term inflationary pressures will intensify, fiscal disciplines on governments will be loosened and all asset prices will be pumped up.

Secondly, the present asset inflation - and rolling financial bubbles - will last for a surprisingly long time. If we are right about the dominant role of Japanese private savers among the zombie bond investors, then global liquidity and asset prices will depend less on US monetary policy or the strength of the dollar, than the level of interest rates in Japan. Given the likelihood that Japanese short rates will remain at or near zero for the next two years - and probably until the end of the decade (see Anatole's Notes from his Japan Trip) - this means US and euro rates long rates will remain "unnaturally" low for a very long time, regardless of what the Fed may do (or say). If this turns out to be true, then the implications for global economic conditions and for equity, property and other asset prices, should be extremely bullish.

Essentially what we are saying is that all global asset prices markets will remain severely distorted as long as the main suppliers of excess savings to the world economy - the Japanese private investors - continue to live in a zero-rate environment. If the returns available to Japanese investors domestically remain at or near zero, they will continue to invest in dollar and even euro assets at what seem to be ridiculously low rates.

The final question is what could change this situation. The obvious answer is new investment opportunities for the Japanese. Rather than an increase in Japanese interest rates, which remains extremely improbable, the likeliest source of new opportunities would be an improvement in equity (and property?) returns in Japan. If Japanese investors rediscovered their domestic equity market, they might stop their zombie bond-buying and conditions in the world financial markets would be transformed. That this is not just an idle speculation is suggested by the surprising correlations shown in the chart below.

The Nikkei and the US bond yield have been moving in tandem for most of the last decade. The correlation of daily movements in these two markets has been 90% since 1990 and 92% since 1996. Intriguingly, the correlation between the Nikkei and the US bond market has been much closer than the correlations between US and Japanese bond markets or between bonds and equities within either Japan or the US.

To judge by this correlation, global bond investors should forget their monthly vigils ahead of the US payroll figures and focus instead on equity prices in Japan. If the Japanese economy regains momentum and the Nikkei manages to break through the 12,000 level, trend-following Japanese may suddenly decide they have better opportunities for investment than US or euro-denominated bonds. The bear market in global bonds would then resume in a big way.