Since the 1950s, when Markowitz first conceived the idea of investing in whole equity markets, indexation has engendered a vast amount of research and debate. By the early 1970s indexation had moved from theory to practice with the launch of the first index fund. Since then index mandates have grown enormously. Total indexed assets world-wide are now approximately US$2.4trn (g2.8trn). Much of this sum is in US equity with a fast-growing portion in non-US equities. Less than 10% of the indexed assets are in bonds. Given that bonds are still the largest asset class for most continental European investors, the question arises: Why is not more fixed income invested
The data available on active versus passive tends to neglect bond markets. Bonds are often managed internally and passive strategies need economy of scale. With the introduction of the euro, many investors reviewed their investment strategies and changed their domestic bond asset class from local currency to euro bonds. The investment opportunity set to actively add value in this asset class has changed (see chart 1). Today, there are no interest and currency bets possible in Euroland. The opportunity to add value is restricted to duration bets only – or if one invests in credits it has been reduced to bets in duration and corporate bonds.
The difference between the government and the corporate yield curve is known as ‘the credit spread’. By definition, the credit spread prices the various risks associated with corporate bonds relative to government bonds. The credit spread compensates for taking more risk. There are a variety of risks, which an investor should be compensated for, when investing in the credit market, such as liquidity risk, sector risk, event risk, spread volatility and, in particular, default risk.
Table1 shows the historical probability of default between 1970 and 1997.
The long-term investor benefits from investing in the credit market, because he can endure periods when spreads increase and corporate bonds perform poorly. Active bond managers, who take strategic bets in credits, can easily outperform a government bond index in the long run, because the required spread to compensate for default is relatively low in comparison to the credit spread as a whole3. Since short-term investors may suffer underperformance relative to government bond investors due to short-term volatility of spreads, diversification across the different credit segments will be valuable in reducing the risk exposure in credits. This makes a case for indexing bonds, including credits, to gain the full diversification benefits. What are the other arguments for indexing bonds?

Three arguments for indexing bonds
Using index funds to implement a European bond asset class is an appealing strategy. It is based on the following three arguments for passively managed funds in fixed income asset classes:
q The risk argument: Many investors’ use modelling techniques to determine a strategic asset allocation (for example, via asset liability
management). These analyses are based on market returns, that is passive returns. If the investor adds additional (active manager) risk to the market (asset class) risk during implementation, the question is: in which asset class and how much active risk should be taken?
One approach is to use a risk budget and define a maximum active risk figure with a way to allocate the active risk (risk budgeting). For example, one could add active risk, especially in high-risk asset classes, (for example, private equities and emerging markets), since here the low correlated active risk will not have a big impact relative to a low-risk asset class. This is demonstrated in the following simplified example:
Adding active manager risk of 3% tracking error to bonds will increase the total bond asset class risk by 0.83%, but only by 0.15% for the equity asset class. Thus active risk is better used in high risk classes rather than in lower risk classes.
q The cost argument: Clearly, an active manager needs to at least outperform the cost hurdle between the active and passive total expense (including transaction and safekeeping costs). In fixed income markets, particularly Euro-zone government bonds, the scope to add value is small relative to the total costs inherent in an actively managed Euro-zone bond portfolio. The ratio of cost to expected asset class return might be more favourable for equity investment. Thus the active fixed income manager is less likely to outperform a benchmark after costs are accounted for than the active equity manager.
q The return argument: Apart from the above more theoretical and forward-looking arguments, what does historical data suggest (see Table 2 below)? The Feri Trust universe of active bond manager returns suggests that the longer the period the more difficult it is to outperform the index. After 10 years, only one manager could beat the international bond index, no manager could beat the European bond index and only one manager could beat the EMU index of the corresponding benchmark. This comparison does not take into account survivorship bias and the strategic credit positions mentioned above. The BVI universe basically suggests the same.
The theoretical arguments suggest passive management is more suitable in asset classes with low risk and high efficiency, such as in Euro-zone fixed income. Historical data supports these arguments even without taking into account that many active bond managers are invested outside their benchmarks. Properly measured, to include credits in the benchmark, the picture would be even more convincing. The search for higher returns through corporate bonds and the need for diversification and the risk, cost and return arguments make a strong case for passively managed Eurozone bonds.
Olaf John is director of European institutional business at Barclays Global Investors in London