In July this year, the debt capital markets marked the thirtieth anniversary of the first ever index for the fixed-income universe.
The founders of that first index, Lehman Brothers’ Art Lipson and John Roundtree, argued for a market-weight construction since the rules are easy to communicate, easy to understand and easy to conform to: the more debt you issue, the greater your position in the index. Consultations with the Bond Portfolio Managers’ Association, an industry body of fixed-income institutional investors, found that the argument had the support of market participants. It was agreed that to ensure the fairest and most realistic depiction of the market, a good index should have the fewest assumptions.
And so the Lehman Government and the Lehman Corporate Bond indices – the forerunners to Lehman’s Aggregate Index – were born. But the wisdom of deciding how much a borrower constitutes in the index according to the amount of debt it has outstanding has been debated ever since.
The concept of market weighting was one transposed almost exactly from the equity markets. However, in equities, the largest companies in the index are taken to be the most successful and the most actively traded. Neither of these assumptions is necessarily true in fixed income. The largest companies in the credit index are the most indebted and are not automatically the most liquid.
As a result, many investors complain that the application of market-weighting to fixed income creates an awkward problem. Fixed-income investors whose performance is measured against a market-weighted index may be forced to hold a large position in a credit that they do not like, or risk taking a large bet against the index – one that may pay off, but may not.
In an index that represents a large market, such as the Lehman U.S. Aggregate, which includes roughly 875 different issuers with 7,452 bonds outstanding, a market weighting does not present too serious a problem. Investors can choose not to hold credits they don’t like because the breadth of the index means that no one corporate issuer represents more than about 1.93%.
However, in smaller markets, market weighting can be a problem. The Italian carmaker Fiat, which was recently downgraded to junk, constitutes roughly 12% of the European high-yield market. And six months before it defaulted in December 2001, Argentina made up 23% of JPMorgan’s emerging market bond index.
As John Kirby, head of the fixed-income index team at Boston-based State Street Global Investors, says: “The problem is that if you don’t partake in these deals, you are taking an active credit decision in your portfolio and if you do, you are taking quite a risk with your clients’ money.”
Nigel Sillis, head of credit research at Baring Asset Management in London, points out that many benchmarked fund managers may not even have that freedom of choice. “Mandated risk controls can prevent managers from deviating too far from the benchmark,” he says. “So by avoiding a credit that is 5% of the index, you may have to surrender all your tracking error in one go.”
Following a period in which the global credit markets have seen record numbers of heavily indebted companies falling out of investment grade and into the high-yield index, the problem has become all too common. And the issue has prompted several banks to create constrained indices, in which there is an upper limit on how much an issuer can represent in an index, often around 2%.
But the concept of a constrained index merely raises another question: how constrained should a constrained index be? As Jack Malvey, chief global fixed-income strategist responsible for the global index group at Lehman Brothers in New York, says: “Any constraint on the weighting is inevitably arbitrary. The index should be an objective measure of market reality, not a subjective view of the ideal market.”
The extreme of a constrained index is an equally weighted index, where every issuer represents, say, 1% of the index; but this is not an accurate representation of the market either. The problem is that both equally weighted and constrained indices mean that if everybody uses them and all investors want to be neutral to the index, they would have to hold more bonds in some companies than actually exist. What’s more, the notion of constrained indices has done nothing to tackle the problem that investors may not be able to buy or sell the bonds in the index.
Being able to buy the bonds in the index is naturally a key concern for investors. To represent what an investor can buy, rather than the whole market, constrained indices have responded in the same way as market-weighted indices: by setting a minimum size for bonds in the index. A larger bond tends to trade more liquidly because, well, there is more of it to trade.
But in reality, this is only a partial solution. As Bob Fuhrman, head of fixed-income indices at MSCI, points out: “There are huge numbers of indexed bonds locked in the vaults of long-term investors. These bonds will never see the market again, but they are still being counted in the indices.”
A different approach
There have been some recent experiments in approaching the problem of liquidity in indices from another angle. Rather than replicate the market as told by outstanding bonds, these products replicate the market by what is being traded.
JPMorgan and Morgan Stanley’s recently launched Trac-x product is an example of this. Trac-x is a basket-linked note that provides investors with exposure to 100 credits via a credit default swap trade. The 100 credits are the names that the credit default swap trading desks at JPMorgan, Morgan Stanley and now BNP Paribas are trading most frequently.
Although this bears little resemblance to indices as we have traditionally known them, the banks have been speaking to both investors and investment consultants to argue that Trac-x could act as a benchmark index as justifiably as one that replicates the outstanding debt in the market.
But liquidity-based indices are not without issues. As Elizabeth Para, fixed-income strategist at Barclays Global Investors in London, says: “While liquid indices are interesting, they have to be rules-based and transparent. With no exchange in the corporate bond market, measures of liquidity based on trading volumes are difficult to verify.”
According to Para, the size of an issue and how recent it is are among the best indicators of liquidity, “but sometimes liquidity is something that is more easily picked up by the human eye.” But while an exodus to a liquidity-based index is clearly not imminent, the idea does illustrate something: the index is simply a snapshot of the market.
In many senses, indices are like maps: they represent the market, and they can choose to represent different parts of the market or to represent it in different ways. A map of America, for example, may illustrate the road network of America, the topography of America or even the socioeconomic breakdown of America.
And just as one map may be useful for one purpose and not another, different indices are useful for different purposes. If you want to get from Los Angeles to New York, a map that displays rainfall trends or population growth would be of little use. A map that shows you everything will be useful whatever your purpose, but it could also be confusing. Just which map you use depends on your objective; in the same way which index you use depends on what you are trying to achieve.
Pension funds, for example, have approximately 30-year liabilities, and will seek to earn as much as possible over that time period. Insurance companies have liabilities of roughly three to five years, and so look for liquid assets that are not excessively volatile over that period. Mutual funds may seek to make as much money as possible within a five-year time frame, while hedge funds just want to make money.
In a market in which the long-term goals and strategies of the investment community are segmented in so many different ways, one index alone is unlikely to match the liabilities of all investors. And the ever-changing characteristics of the index through time are unlikely to meet any single investor’s needs all of the time.
As MSCI’s Fuhrman says: “As the market evolves, there may be more or less credit risk, more or less governments or more or less duration. What is it about that changing universe that should entice you, the investor, to follow it around?”
This is precisely why Tom Klaffky, head of the yield book group at Citigroup in New York, argues that there is an important distinction to be made between the index for the market – the map that shows everything – and the best benchmark for any given fund – the map that shows what you need. “The key to successful credit investment is careful analysis of long-term goals and the choice of a suitable benchmark for one’s risk parameters,” he says.
For a security selector like Bill Lissenden, an investment-grade bond manager at Deutsche Asset Management in New York, the best benchmark is a broad index like the Lehman Aggregate because it describes the broadest possible opportunity set. But for those investors with very long-term liabilities such as pension funds, the Lehman Aggregate may be less suitable.
Since each investor’s liabilities are unique, perhaps the best way to measure fund performance would be a myriad of individual indices each suiting a particular fund’s objectives. Most index providers already offer their clients these customized benchmarks, but they inevitably exclude any straightforward comparative measure between managers. And if the only thing that really interests clients is whether the investor is earning more money than other asset managers, then should indices be used at all? Shouldn’t investors be measured against how well they are performing in comparison with other fund managers with similar objectives?
Second- and third-guessing
But this could very well influence the investment process as much as comparison with an index. The eminent British economist, John Maynard Keynes, explained this phenomenon by comparing it to a competition in which entrants have to select the six prettiest faces from a hundred photographs, with the prize going to the competitor whose choice most closely corresponds to the average preferences of the competitors as a whole.
The result is that each competitor does not look for those faces that he finds prettiest, nor does he look for those that he thinks the average opinion might find prettiest. The competitor actually reaches the third degree whereby he tries to anticipate what the average opinion expects the average opinion to be.
But this analogy is not unique to fixed-income investing or peer review, it is the natural outcome of any investing. And since all investors, clients and investment consultants require some metric against which to compare fund managers, it seems likely that if the banks did not provide an index, the investment community would still look at the average of how the market is performing and measure themselves against it – which is in itself an index.
Citigroup’s Klaffky argues that the crucial point is that “the index cannot drive investment decisions, it is the people choosing the benchmarks and guidelines that drive which decisions the manager can make.” In Keynes’ analogy: the rules of the beauty contest determine how the choices are made, rather than the 100 faces in the photographs.
So the problem is not really the index. The problem is not whether one is able to track the index. The problem is not even that, say, WorldCom makes up 20% of the index. The problem is the investment mandate that states that the fund manager must not deviate too far from the index. It is the rules and restrictions of the game that are at the heart of investors’ complaints.
Clients paying for active management naturally want their managers to make their own decisions, to seek out intrinsic value where they can find it and to outperform the rest of the market. But, all things being equal, since the index is the sum of all individual investors, for every investor that is outperforming the market, there must be one or more underperforming the market.
When a client chooses active management, therefore, they tackle the risk of underperformance with mandated restrictions on the investments that can be made. Those restrictions may prevent the manager taking off-index bets, or may demand that the fund manager track the index within, for example, 2%.
And herein lies the problem. Clients restricting their fund managers to only 2% deviation from the index limit their underperformance by 2%, but limit their outperformance to the same degree. As a result, investment mandates that tie fund managers to the performance of an index are consigning the fund to follow the credit cycle, be it positive or negative.
As Nigel Sillis at Baring Asset Management argues: “Mandates do not take into account possible developments in the market, where telecoms may suddenly make up 20% of the high-yield index, for example. As a fund manager, even if you are smart enough to figure out how bad that situation could get, you still have to get involved.”
Investment mandates are therefore forcing many money managers to make sub-optimal investment decisions. And while the mandate may stipulate a 2% tracking error, if the index is down 6%, clients are very rarely pleased to lose 4% of their savings. It seems that many investors may be forking out for active management but getting passive performance.
Some investment consultants are beginning to recognize this problem and are advocating unconstrained investment strategies for fixed income, but there is still a long way to go. Clients remain wary of changing the benchmarks and guidelines with which they are familiar, and of losing their ability to control the investments.
But if investors want their assets managed well, their fund managers not only need an index that better matches their liabilities in order to eliminate some of the instrumental problems associated with indices, but they also need mandates that reflect the true nature of active fixed-income investment: careful analysis and deviation from the benchmark when necessary or beneficial.
Until then, worldly wisdom will continue to follow Keynes’ conclusion to his beauty contest: “that it is better for reputation to fail conventionally than to succeed unconventionally.”