Spend, Save and Invest Smartly
There are a many indices that follow the hedge fund industry, and these fall under into three main categories. In their historical order of development they are Non-investable, Investable and Clone.
In established equity investment, indices play a central and clear-cut role. They are extensively accepted as representative, and products such as futures and ETFs provide investable access to them in most developed markets. Nevertheless hedge funds are illiquid, heterogeneous and ephemeral, which makes it hard to construct a reasonable index. Non-investable indices are delegate, but due to various biases their quoted returns may not be available in practice. Investable indices achieve liquidity at the expenditure of limited representativeness. Clone indices inquire about to replicate some statistical properties of hedge funds but are not directly based on them. None of these approaches is entirely satisfactory.
Non-investable indices are analytical in nature, and aspire to represent the performance of some database of hedge funds using some measure such as mean, median or weighted mean from a hedge fund database. The databases have varied selection criteria and methods of construction, and no single database captures all funds. This leads to noteworthy differences in reported performance between different indices.
Even though they aim to be representative, non-investable indices suffer from a lengthy and largely unavoidable list of biases.Funds’ participation in a database is voluntary, leading to self-selection bias since those funds that choose to report may not be typical of funds as a whole. For instance, some do not report for the reason that of poor results or because they have already reached their target size and do not wish to raise further money.
The short lifetimes of many hedge funds means that there are many fresh entrants and many departures each year, which increases the problem of survivorship bias. If we look at only funds that have survived to the present, we will misjudge past returns because many of the worst-performing funds have not survived, and the observed association between fund youth and fund performance suggests that this bias may be large.
When a fund is additional to a database for the first time, all or part of its past data is recorded ex-post in the database. It is to be expected that funds only publish their results when they are favorable, so that the average performances displayed by the funds during their incubation period are inflated. This is called as "instant history bias” or “backfill bias”.
Investable indices are an effort to reduce these problems by ensuring that the return of the index is available to shareholders. To generate an investable index, the index provider selects funds and develops structured products or derivative instruments that distribute the performance of the index. When investors buy these kinds of products the index provider makes the investments in the underlying funds. This makes an investable index related in some ways to a fund of hedge funds portfolio.
Only Hedge Funds that be in agreement to accept investments on terms acceptable to the constructor of the index are included in the index, so that the provider can sell products based on it. This guarantees that the indices are investable, which is an attractive property for an index because it makes the index more appropriate to the options available to investors in practice. Though, investable indices do not symbolize the total universe of hedge funds. Most seriously they may under-represent the more successful managers since these may find the index terms unattractive, for instance due to reduced fees or tedious redemption terms being demanded by the provider.
The most recent accumulation to the field approaches the problem in a different manner. As a replacement for of reflecting the performance of actual hedge funds they take a statistical approach to the analysis of historic hedge fund returns, and use this to construct a model of how hedge fund returns respond to the movements of various investable financial assets. This model is then used to build an investable portfolio of those assets. This makes the index investable, and in belief they can be as representative as the hedge fund database from which they were constructed.
Regrettably they rely on a statistical modeling process. As clone indices have is relatively short history it is not so far possible to know how reliable this process will be in practice.