Lower market expectations for asset classes are forcing pension funds to revise their investment policy. This is boosted by the Financial Assessment Framework (FTK) of the Dutch Central Bank (DNB). Under the FTK regime pension funds are judged on the ability to meet their liabilities, which are difficult to beat in the current economic environment notwithstanding the performance of 2005. In meeting liabilities in the future, pension funds need more than just the market expectations of their current investments.
In order to enhance the performance of pension fund investments and meet, or preferably outperform the liabilities, funds need to implement a more advanced investment strategy. In the past strategic asset allocation (the conventional approach) dominated the amount of manager risk the funds could take. Market picks were driven by market capitalisation (for instance US equities large cap) and didn’t therefore always face substantial out performance possibilities. The strategy we discuss in this article is based on the separation of market risk and manager risk and is known as the ‘Portable alpha and beta’ concept.
In the portable alpha and beta concept, in contrast to the conventional approach, the exposed manager risk is no longer a result of the strategic asset allocation (exposed market risk) decision. In constructing the optimal portfolio within the applied concept both decisions are made independently. This separation leads to a higher allocation to manager risk, where as in the conventional approach typically more than 95% of the risk budget is assigned to the market. Porting alpha and beta leads to a better balance between market risk and manager risk and significantly increases the total expected portfolio return of the pension fund (and with that its information ratio).
Higher portfolio returns do not automatically mean an improved ability to outperform liabilities. To comply with the focus on risks against liabilities, we prefer to use the strategic asset allocation part of our decision to limit those risks through a better match of assets and liabilities. In comparison to the conventional approach, the strategic asset allocation part is smaller, less dominant and not used to add return over liabilities. Therefore, a substantial larger part of the portfolio is dedicated to a ‘return seeking portfolio’. The main activities in the return seeking portfolio are played in the field of exotic beta (return from alternative asset categories) and alpha (return from pure manager skill) competing with each other.
To assess the competition between beta and alpha, and to allocate the desired beta and alpha to the market where each add most value, we should understand the pros and cons of both. We see substantial differences between alpha and beta. First, alpha has a larger impact on the information ratio of the total portfolio, mainly caused by smaller tracking errors. The choice for a successful manager therefore improves the (excess) return of the total portfolio significantly. On the other hand, on an aggregate level alpha is negative, more costly then beta exposure and achieving positive alpha is hard and a time consuming exercise. Therefore, the search for alpha demands high governance budgets. However, a big ‘plus’, low ‘inter-alpha correlations’ make multiple sources of manager risk an extremely interesting part of the portfolio.
To outperform liabilities of a pension fund we need beta as well as alpha. The portable alpha and beta concept is especially designed to create an optimal balance between the two risk components. One needs to find a way to divide and allocate the total risk budget over these two sources. For example, beta risk is preferably taken in those markets with high expected returns. These are not necessarily markets with the highest alpha potential for managers.In this phase the identification of profitable markets for alpha and beta (separated from each other) is crucial.
Pure alpha strategies (that is, uncorrelated to market exposure) are hard to find. Typical alpha strategies like market neutral hedge funds seem to be significantly correlated to beta1 . So, pureness is an issue.
Another issue is the absolute level of alpha potential in inefficient markets. In efficient markets (for instance US government bonds) consistent alphas are rarely found. In the search for alpha emerging markets equities and small cap equity markets seem to qualify better. Other, more unconstrained strategies such as long term, absolute return mandates do too. Unfortunately, these strategies are often capacity constrained and therefore pension fund boards need to act quicker than in the traditional case to build an optimal portfolio.
From a beta perspective, the more inefficient markets bring some other difficulties to the table too. The most important one is the difficulty to remove undesired beta exposure. In efficient markets it is easy to replicate the key benchmark and/or hedge out undesired market exposure via swaps and futures. This is not the case in most inefficient markets. Why?
The costs of swaps, futures and forwards highly depend on their liquidity and they are different in each market. For this reason, it is important to compare costs (bid-offer and transaction) and capital requirements of the used instruments. Unfortunately, in inefficient markets, where the need for removing undesired beta is most required, the implementation of the programme is the most difficult and expensive. This is another key challenge of the portable alpha and beta concept.
If identification and implementation issues are overcome the pension fund should continue to focus on the selection of best-in-class managers in order to ‘reap’ from the alpha rewards. Understandably, without an outperforming manager the whole concept is pretty useless.
Some concluding remarks. Theoretically, portable alpha and beta bring us new possibilities to increase the information ratio of portfolios. Practically speaking, we need to overcome essential implementation issues.
A better and more efficient balance between market exposure and active management could be the solution for pension funds to achieve their long-term expected return over liabilities. Implementing a liability matching approach in combination with a return seeking approach, thereby using portable alpha and beta techniques will probably help pension funds in meeting their goals.
1. Correlation CSFB Market Neutral Index versus S&P 500: 0.4 (1994 – 2003)
Correlation CSFB Market Neutral Index versus MSCI EAFE: 0.3 (1994 – 2003)
Sander Gerritsen and Gerard Roelofs work at the Investment Consulting Practice of Watson Wyatt the Netherlands.