An Alpha+Beta Framework
| Nov 16th, 2006 | Filed under: Portable Alpha & Alpha/Beta Separation | By: Alpha Male |
By: Edward Kung, Babson Capital & Larry Pohlman, Wellington Management Company
Published:á=Winter 2005, Journal of Investing
This article provides a ringing indictment of strategic asset allocation (SAA). The authors argue that an Á-°Alpha BetaÁ-? solution is far superior to a simple allocation to stocks/bonds/cash and sub-allocations to various equity classes.
Á-°A policy portfolio needs to meet all of these objectives [short & long term liabilities, growth, risk budgeting, out performance] within a single, traditional alpha and beta Á-?bundledÁ-? product structure is inefficient if not inappropriate.Á-?
Essentially, Kung and Pohlman say that SAA falls short on six fronts:
-
- Á-°The alpha and beta decisions are tangled when implementing an asset allocation.
- Á-°Alpha selections are constrained and become a byproduct of asset allocation decisions.
- Á-°Investors may be inefficiently obtaining alpha through the traditional alpha and beta bundled product structure.
- Á-°Various constraints, such as the inability to short positions, in the traditional asset allocation process make it difficult for investors to efficiently allocate and utilize a risk budget.
- Á-°A traditional 60/40, 50/50, or 30/70 asset mix benchmark may not realistically represent a planÁ-?s liabilityÁ-¬
- Á-°Pension plans and endowments and foundations using traditional asset allocation may be inefficient in managing their investment fee structure.Á-?
á=
But the authorsÁ-? central argument is that alpha/beta bifurcation increases the available opportunity set. To may their point, they compare a Á-°bundledÁ-? alpha/beta portfolio whose return can be represented by…
á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á= 
Á-¬and an Á-°unbundledÁ-? alpha/beta portfolio whose return isÁ-¬
á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á=á= 
Portable Alpha skeptics will often point to the superior importance of beta.á= After all, they say, systematic market risk is positively compensated in the long run, “alpha is a zero-sum game”. Kung and Pohlman acknowledge this, stopping well short of advocating an all-alpha portfolio.
Á-°From a macroeconomic point of view, we believe beta is a necessity. It reflects the opportunities to participate in the growth or profitability of businesses in the real economy.Á-?
As a result, the authors advocate using a 100% (synthetic) beta portfolio.
Á-°All asset class exposures for the policy portfolio should be implemented via passive (preferably synthetic) instruments such as index futures, swaps, Exchange Traded Funds (ETFs), or index funds for major asset classes to minimize active beta risk and fees. Synthetic beta exposures enable investors to redeploy cash more efficiently into pure alpha strategies.Á-?
To this beta portfolio, they add a fully-funded alpha portfolio:
Á-°Skill-based pure alpha portfolios are typically represented by absolute return strategies that are free of traditional benchmark and long-only constraints and are not correlated with traditional asset classes.Á-?
This article draws an important link between Liability Driven Investing (LDI) and the alpha beta framework. LDI strategies use the present value of a pension planÁ-?s future liability stream as a benchmark. While this makes conceptual sense, it comes with operational challenges. Specifically, such a benchmark would require frequent re-balancing of the asset mix and liabilities fluctuate with interest (discount) rates. Kung and Pohlman point out that a separately managed beta portfolio facilities this frequent re-balancing without having to change active managers (or worse, interfering with the active managersÁ-? investing decisions).
In conclusion, the authors are pragmatic about the short-term adoption of their framework:
“Radical departures from the traditional approach are unlikely to be widespread in the short run. Initially, institutional investors need to take steps to implement separate pure alpha and beta allocation for major asset classes.”
Read Full Article (subscription required)
View previous postings on articles by Kung and/or Pohlman (here, here, and here)
Email This Post
Print This Post




