Why a global scope assures GTAA returns

In spite of recent reverses, global tactical asset allocation has developed from an add-on strategy to an integral part of a pension fund’s investment portfolio, according to a recent study. David White reports

Like a courtier falling in and out of favour in the courts of 16th century Europe, global tactical asset allocation (GTAA) has experienced both advances and reverses. In the 1990s, GTAA suffered some reputational loss following the poor performance of value-orientated managers who mis-read the equity boom.

Last August, IPE reported that GTAA strategies adopted by some Dutch pension funds had underperformed in 2006 and in the first quarter of 2007.

So is the place of GTAA secure in portfolio investment, or has recent performance put a question mark against the claims made for it?

On cue, a new study ‘Global Tactical Asset Allocation: Exploiting the Opportunity of Return Differentials across Asset Classes and Global Financial Market’ attempts to answer these questions. The authors are Daan Potjer and Chris Gould, respectively head of tactical asset allocation portfolio management and head of risk for TAA at ABN AMRO Asset Management.

They argue that GTAA has developed from a co-management role for balanced mandates into a specialised part of the investment management industry, with a range of specific products.

They suggest that most institutional investors have only a hazy idea of what GTAA is. This is not surprising, since the term is used to describe a wide range of products, including GTAA overlay, dynamic growth, dynamic asset allocation and fully flexible allocation. In their study, they try to distinguish the specific features that are common to all GTAA products.

In its broadest context, what differentiates GTAA from strategic asset allocation (SAA) is its time horizon Normally SAA takes a long term horizon. Potjer and Gould suggest the horizon of GTAA should be less than a year.

Yet the defining feature of GTAA is its scope. In the mid 1970s, the tactical asset allocation (TAA) strategy marketed by its inventor, William Fouse at Wells Fargo, was limited to US equities and US bonds. By the late 1990s and 2000s, GTAA had widened the scope of TAA to include currency selection, individual country selection, corporate and high yield bonds, and equity styles such as value/growth and large/small cap.

The results of this have been impressive in terms of the improvement in the information ratio (IR), the key test of manager skill. This measures the efficiency of converting additional risk into additional return.

Potjer and Gould illustrate this improvement with a Monte Carlo simulation that takes a random monthly return out of a dataset of 10 years of monthly return data on various indices. The simulation calculates the return of a random long/short strategy with a success ratio of 55% (only slightly better than the 50% success ratio of spinning a coin). This is done until all the months in the dataset are used, and then 1,000 times in a random way. The average result of the simulations is then used to calculate the IR.

The simulation shows that the broader the scope the higher the IR (see left). Using only two strategies - US equity versus US bonds and vice versa - produces an IR of 0.41. Allocating over various bond regions and including cash as an asset class, the IR is 0.58. Allocating over five equity regions, using 10 different strategies that can be used both long and short, increases the information ratio to 0.60.

The most dramatic improvement in IR is achieved when different asset classes and regions are combined. By using 45 different strategies, which can be used both long and short, the information ratio increases to 0.84.

The importance of diversification, rooted in the Fundamental Law of Active Management, is the basis of Potjer and Gould’s definition of GTAA as “an investment strategy that exploits the opportunity of return differentials or market inefficiencies across subsets of asset classes, global markets, investment styles, currencies and commodities”.

The diversification provides room for mistakes, Potjer and Gould point out. It is not necessary to make a right call on every long/short strategy. What is important is that the balance of calls should be right.

GTAA’s scope of global asset classes, markets, currencies and commodities with the use of long/short positions means that it shares some similarities with global macro hedge fund strategies.

There are significant differences that make these two products quite distinct from each other, Potjer and Gould maintain. The main distinction is that hedge funds use cash leverage - that is, they borrow cash either to invest in other assets or for margin to achieve a larger leverage of exposure than would otherwise be possible.

GTAA funds make use of leverage exposure - that is, taking a long or short position via futures or forwards where the net exposure is larger than the underlying portfolio, but they do not borrow cash.

What also distinguishes GTAA strategies is that they focus on liquid rather than illiquid markets. Most GTAA managers do generate their returns - often exclusively - from the most liquid parts of the investment strategy.

This has two advantages. First, there are no constraints on capacity. As long as they can make money consistently through a process that works in the most liquid global markets, GTAA managers can add to assets under management without hitting capacity constraints. Consistent strong returns from investment in liquid markets also means that GTAA managers can meet competition from new players.

The attractions of GTAA in an investment portfolio are linked to its defining characteristics, Potjer and Gould suggest. GTAA strategies are always long/short, and they can be scaled in terms of size and the risk budget given to the GTAA manager. In a futures overlay mandate, for example, a GTAA manager can as easily buy or sell, two, 20 or 2,000 futures.

An added attraction is that this scaling can be done by clients as well as the GTAA manager. A pension fund can specify the precise amount of active risk it wants a GTAA manager to take. The GTAA manager can adjust long/short positions quickly to the new risk budget.

One of the biggest changes in GTAA over the past 20 years has been in the way it is managed. In the past, the qualitative element of the alpha generation process was provided by the monthly meetings of a committee. The committee might include the CEO, the CIO, and heads of the specialist teams managing different asset classes.

The purpose of these monthly meetings was to provide a consensual, ‘big picture’ view of the global economy and the outlook for markets over the medium term horizon of three to 12 months.

This was appealing in theory, since it captured the potentially high return, longer-term investment themes of market as provided by a diverse group of specialists. In practice, the monthly, weekly or even daily agenda of committee meetings risked missing important short term changes in the market. “The market never sleeps,” as Potjer has observed.

The committee has been replaced by a dedicated GTAA team, which can react quicker to events since it conducts its own research and formulates its own asset classes assessments.

The real test, however, is performance. Potjer and Gould point out that comparing the performance of different GTAA managers is difficult, because of the varying degrees of scope and time horizons. Depending on the risk budget, GTAA portfolios can produce wildly divergent performance.

The guide should be the information ratio, since this provides greater insight than pure performance. Consultants and their pension fund clients should look at the high IRs first, and then look at the actual risk budget and alpha relationship.

Another problem with assessing performance is that investment processes are inevitably obsolescent. Potjer and Gould suggest that it is unlikely that the structure, content and process used within a GTAA portfolio today is the same as it was five or even three years ago.

Ironically, GTAA is one of the few areas of investment where a short track record may tell the investor more than a long track record The investment results obtained in the past may reflect an entirely different process from the one in use today.

For this reason, Potjer and Gould suggest, it may be more useful for pension funds to consider a shorter performance track record for GTAA managers than they would for the managers of other investment strategies. In other words, the jury may be still out on the performance of GTAA managers.