At ING Investment Management (ING IM), we consider Tactical Asset Allocation (TAA) to be a multi-disciplinary process. In this process, it is necessary to combine optimally fixed-income, equity and currency views together with macroeconomics, psychology, statistics and risk management. TAA should involve careful identification of all investment opportunities and their interdependencies. Expected return targets and investment risks must also be in harmony with the complete individual portfolio characteristics and client-specific objectives. Successful TAA is the result of a solid, comprehensive and yet tailored investment approach where the partnership between TAA manager and client is extremely important. This holds for the case of a single asset manager approach as well as for a multi-asset manager approach.

TAA – risk budget
Starting point in our approach is that every pension fund has its unique strategic goals and objectives for its pools of assets. By using amongst others ALM-studies, these client-specific goals and objectives result in a strategic asset allocation that maximises the chance of matching current and future liabilities and assets. Key in this respect is diversification over a wide set of asset classes, thereby minimising portfolio risk but maximising investment opportunities. This is exactly what Dutch pension funds have done on a large scale in the last couple of years. By far, the largest part of the 300 Dutch pension funds for which we manage assets, has greatly diversified its investments. Investments, for example, in real estate, private equity, emerging market debt, asset-backed securities, high-yield bonds, active currency and hedge funds, different style products, etc. are nowadays common or considered actively by our clients.
As soon as long-term goals will be obtained with a sufficient certainty, then short-term adjustments of the strategic allocation deserve attention. Picking-up additional return by shifting exposures over asset classes increases the likelihood that long-term goals are met and limits the contributions of the plan sponsor. Yet at the same time, short-term exposure shifts should under no circumstance endanger the long-term fund objectives. To manage this process effectively TAA should be managed by using a risk budget, which is an effective methodology to determine which exposures give most value. A good TAA risk budgeting plan requires at least an assessment and identification of:
1. Scope and impact of investment decisions, taking into account:

o Level of skills of the asset manager, or information ratio;

o Correlations between the tactical allocation decisions;

o The maximum allowed contribution to risk of each asset allocation decision;

o Frequency of expected reallocations over different asset classes.
2. Portfolio implications, taking into account:

o Impact on asset mix because of cash flows (eg, dividends, withdrawals)

o Impact on asset mix because of performance of asset classes.
It is important to realise the dynamics in this process. For example, market risks are time-varying and correlations between prior decisions and the nature of asset classes are not constant. At the same time, the client’s objectives and financial situation are also subject to change.

Investmentpolicy – ING IM’s TAA process
In its very essence, TAA allows investors to profit from an asset-class’s deviation from its fair value by actively either over- or underweighting this asset-class versus its strategic portfolio weight. In hindsight, examples of these opportunities are easy to give, eg, the overvaluation of Japanese stocks in 1988, the infamous tech bubble and the overvaluation of the US dollar between 1999 and 2001. Less pronounced deviations from fair value are even more common and can be interesting for tactical allocation decisions as well. Many small tactical asset allocation bets can provide a stable return enhancement. To profit from these price-deviations and the relative attractiveness of a certain asset class, a consistent, systematic process has to be applied which takes into account the client specific situation.
Fundamental for a good TAA-process is the skill to estimate the relative degree of an asset class’s under- or overvaluation versus other asset classes and versus its historic average. Our estimates are based on quantitative and qualitative processes and the experience of our investment staff. Models for price momentum, market valuations and investor’s risk tolerance are supportive for these quantitative decisions. Qualitative analysis focuses on difficult-to-capture elements, as, for example, the effects of the war in Iraq and market sentiments of rare nature like fear, panic, despair, greed, comfort, overconfidence and speculation. By doing so we are able to identify, interpret and judge rare events and circumstances. Consequently, the final allocation decision is the outcome of both the quantitative signals and human assessment.
In our process we use two types of analysis: fundamental and market analysis.
o Fundamental analysis investigates the relations between macroeconomic indicators and expected stock and bond returns. It relies on the premise that changes in earnings and changes in long-term interest rates fundamentally drive equity and bond returns.
o Market analysis describes the connection between trends in share prices, risk appetite (as measured by short-term changes in bond yields), market valuations and expected stock returns.
This process is illustrated in the chart below:
Fundamental analysis is based on analysis from our macro economists. Their analysis on key economic variables are used to forecast the three return drivers of bonds and equities:

o Future interest rate development, for which we use:
The ISM index, as an indicator for confidence
Fed funds/ short term interest rates
Inflation
Budget deficit

o Earnings growth, for which we analyse:
Change in the unemployment rate
Change in the capital utilisation rate
Leading indicator, as an indicator for GDP growth
ISM index, as an indicator for confidence

o Expected change in inflation

The fundamental analysis gives a positive, neutral or negative signal for equities (compared to bonds). The signal will be positive, when the expected return on equities is 3%-points higher than the expected return on bonds.
Market analysis is important because fundamental, long-term trends do not necessarily predict short-term trends. Although we firmly believe that long-term forces drive each asset class back to its fair value, short-term forces might cause that the deviation from fair value is more pronounced tomorrow than today. A clear example of this is of course the tech bubble during which the market was overvalued on fundamental analysis, but the overvaluation became larger and larger almost every day. Missing the tech rally would have provided a significant deviation from peers and potential investment opportunities would have remained unused. Also the 1999 overvaluation of the US dollar lasted at least through 2001. We think therefore that our analysis should also focus on the probability whether the misvaluation is likely to disappear relatively soon so that it makes sense to take already active positions.
Market analysis is based on three (lowly correlated) indicators:

o Price momentum
Research shows that markets tends to trend over periods of 3 to 9 months. This is found for several decades and in many equity markets. In other words, markets do not incorporate all information immediately in the current price. Markets are slowly digesting new information.

o Risk appetite
Over the past decades, bonds have become a very systematic indicator for risk aversion of investors. As soon as systematic, macro-economic or geopolitical risks are perceived, investors flock to government bonds, driving yields down. When risk aversion relents, bonds are traded in for stocks and consequently yields rise. In this case, it is not the yield level matters but the turning points and trends. As a consequence, yields generate buy and sell signals.

o Valuation reversion
The valuation reversion indicator uses a discount model, aggregated earnings forecast and interest rates to asses whether the current market valuations are correct. The theoretical fair value is compared with the actual price providing a valuation score. The valuation reverse indicator looks at the momentum in the valuation score. The model gives a positive signal for equities when the undervaluation weakens and a negative signal for equities when the overvaluation becomes less pronounced. That’s why we speak about a reverse valuation indicator.

Development overall view
The results of both analyses are recorded in a so-called scorecard, which is an important aid in the decision making process. This scorecard also serves as an efficient way to communicate the TAA decisions and rationale for these decisions.
In its decisions on the expected outperformance of equities compared to bonds, five ranges are distinguished – positive, mildly positive, neutral, mildly negative, negative
The magnitude of the bet depends on the views whereby the fundamental signal and the market signal either restrict or support the decisions. If one signal is positive, equities will not be underweighted; if one signal is negative, equities will not be overweighted; if signals are contradicting, equities will be neutral.

Risk concentration
Tactical asset allocation is not a decision that can be seen in isolation from other return generating strategies like stock picking and yield-curve positioning, as they are probably highly correlated among each other. Therefore, tactical return targets and acceptable risk levels should be determined ex ante, avoiding undesired risk concentration. From a portfolio point of view, risk concentration can be one of the most challenging elements to manage. A good example of undesired risk concentration is seen in the tech-bubble period. Suppose that the following TAA-allocations are consistent with a mandate that only specifies restrictions on individual decisions:

o Overweighting equities versus bonds,

o Overweighting US versus Europe,

o Overweighting the Information Technology sector versus the Financial sector,

o Overweighting Growth stocks versus Value stocks,

o Shortening duration.

The accumulated risk of all these separate decisions, which in itself might be perfectly understandable, can be large. At the end of the bubble the investment results of all these decisions were highly correlated and depended to a very large extent on a single economic factor, which was economic growth. The non-materialisation of economic growth expectations at the turn of the previous century has significantly negatively affected the returns from all these positions. This shows that a wrongly designed TAA process could result in non-intended and unnecessarily risk taking.

Client portfolio
For TAA to be optimally effective, the client-specific situation has to be taken into account, ensuring that assets match liabilities. This differs from an asset-only approach. ING Investment Management believes that professionals from different disciplines together with the client should determine the attainable investment goals and set targets and policy guidelines. To provide a solid TAA solution therefore means that the TAA risk budget should be verified, monitored and discussed regularly. This can only be done properly when the asset manager actively builds on a true partnership with its clients. Partnership with dedicated and trained professionals and skills to manage the scope of investment decisions therefore are key when it comes to successful Tactical Asset Allocation.
Ronald van Dijk, head of research, ING Investment Management Europe