Asset classes, strategic asset allocation and portfolio construction

How are asset classes specified?

Memorize

An asset class is a category of assets with similar characteristics, attributes and risk–return relationships. The strategic asset allocation is stated as percentages of asset classes, so how the classes are defined decides how much control the investor has over the risk and return of the final portfolio.

Intuitively, a class should hold similar assets and diversify the other classes. In statistical terms, the following criteria apply.

Asset classa category of assets with similar characteristics and risk–return relationships

• Similar inside Memorize similar risk and return expectations, relatively high correlation between members
• Different across Memorize lower correlation with other classes, so they diversify
• Mutually exclusive Memorize an asset belongs to one class only
• Cover the universe Memorize together approximate the investable universe

Splitting is a choice. Bonds can be one class, or government and corporate bonds, or four classes: domestic and foreign government, investment grade and high yield corporates. The finer split lets you state separate risk and return expectations and correlations for each, which matters when their behavior really does differ. Equities are often split into domestic and international, developed and emerging markets, or large and small caps.

Correlations tell you whether a split is worth it. Equity classes tend to be highly correlated with each other, while government bonds are far less correlated with equities and can even be negatively correlated. A split can still be justified for other reasons: managers and products are organized around specific areas of the market, and some regulations make the definitions mandatory.

Asset classes are usually represented by benchmarks or universes from index providers. If a client excludes certain companies or sectors (negative screening), the standard benchmark may not describe the opportunity set any more, and a custom benchmark that reflects the exclusions is needed.

How does the IPS become a strategic asset allocation?

The capital market expectations are the investor’s expectations of the risk and return prospects of asset classes: expected returns, standard deviations and correlations. The expected return of a class is the risk-free rate plus one or more risk premiums. Expectations come from historical estimates, economic analysis and valuation models.

The strategic asset allocation (SAA) is the set of exposures to IPS-permitted asset classes that is expected to achieve the client’s long-term objectives given the risk profile and constraints. It combines two inputs: the objectives and constraints from the IPS, and long-term capital market expectations. In theory the two are combined by optimization (mean–variance) or simulation. The SAA may also state a policy on hedging risks, such as currency risk, interest rate risk or inflation risk, often through overlays of derivatives.

The IPS draws the box (the return you must earn, the risk you can accept); the frontier picks the mix. If no frontier portfolio sits inside the box, the objectives are inconsistent and the IPS needs revisiting.

Why does the SAA get so much attention? A portfolio’s systematic risk accounts for most of its change in value over the long term, and the returns of groups of similar assets predictably reflect their exposure to systematic factors. Asset allocation is how those exposures are chosen and sized to match the risk and return objectives. Adding classes with low correlation improves the risk–return trade-off. In general, risk-averse investors hold large weights in government bonds and cash, while those with more willingness and ability to take risk hold more equities and alternatives.

Go deeper Utility and the optimal mix

An investor’s preference can be written as a utility function that rises with expected return and falls with risk, for example $U = E(R_p) - \lambda\sigma_p^2$, where $\lambda$ measures risk aversion. Plot the efficient frontier from the capital market expectations; the optimal SAA is the frontier portfolio that gives the highest utility and respects the IPS constraints. The lessons on risk aversion and the efficient frontier cover the mechanics.

Go deeper Asset–liability allocation for pension plans and insurers

Institutions with liabilities often frame the allocation relative to those liabilities, using the correlations between the asset classes and the liabilities. The liability-driven approach uses assets, often bonds, whose cash flows track the obligations and invests the surplus for return. The SAA is then reviewed against the funding position, not only against the market.

What do tactical allocation and security selection add?

Memorize

The SAA is not yet a portfolio. The next step, especially for quantitatively oriented managers, is risk budgeting : deciding the overall risk to take (set when the IPS was written) and dividing it between the sources of return. Apart from the exposures in the SAA, there are two:

• tactical asset allocation (TAA) : deliberately deviating from policy weights to add value from forecasts of near-term returns, for example holding more equities for a quarter because you expect them to beat bonds;
• security selection : trying to beat an asset class’s benchmark by choosing securities with higher expected returns, for example holding more of a stock than the benchmark does and funding it by selling another.

Both add uncertainty over and above the risk in the policy portfolio, so the IPS should set risk limits and desired payoffs for each activity.

Portfolio constructionIPS + capital market expectations → strategic asset allocation → risk budget → implementation

• Strategic asset allocation Memorize long-run policy weights; exposure to systematic risk
• Tactical asset allocation Memorize deliberate short-term deviations from policy weights
• Security selection Memorize choosing securities within an asset class to beat its benchmark
• Rebalancing policy Use it rules for bringing weights back after drift
  • Calendar or corridor rebalance periodically, or when a weight breaches its band

Is security selection worth its share of the budget? Unlike the SAA, selection has no long-run payoff to risk: it is a zero-sum game before costs, because all investors in a class together earn the market return. Active managers trade more and pay for research, so the average one trails the market after costs. That does not mean no manager has skill. The chance of adding value depends on the manager’s skill and on how efficient the market is: a market for large-cap equities is very efficient, while some regional bond and equity markets are less so. Sometimes the choice is made implicitly, for example in non-listed real estate and infrastructure, where a diversified index-like exposure is hard to buy, so taking part means selecting assets.

The return from TAA is the difference between the actual portfolio and a portfolio held at policy weights. For each asset class, multiply the weight difference by the class return and sum.

$$C_{TAA} = \sum_i \left(w_i - w_i^{policy}\right) R_i$$

Because the two weight differences are equal and opposite, a two-class TAA contribution is just the weight shift times the return gap: 0.05 × (8% − 2%) = 0.30%.

When do you rebalance?

Use it

As asset-class returns differ, the actual weights wander from the policy weights. This is drift . The rebalancing policy in the IPS sets the rules: rebalance periodically, or whenever a weight breaches its band (the corridor), selling what has grown too large and buying what has shrunk. Even without a formal risk budget, a rebalancing policy is an important element of risk management, because drift quietly changes the risk of the portfolio. Weights also move for a second reason, deliberate TAA bets; monitoring separates the two.

Go deeper How wide should a corridor be?

A narrow corridor keeps risk close to policy but triggers frequent trades with transaction costs. A wide corridor saves costs but lets risk drift. The right width depends on trading costs, volatility and how much the asset classes are correlated. The portfolio is also reviewed as capital market expectations and the client’s circumstances change, which can lead to a new SAA.

What newer approaches exist?

Know the idea

• Core-satellite. Many portfolios use a core-satellite structure: a low-cost, passive or low-tracking-error core holds broad exposure, and a few smaller active satellites aim for extra return.
• ETFs and robo-advice. Exchange-traded funds track an asset class index cheaply and trade easily. Combined with algorithm-based advice (robo-advice ), they give retail investors fast, inexpensive, liquid and diversified exposure.
• Risk parity. Critics of mean–variance methods point to their sensitivity to small errors in forecast returns and correlations. Risk parity weights asset classes by their contribution to risk. Supporters note that a typical portfolio with 60% or more in equities gets most of its risk from equities, which are far more volatile than bonds. Opponents argue that its good results after the 2007–2009 crisis came from the long fall in interest rates, which helped bonds.

Exam tip Strategic is the plan, tactical is the deviation, selection is the stock-picking. Name which one the question describes before you calculate.