Financial Education

Black-Litterman optimization essentially reverse engineers a diversified portfolio to derive the implicit assumptions regarding risk, expected return and covariance among the constituent asset classes. Then investors can use their own views, combined with the strength of their belief in those views, in a Bayesian process to adjust those weights.

Some studies have shown that asset allocation decisions explain 90% or more of the total return to portfolios, leaving little benefit for time spent on market timing or security selection decisions. However, a flaw in these studies is that many pertained to institutional investors whose decisions were primarily allocation decisions. Naturally, allocation explained a good deal of these returns.

Cross-sectional studies showed a lower but still significant (about 40% of total variation in portfolio returns) contribution from asset allocation.

Later studies have shown that while asset allocation is important, the dispersion of results can vary more from successful security selection or market timing decisions. These studies suggests that truly skillful investors will achieve more benefit by spending their time on timing and security selection rather than asset allocation.

Economic growth trends refer to the long-term path of GDP, and is an important input to discounted cash flow models of expected return. This is distinct from the fluctuations of the business cycle, which move around the long-term trend. Trends are typically easier to forecast than cycles, but can be altered by sudden changes called shocks.

Economic growth trends are determined by growth in labor inputs (the size of the potential labor force and the growth in labor force participation) and from labor productivity (capital inputs and total factor productivity growth).

Governments can influence economic trend growth by implementing sound fiscal policies, avoiding intrusion into private sector activities, encouraging private sector competition, supporting infrastructure and human capital developent and having simple, transparent and stable tax policies.

Shocks are events external to the economy that nonetheless alter the economic course. ALthough they cannot be predicted, they occur fairly regularly and could have short-term or long-term impacts. Most frequently shocks are related to either oil or financial crises.

Investors must rely on economic data to form their expectations for investment performance and develop a strategic asset allocation. It is important to realize the limitations to using such data, including the definitions, construction, timeliness adn accuracy of the data, including any biases.

Much data is reported with a time lag. The greater the lag before information is reported, the greater the risk that it is not pertinent to the present situation.

Given the uncertainty related to data collection, data are often subject to revision. In essence, the initial report is an estimate, with the revision providing better data at the cost of a greater time lag.

Definitions and calculation methods can change over time. This can affect the validity of time-series data.

Finally, certain data get re-indexed periodically to reflect more current bases. This introduces the risk of mixing data indexed to different bases.

Insurance companies have obligations to pay out certain benefits (death benefits, annuities) in the future. Their investment policies should be set with these obligations in mind.

Risk Objective

Future liabilities and interest rate sensitivity generally result in lower risk tolerance. Many insurers segment funds, with liabilities covered by safe assets of similar duration and surplus funds invested in riskier securities.

Return Objective

Earn a positive spread over the rates paid to policyholders.

Constraints

Liquidity is not typically a concern. However, liquidity needs can become important if the insurer is responsible for high annuity payments or if there is significant interest rate volatility (reducing liquidity of assets.) The overall time horizon is long-term, though many insurers prefer to segment time horizons according to product classes. Insurers are commercial entities and subject to taxation, so after-tax returns are the critical measure. Insurers also face heavy legal and regulatory constraints on investment policy ranging from adherence to the prudent investor rule to restrictions on investments and valuation methodology. Unique circumstances vary by insurer.

The ability of a defined benefit retirement plan to accept investment risk is governed by several factors, including plan status, the sponsor’s financial status, common risk exposures between the sponsor and the plan, plan features and workforce characteristics.

Plans with a surplus (more assets than liabilities) are in a better position to accept investment risk. In such cases, however, the need to accept such risk (presumably in exchange for higher returns) is reduced.

A sponsor with less debt and higher profitability will be able to make up investment shortfalls with additional contributions. Plans sponsored by such employers can tolerate a higher level of investment risk.

A high (low) correlation between the sponsor’s operating results and the plan’s investment returns implies a lower (higher) tolerance for investment risk.

Plan features such as early retirement options or the ability to accept a lump-sum distributions shorten the plan’s time horizon and reduce risk tolerance.

Plans serving relatively young workforces, or workforces with a high ratio of active employees to retirees, can accept a higher level of investment risk.

Whether to rebalance a portfolio depends on the costs associated with rebalancing and the reduced tracking error resulting from doing so.  Research has demonstrated strategies to minimize transaction costs for a given level of tracking error, resulting in similar performance at lower cost relative to naive strategies observed in practice.

In the September/October 2007 Financial Analysts Journal, Brown, Ozik and Scholz demonstrate a derivatives-based strategy that offers significantly lower cost at the expense of being more difficult to design. This is due to the fact that the transaction costs and tracking error of such a synthetic strategy depend both on the cost of trading the derivatives and the length of time the derivatives are held.

Behavioral finance suggests that investors are loss averse, hold biased expectations and segregate assets. To accommodate behavior, portfolio construction should allow for both objective and subjective constraints. In addition, an integrated portfolio can be formed from layers of segregated assets.

Loss aversion means investors do not view risk as uncertainty but rather as the potential for gain or loss. Investors tend to place more weight on losses than on gains, and will actually seek risk to avoid a certain loss but avoid risk to achieve a certain gain, even when probabilities favor the opposite course of action.

Biased expectations arise from overconfidence about ones predictions of future outcomes and from overestimating the significance of rare events.

Asset segregation, or mental accounting, tends to consider different assets according to purpose or preference. The interaction between investments is often ignored.

It is generally accepted in corporate finance that higher levels of operating risk (operating leverage) and higher use of leverage (financial risk) will increase equity risk and thus investors will require a higher return from the stock.

In the May 2007 Journal of Accounting Research, Penman, Richardson and Tuna examine the positive relationship between the book/price ratio and subsequent stock market returns by separating book/price into operating leverage and financial leverage effects.

The authors define “enterprise book-to-price” as the net operating assets divided by market price and use it as a proxy for operating risk. Leverage is net debt divided by the market value of equity.  They find that operating leverage does has a positive relationship to subsequent stock returns, but that financial leverage has a negative relationship.

In the June 2007 Journal of Banking and Finance, Miffre and Rallis compare strategies for investing in commodity futures based on short-term momentum and long-term reversal, based on a variety of formation and holding periods.

Momentum strategies based on selling past losers and buying past winners generated positive and significant returns in 13 of the 16 combinations of formation and holding periods, with a significant portion of that return being derived from short positions in the losers. These strategies generate positive alpha and have low correlations with the returns on equity or fixed income securities.

The reversal strategies do not exhibit consistent outperformance in this study.