| Dynamic Asset Allocation 101 |
|
|
| 10/06/2007 | |
|
Dr. John R. Birge, of the University of Chicago Graduate School of Business and Quantstar Corporation, answers this week's questions on dynamic asset allocation. 1.What is dynamic asset allocation? Dynamic asset allocation is an investment strategy that adjusts the allocation of funds across asset classes in response to changes in forecasts of future returns or to changes in the portfolio's position relative to the investor's objective. As an example, suppose an investor has a goal of earning enough from investment in either stocks or bonds to grow his or her portfolio enough in two years to make a down-payment on a "dream" house. With a fixed fraction invested in the stocks in each year, the investor may not be able to achieve the goal with any greater than a 50/50 chance. A dynamic strategy, however, may be able to achieve a substantially higher (75% or more) chance of achieving the investor's goal. In this case, the investor "locks in" the home down-payment if stocks perform well in the first year by shifting to a 100% bond portfolio with little risk of missing the housing goal. In the other situation where stocks perform poorly, the investor still has a chance of meeting the goal by increasing the stock investment so that good stock performance would lead to achievement of the goal. 2.What are its benefits to investors? In general, dynamic asset allocation can improve the overall performance of the investor's portfolio. Investors can benefit by reducing risks while achieving increased returns. In particular, dynamic asset allocation can provide protection against losses while still allowing for growth. 3.What are the theoretical and practical constraints and challenges of dynamic asset allocation? Traditional theory states that investors should hold a fixed proportion of their portfolio in a properly diversified risky portfolio with the remainder in a riskless asset. Dynamic asset allocation addresses the limitations of that theory by considering practical issues such as varying interest rates, incomplete information dispersion, diverse investor preferences, and transaction effects including limited liquidity and taxes. These variations to standard portfolio theory present challenges to computing optimal allocations even in theoretical models. To enable practical implementation, computational models necessarily use simplifying assumptions about market behavior and the relationships among prices, economic conditions, and investor preferences. Transaction costs present one example of a challenging consideration for practical implementation. The total cost of a transaction depends not only on the spread between bid and ask prices but also on the depth of the order book at the time of execution. Small changes in these values can lead to substantial changes in asset allocations. The following graph gives one example for low, medium, and high transaction costs and the effect on risky investment allocations in an optimal portfolio.  a4.Can you outline some of the more recent developments in the field of finance in relation to Dynamic Asset Allocation? Recently, several new methods and models for dynamic asset allocation have been developed and enabled by rapid increases in computational capability. These developments particularly relate to structured products, such as annuities with guarantees. The new computational methods build on statistical methodologies that impose fewer restrictions on possible relationships and on optimization technology that takes full advantage of model structure. The statistical methodology builds on recent work in the use of Markov Chain Monte Carlo methods for building posterior probability distributions without resorting to restrictive assumptions on the class of distributions. The new optimization work has progressed in both continuous-time extensions, where simulation and dualization of the optimization problem can allow consideration of more realistic conditions, and in discrete-time (stochastic programming) models, where combinations of simulation and advance numerical routines enable solution with increasing numbers of assets and re-balancing periods under consideration. |
|
Articles of the same Serie : 101 |
|
© Copyright 2008 bfinance. This document is for your personal non-commercial use. Any further copying, reproduction, distribution is strictly prohibited. To obtain permission please contact This e-mail address is being protected from spam bots, you need JavaScript enabled to view it