Wednesday, December 19, 2012

“Alpha, Beta, and Now… Gamma”

Before getting started, a quick note: my third column is now available at RetireMentor. It provides my latest effort to summarize my overall thinking on the topic of retirement withdrawal rates and stock market valuation levels.

Today at GRIPS in Tokyo, David Blanchett presented a paper he authored with Paul Kaplan called, “Alpha, Beta, and Now... Gamma.” Just for the record, GRIPS has transformed itself into quite a center for financial planning research this year, with seminars from luminaries including Somnath Basu, Joseph Tomlinson, and David Nanigian, and next week we will cap off the year with presentations from Michael Finke and Sandra Huston. If you happen to be in the Tokyo area and (for whatever reason) haven’t contacted me before, please note that these seminars are open to the public. Contact me for further details.

The aim of Blanchett and Kaplan is to show the value to retirees from making better financial decisions. They do this by comparing outcomes for a naïve investor in retirement again someone who makes good decisions. Financial planners could use the results of this article to help explain to clients how they can provide value. Do-it-yourself retirees can quantify how better financial decisions improve outcomes.

Blanchett and Kaplan use the Greek letter gamma (which follows alpha and beta) to quantify the benefits received from making better financial decisions in terms of the potential for increasing retirement income on a utility-adjusted basis.

The dimensions for improving financial decisions considered in their paper include:

Total Wealth Asset Allocation

What’s the issue: Making asset allocation decisions after considering total wealth including human capital

Naïve investor: Makes asset allocation decisions while ignoring the role of lifetime human capital. The assumption is that a naïve investor holds a fixed asset allocation of 20% stocks and 80% bonds.

Improved Outcome: Calculate the present discounted value of lifetime earnings which will be saved. Determine the characteristics of lifetime income in terms of whether it is more bond-like or stock-like. Consider this as an asset in your portfolio and then figure out the asset allocation for the financial portfolio in order to obtain the final overall desired asset allocation for wealth. The overall market portfolio suggests that the overall asset allocation including human capital is 45% stocks and 55% bonds.

Dynamic withdrawal strategy

What’s the issue: Making withdrawal decisions using the mortality-updated constant probability of failure

Naïve retiree: Uses the 4% rule. Takes out 4% at retirement, and then increase that amount by inflation in subsequent years for as long as possible until wealth is depleted.

Improved outcomes: Dynamic decisions based on a circular process. Each year: Determine retirement horizon, determine asset allocation, and then determine the maximum withdrawal percentage for a given target probability of failure. Repeat this process annually. (For more details, this is the approach described in an article I reviewed before at the blog called, “Optimal Withdrawal Strategy for Retirement-Income Portfolios.”)

This lets you spend more when markets do well and make cuts to preserve the portfolio when markets do poorly. On a utility-adjusted basis, it can provide higher overall lifetime income that the simple 4% rule. [more about utility-adjusted spending can be found in the above blog post… simply put, different strategies result in different amounts of volatility for spending, which people don’t like, and this makes an adjustment so that different spending strategies are comparable]

Annuity Allocation

What’s the issue: Using product allocation to devote some financial assets to purchase guaranteed income products may improve outcomes

Naïve investor: Views annuities as a gamble about dying too soon and ignores them as a retirement income option.

Improved outcomes: View annuities as insurance against outliving one’s wealth by relying on the guaranteed income for life. In this analysis, the role of annuities is investigated by assuming one uses 25% of financial assets to purchase a SPIA. Specifically, since SPIAs provide bond-like income, bonds are sold to purchase the annuity, increasing the stock allocation as a percentage of financial assets, but keeping the same stock allocation from the perspective of total assets including the present value of the annuity.

With total wealth allocation, the total market portfolio consists of 45% stocks and 55% bonds. If you buy an annuity, you transfer some assets into a guaranteed income for life. If 25% of assets are annuitized, the nature of the income is very bond-like, and the suggestion would be to consider that SPIA allocation as a bond. The asset allocation is then 45% stocks, 30% bonds, and 25% SPIAs.

Tax Efficiency Through Asset Location and Withdrawal Sequencing
What’s the issue: Creating better tax efficiency by locating assets in the most tax efficient places and withdrawing assets in a more tax efficient manner.

Naïve investor: Ignores these issues by keeping the same asset allocation for both tax-deferred and taxable accounts, and then withdrawing proportionately from each account in retirement.

Improved outcomes: Efficient asset location is to fill tax-deferred accounts with bonds, while stocks would be used in taxable accounts as much as possible. Efficient withdrawal sequencing is to first spend down taxable accounts, and then spend down tax deferred accounts.

Liability Relative Optimization
What’s the issue: The true risk for a portfolio which exists to financial retirement lifestyle spending goals is NOT the standard deviation of the asset portfolio, and is NOT the performance of the asset portfolio relative to peers. Rather, it IS the risk that you won’t be able to meet your spending goals.

Naïve investor: Makes asset allocation decisions without regard to spending goals by focusing instead on single-period Modern Portfolio Theory concepts.

Improved outcomes: Make investment decisions specifically with spending liabilities in mind. This could result in a portfolio with a lower expected return and/or a higher volatility than one on the traditional efficient frontier of Modern Portfolio Theory, but it might nonetheless do a better job meeting lifetime spending needs. Adding a liability creates a different efficient frontier with portfolios that would have previously seemed suboptimal. For instance, TIPS might not play a role in mean-variance optimization, but it might do a better overall job meeting spending needs, especially in high inflation environments.

Improved Financial Decisions
By making these improved financial decisions, retirement income can be increased dramatically. On a utility-adjusted basis, a comparison of the naïve case to the improved income case shows an increase in retirement income of 29%.

David also discussed in his presentation about an additional result not found in this paper: optimizing the Social Security claiming decision has the potential to raise retirement income by another 9%. Now we are talking about a 38% increase in income, in utility-adjusted terms.

Leaving Social Security out, The breakdown about how improved decisions can improve outcomes from the baseline include:

Dynamic Withdrawal Strategy        8.5% more income
Tax Efficiency                     8.2% more income
Total Wealth Asset Allocation      6.1% more income
Annuity Allocation                 3.8% more income
Liability Relative Optimization    2.2% more income

They also ask the question: how much is this worth in alpha terms? In other words, how much would portfolio returns need to be increased to support a 29% larger spending level? The answer is that over a 30-year period, those starting with a 4% withdrawal rate would need to earn 1.82% more per year (in the median case) to increase income by this amount. This is the “gamma-equivalent alpha.” Improving financial decisions has the potential to improve outcomes dramatically. It’s like earning 1.82% more per year on your financial portfolio, relative to the baseline case for the naïve investor.

No comments:

Post a Comment