7 Questions for Scott Burns on AssetBuilder

I was invited to interview Scott Burns of Couch Potato Investment Portfolio fame about a month ago. I held off posting the interview because of the credit crisis.

I will post it in two parts. Part 1 focuses on  7 questions I asked him about AssetBuilder and the book, “Spend ’til The End” he co-authored with Laurence J. Kotlikoff.

clip_image002

For the purposes of this Interview I assumed I was a 60 year old Baby Boomer in retirement, living in the USA, own my own home, have no debts and have $500,000 nest egg ready to invest. My wife does not work.

I filled out the 4 Question Survey as follows:

  1. $500,000
  2. 1 year or less because I am in retirement
  3. $1,667 per month (I’m using the 4% rule as I have no information on my consumption smoothing yet)
  4. Conservative. I want Capital Preservation at least in the early years. I do not want to lose money in retirement in the early years, because it will be too hard to replace it if I have no other source of income.
  5. Results are:
    • Expected Return $1,850,647.00 with Expected Standard Deviation of 5.36%
    • I understand these results include Fund Fees but not AB’s Advisory Service Fees
    • According to the chart I can increase my withdrawal amount to $4,000 a month in which case I will still have $310,163 left after 20 years.
    • The chart is very smooth so I assume it is using average returns. Is this correct?
    • How “real” are the expected returns given the increasing market volatility of late?

The chart is the result of using the figures above. Unfortunately the two questions above did not get responses. I believe they were just missed when Scott went through the emailed list of questions.

Here are the other 7 I asked;

1. The Performance Table above is representative of a compilation of the selected funds to achieve a probabilistic return for a measured level of risk. How is it done and how does it compare to “real” returns for those funds compiled together?

The expected return and standard deviation are just that, “expected”— they are the result the you would expect over a long period of time. Results will vary significantly over shorter time periods. You can get an idea of how much variation by considering the standard deviation. If the expected return is 10 percent and the standard deviation is also 10 percent, that means your return will range between 20 percent and 0 percent in two of every three years. In the third year the performance will be outside that range.

2. How does AssetBuilder cater for Retirees as it is a 20 year expected return and not a 20-30 Year income distribution chart? Will my nest egg survive?

Portfolios in distribution face a significant danger of exhaustion and that danger increases as the time period increases. If you select a portfolio with a relatively low standard deviation, you reduce the danger. Select a portfolio with a relatively high standard deviation and you increase the danger. If your withdrawal rate exceeds the current income production of the portfolio, securities will have to be sold. In down markets this amounts to reverse dollar-cost averaging. Statisticians call this effect a “variance sink.” Biologists call it “the extinction problem,” referencing what happens to a population when a significant loss overcomes its reproductive capability.

The lower your withdrawal rate, the greater the odds of long term portfolio survival. The best you can do is play the odds. Remember, in a scenario that estimates a 90 percent survival rate, 10 percent of portfolios will be exhausted prior to death but the other 90 percent will have a broad distribution of final values.

3. This is just a single instance chart of expected future returns. Or is there a Monte Carlo Simulation behind it? How are these expected returns calculated for each asset class into the future?

We start from historical returns and standard deviations, then make adjustments to reflect recent experience. We also back-test to see how much an significant change in expected return for an asset class would effect overall portfolio performance.

4. Why not use real returns and show how the portfolio would have performed over the last 10 years? (I understand some funds are less than 10 years old so should only be used from their start dates)

Return figures for the model portfolios are posted monthly on our website. We don’t go back 10 years because many of the funds don’t have that much history.

5. How would I take a pension from my AssetBuilder invested funds in good and bad years?

There are a variety of approaches to doing that, all of which work by different “rules” for modifying your withdrawals. You can find columns about some of these approaches on the website. You can also read the original papers in sources like the Journal of Financial Planning. Backtesting with different portfolios indicates that withdrawal rates can be as high as 6 percent if some restrictive rules are followed.

6. You mention in the book that diversifying your portfolio is generally a bad idea but Asset Builder does that doesn’t it? Can you explain why the book says one thing and the AssetBuilder says another? Or have I misunderstood something?

I think you have forgotten the context. In the book we discussed diversification in the context of incorporating your human capital into your planning. At AssetBuilder we design portfolios with a range of expected returns and risks, hoping to provide the highest possible return for any given level of risk. That’s what we mean when we talk about being “risk efficient.”

7. I have seen the charts on the web site and I note that none of them made a loss during the 2000-2033 stock market slump. I know these are back-tested results  and some funds have been adjusted back to 2000, but how much confidence is there that this is what would have happened to a real portfolio?

You can be very confident that those are the results that a real portfolio would have had during that specific period of time. That should not be interpreted, however, as a guarantee against losses in any future period. To us, the issue is providing the highest return at the least risk— maximizing the odds that loss periods will be limited.

8. You’ve written a report with Brook Hamilton called “Reinventing Retirement Income in America”. In it you proposed a new 401(k) plan called  the American Freedom 401(k) and then go on to list many of the attributes that should be in that plan. It seems a good idea. How far along has this progressed and is it something you will be adding to AssetBuilder to expand it’s reach into the employee 401(k) business? It seems to me from my readings that this area is still lacking the attention it deserves – getting employees to maximize their employee 401(k)’s or even to contribute in the first place.

Many of the recommendations in that study are becoming standard in company plans, particularly automatic enrollment. In addition, new legislation will force providing more information on costs and the mutual fund companies are clearly reducing expense ratios, at least for companies with large asset accumulations.

I’ve not added any comments to the answers given by Scott. Some of his answers have generated several more questions in my mind. But for now I do appreciate the time and effort Scott took to respond to them.

Part II will follow in a few days.

Leave a Reply

CommentLuv badge