Harold Evensky, an internationally recognized authority
on investment and financial planning topics, explains
why traditional concepts,
such as the income portfolio
and Monte Carlo simulation, can
lead to imperfect
decision making. Our interview with Harold
how his firm, Evensky & Katz, constructs portfolios to
fund the Distribution Phase of a retirement account.
Let’s begin by discussing what seems to be a common problem in financial planning – the paycheck syndrome. Can you explain this syndrome and how it leads to the use of income portfolios, a practice that you have widely criticized?
My partner, Deena Katz, and I co-edited a book called the Investment Think Tank (View Link) a number of years ago. Our focus in that book was on how to save and invest money, not how to get it out once you are retired. Now we have turned our attention to the retirement phase of financial planning, which we believe is and will continue to be a great challenge for the financial planning profession.
Deena coined the phrase ‘paycheck syndrome.’ Our research shows that people are used to having income from a paycheck that is regular and dependable, and in retirement they look for the same kind of solution. Our profession traditionally addresses this through the income portfolio, which is constructed of securities that pay dividends and interest, to make it feel like a paycheck. We believe this is nonsensical and dangerous. Income portfolios are highly dependent on interest rates and dividends, not what the client really needs, which is financial security. Income portfolios ignore modern portfolio theory, in that they do not seek an optimum risk/return tradeoff. The biggest problem is that when rates go up, clients get more money, and then spend more money. But this is precisely when the value of their portfolio is going down, creating potentially disastrous consequences. Professionals know that a portfolio should be designed around a total return strategy, and a cash flow strategy should be build on top of that.
You have articulated a number of risks that are inherent in the distribution phase of a retirement account. What are these risks, in particular those that are not well understood or recognized as being a risk?
Lots of risks are unique to distribution phase – for example, longevity is now a well-understood risk, as people are living longer. But something else is happening along with that, and it not well understood. People are staying healthy, and they continue to spend money until they die. I took a cruise not long ago and I was the youngest person around. There are a lot of older folks spending a lot of money.
Advisors focus on the funding risk, or the risk of running out of money in the distribution phase of a retirement account. But there is also the risk of under-spending your money. I don’t want to see situations where retirees could have taken more cruises or seen their children more often.
Everyone understands inflationary risk. But not many understand that inflation can affect retirees more dramatically. A new statistic, called the CPI-E (the Consumer Price Index for the Elderly) shows that inflation may run a half a point higher for the elderly segment of our population.
Volatility risk in a retirement portfolio is more important in the distribution phase. Account sizes are bigger but, in the distribution phase, withdrawals are occurring, and the impact of poorly timing these withdrawals can have very harmful consequences. Accounts in the distribution phase are very sensitive to the size of withdrawals in the early years, as well as to when withdrawals start. Growth in a retirement account, like any account, is dependent on asset allocation and security selection but, as the withdrawal rate increases, retirees need consistent investment success (some might call this ‘luck’) to insure that their investments last as long as they do.
We pay careful attention to the fundamental rate of return assumptions used in planning the distribution phase of a retirement account. Equity returns over the last 20 years have been extremely strong, about 12%, but we are projecting real returns of 6% going forward (Editor’s Note: our recent interview with Josh Coval, Harvard Business School Professor of Finance, reflects similar thinking – View Article). With a balanced portfolio of stocks and bonds, after allowing for taxes, inflation, and fees, the net return to the client, after expenses, tax and inflation is likely to be around 2.5%. Our portfolio construction reflects this new reality.
We cannot influence returns, but we can control fees and lower taxes, which we doing through more extensive use of ETFs and less reliance on active managers. It can be a boring investment strategy, but we know it is an intelligent strategy and carries a greater probability of meeting our clients’ goals.
How does Pascal’s wager enter into this?
Let’s say you are told there is a 50/50 chance that God exists, and you can choose to live your life accordingly. If you choose to live as if God exists, and you are right, you are better off. If you choose to live as if God does not exist, and you are wrong, you are far worse off. If God does not exist, you gain or lose nothing. Therefore, it is safer to assume God exists. The point is that we need to look at the consequences of an outcome as well as the probability of an outcome. If we decrease our reliance on active management we may risk seeing our returns go from, say, 7.5% to 7%. The potential adverse consequences (the loss of 50 basis points of return) are minor, in comparison to the benefit of adding ½% to a 2 or 3% return on a portfolio.
How exactly do you decrease your reliance on active management?
In the equity arena we once had 12 different managers, plus more for foreign markets. More managers translates to more transactions and taxes and expenses. Now we take a sizeable portion of client portfolios, often around 80%, and invest primarily passively and in a tax efficient manner. The other 20% - the satellite portfolio – is where we introduce risk. It used to be that we spread risk throughout the portfolio; now it is focused on just the smaller satellite portion. We are not eliminating active management, just concentrating it where we think it can be justified.
Monte Carlo analysis is a staple of financial planning. You advocate its use, but only up to a point. Can you explain where and why Monte Carlo falls short, and how advisors should properly use Monte Carlo?
Monte Carlo is one of my favorite topics, and also something that annoys me the most. Many in our profession treat it as the Holy Grail. It is very powerful, but too often it is utilized as if it is an absolute answer. Consider the problem of funding a college education five years from now. We need to consider the cost of that education, adjusted for a rate of inflation, and the rate of return on the portfolio that will fund that education. With some simple cash flow discounting, we can determine a point solution. Monte Carlo allows us to introduce changes to the underlying assumptions – the inflation rate and the rate of return on the portfolio – using upper and lower boundaries and the distribution. The default distribution is normal; unfortunately reality is not necessarily normally distributed – for example, there may be a much greater probability of high inflation than low inflation. So, we need to estimate the range and shape of probability distributions. The second problem with Monte Carlo is not looking at the consequences of failure – for example, it the Monte Carlo analysis indicates an 80% of success, does the 20% chance of failure mean that our client cannot fund any of the cost of education or can “only” fund 97%?
Monte Carlo should be used as an educational tool and as a way to show a distribution of outcomes. But it cannot be the sole, or even the primary, input when making investment decisions.
What is the “Alpo Diet” and how do you protect your clients from this kind of a culinary regime?
I am worried that too many retirees will be eating dog food instead of steak in the next couple of decades. Many people who are not retired are not going to have the resources to maintain their standard of living. In our society there will be an unpleasant wake up call. The solution is good planning and dealing with reality. A recent study by New York Life asked potential retirees how much they expect to withdraw annually from their nest egg. The data was shocking – showing, first, that far too many people overestimate the rate at which they can safely withdraw funds and, second, that as many as 40% of the respondents did not have a clue as to the correct answer. This is the real challenge – educating people to think properly about retirement, factoring in inflation and all the other risks I have mentioned.
Can you explain the history of the Evensky & Katz Strategy (the E&K S)? When did this unfold and why is it better than traditional methods of financing retirement planning?
In the early 1980s we recognized that volatility was the biggest problem, exacerbated by withdrawing money at the wrong time. We determined that the income portfolio made no sense, and we needed a design that created total return, but with the ability to withdraw funds as needed. Clients needed consistent, real cash flow, not dividends and interest that are dependent on market levels. We needed a strategy that offered financial flexibility, because needs change over time. Lastly, we had to have something simple; if it is too complex, clients won’t understand it and will not buy in.
The E&K S bifurcates the nest egg, based on our five year mantra. We believe that investments should be made with a five year time horizon for liquidation. We create an investment portfolio (the IP) and a cash flow reserve (the CFR). The CFR is designed to fund a two-year time horizon, and uses money markets and short term fixed income. Each month we withdraw 1/24 of these assets and forward these funds to our clients’ checking account (i.e., their “payroll check”). In the IP, asset allocation is adjusted, as necessary, based on the client’s risk tolerance and investment objectives to adjust for any opportunity cost associated with the CFR account. The key is that we deal with cash flow issues independently of investment decisions, and we control cash flow very easily.
Lastly, you have stress tested the E&K S under some very harsh conditions. How did it fare?
The E&K S has survived the 1987 crash and the bear markets of 2000 and 2001. Clients knew where their grocery money was coming from and never had to sell portions of their IP. The bottom line is that it works.
Many advisors today employ similar techniques, and we do not assert that the E&K S is the best strategy for everyone. The most common mistake advisors make is not paying attention to the process they are using. Most everyone agrees with the basic assumptions we use, but often err by not looking at the consequences of those assumptions under all scenarios. Advisors need to consider the probabilities of success and failure, and the consequences of each possible outcome.
Harold Evensky’s most recent book is Retirement Income Redesigned: Master Plans for Distribution: An Adviser’s Guild for Funding Boomers’ Best Years. It discusses the above topics, plus much more, and is available at Amazon.