I kicked a hornet's nest last week when I asked my colleagues about retirement calculators. It all started when I quoted Paul Kaplan.
Paul Kaplan, Ph.D., CFA, is Director of Research at Morningstar Canada. Paul is also author of "Frontiers of Modern Asset Allocation", which I reviewed last year.
I quoted Paul in an article last week about the simultaneous plunge in gold, stocks, and commodities. I said that we shouldn't read too much into Monday's convergence of correlations, and this is true. But then I quoted Paul to make a point about online retirement tools, and I wound up quoting him out of context.
My Critique of Retirement Calculators
Online retirement calculators are popular tools that help investors project the growth in assets over time, based on different mixes of assets. My concern arises when these calculators are used to:
- Mindlessly extrapolate the past
- Offer a precise estimate of the future
I am always wary of the potential for mathematical models to be misused by investors (including myself). So I made my point about retirement calculators by quoting Paul Kaplan, who wrote in his book "Historical results should not be blindly fed into an optimizer." (I love this quote and use it often.)
Unfortunately, Paul was referring to Mean Variance Optimization (NYSE:MVO), a process that uses expected returns, volatility, and correlations to optimize asset allocation. Paul pointed out this out to me, and he also noted that retirement calculators typically use Monte Carlo simulations, not optimization.
I checked the retirement calculator at Charles Schwab and it turns out that Paul is right: Schwab uses Monte Carlo simulations. Moreover, Schwab's retirement calculator uses assumptions about long-term market returns that seem quite reasonable to me, and which are not an extrapolation of past returns. (Extrapolation is especially worrisome for U.S. government bonds, which have enjoyed a 30-year bull market, and which are not likely to stage an encore.)
Oh, and One More Thing…
In my prior article I also said: "Monday's broad sell-off across asset classes is a good time to highlight the fatal flaw in static optimization models: Correlations change."
Paul responded thus:
I disagree, because the models themselves make no predications about such short-term events. You then say "Days like Monday are a reminder of the weaknesses of automated tools that help investors plan for retirement."
What you call a weakness I call a strength! If using a long-term planning tool helps keep investors' eyes on the long run and helps them ignore the financial journalists when they read too much into a one-day event, then I'd say they've done investors a service!
I stand corrected. My thanks to Paul for pointing this out, and for highlighting a proper role for retirement calculators.
Volatility Management for Retirees
I find it useful to run a retirement scenario for clients since they are surprised to find out how devastating a bear market can be during retirement. It all depends on the timing of the downturn: The portfolio is especially vulnerable early in retirement, since withdrawals at a cyclical bottom cannot be recovered. This makes it much less likely that the portfolio will last for the client's lifetime, so volatility management is critical for retirees.
A 60/40 mix of stocks and bonds is the classic solution, since bonds have historically had a low correlation with stocks. Unfortunately, most retirement calculators limit the asset classes to stocks, bonds, and cash, which I see as a problem:
- With Bonds: Let's say I use a 60/40 mix of stocks and bonds. Right now that means owning a lot of bonds when interest rates are at historic lows, and when Treasurys offer negative real returns. That's not a recipe for a happy retirement.
- Without Bonds: If I do not buy bonds, especially government bonds, then the portfolio will be much more volatile. This leaves it vulnerable to a market downturn early in retirement (the dreaded "sequence of returns" problem).
Annuities and Alternatives
Two of my favorite solutions for volatility management are annuities and alternative investments. Annuities get a bad rap because of high commissions, and retirees certainly must be wary of unscrupulous salespeople. But since annuities can offer guaranteed income for life, they are a valuable ingredient in the retirement planning, depending on the client's liquidity needs.
My other favorite solution is alternative investments, and I use the following:
- Real Assets to Hedge Inflation: These assets include precious metals, Real Estate Investment Trusts, and Master Limited Partnerships. I own the iShares Gold Trust (NYSEARCA:IAU), the Schwab U.S. REIT ETF (NYSEARCA:SCHH), and the JP Morgan Alerian MLP ETN (NYSEARCA:AMJ). These assets provide both income and an inflation hedge. In addition, IAU has a particularly low correlation to stocks, and it provides a hedge against tail risks.
- Alternative Assets to Reduce Correlation: This includes hedge fund and private equity strategies, which typically are not highly correlated with the stock market (depending on the strategy). My holdings include the Barclays S&P 500 Dynamic VECTOR ETN (NYSEARCA:VQT), the Arbitrage Fund (MUTF:ARBFX), and the KKR Alternative High Yield Fund (KHYZX). The VQT dynamically reduces exposure to equities as volatility rises; ARBFX is a merger arbitrage fund, and KHYZX is a high-yield strategy that relies on the same fundamental credit research that KKR uses for its private equity strategies. Thus, even though it is a high-yield fund, it tends to be more defensive than most, since the portfolio management process is focused on downside risk.
Since I am strategically underweight in U.S. government bonds, I find these assets helpful in providing volatility management for retirees. (This example is for small accounts; high net clients may add traditional hedge funds, venture capital, etc.) Retirement calculators may not be able to use all of these strategies, but I find them a useful exercise to get clients to focus on long-term goals.
A Stinging Critique From Ed Stavetski
Not everyone is a fan of these calculators, however, as I learned when I asked another colleague in the investment business, Edward Stavetski.
Ed is the founder of PCM Partners LLC, a firm that provides asset allocation research, due diligence, and business valuations. So for Ed, retirement estimates are not an abstract math problem. Ed reframes the issue of retirement planning like this:
I prefer to start an asset allocation based on what the client requires, rather than starting with capital markets. I take their needs and add a premium (call it an inflation premium or a risk premium). Then I adjust the market exposures based on the client's risk profile.
By the way, risk is not relative to an index, and risk is not volatility: Risk is the permanent loss of capital.
I like Ed's perspective since it shifts the focus onto the goal of the portfolio, and since it redefines risk in a way that makes sense for the client. Sometimes people can be too glib about investment volatility, especially when they imply that "it all works out in the long run." This overlooks the fact that people retire at a relatively fixed point in time, and volatility early in retirement can be a killer.
The impact of volatility on retirement income was called the Retirement Red Zone by Prudential. In this marketing campaign, Prudential compares the years before and after retirement to the red zone in American football. I think this is a clever way to highlight how volatility reduces returns when an investor is withdrawing money. And let's face it, talking about the Retirement Red Zone is a lot catchier than talking about "the sequence of returns problem for portfolios in distribution."
Models Behaving Badly
Ed's critique of retirement calculators is actually a general critique of quantitative analysis, particularly when it is misused to the detriment of clients:
Modern Portfolio Theory, Mean Variance Optimization, and Monte Carlo are all the same. Whether it's expected returns, historical returns, volatility or correlations, they are all based on historical views and forecasts of future reactions. Academics for decades have ridiculed tactical asset allocation and market timing as nothing more than wild guesses as to what markets will do, and which are unreliable at best. But at the end of the day, what is MPT, MVO and Monte Carlo? Wild guesses based on precise mathematical calculations out to four insignificant decimal places.
Ed is actually critiquing something much broader than retirement calculators. He is criticizing the false sense of precision that sometimes emerges from quantitative finance. Warren Buffett also has a disdain for complex math, and this article collects his quotes on the topic. It also reminds me of Models Behaving Badly, which is subtitled "Why Confusing Illusion with Reality Can Lead to Disaster, on Wall Street and in Life." This book explains why mathematical models can never truly capture human behavior, and why a heavy reliance on models can be so dangerous.
The Portfolio Manager as Fiduciary
More fundamentally, Ed is raising philosophical issues about the fiduciary role of a portfolio manager. As fiduciaries, it is our responsibility to preserve client wealth and purchasing power in the face of uncertainty. We can't escape our responsibility by using retirement calculators as a substitute for sound judgment. We can't pass the buck and tell our clients: "Sorry, the model failed and your portfolio crashed. Your retirement will feature cat food, not caviar."
Links to 21 Retirement Calculators
With that caveat in mind, I asked Alan Johnson for his opinion. Alan is the founder of Johnson Harper, LLC, a registered investment advisor in NJ, and a contributor to Seeking Alpha.
Alan has spent a lot of time comparing different tools on the web, and his handy reference article compares 21 different retirement calculators. The post has links to each retirement tool, and he notes whether each is a calculator or a simulator (which allows the user to adjust historic returns and volatility).
Closing Thoughts About Use vs. Misuse
Alan and I tend to think alike about retirement calculators. Calculators are necessary tools, because you have to make some kind of estimate to create a financial plan for clients. These estimates may spur people to save more, to think about long-term goals, frame, and to focus on the aspects of retirement that they can control (such as their retirement age and spending habits. So in this sense, retirement calculators have a useful role.
But these financial estimates of the distant future are very sensitive to the inputs, so the outputs should be considered guesstimates. Consequently, I do not like the commercial from ING that has people walking around with their "number". To me, this implies that investors can achieve financial peace of mind by finding a number that precisely quantifies their retirement needs. This is a misuse of retirement calculators.
Then again, if an investor genuinely expects to achieve security in life through a number, they probably need a reminder that retirement planning is merely a means to an end. As Philip Fisher said:
"The stock market is filled with individuals who know the price of everything and the value of nothing."
Disclosure: I am long IAU, SCHH, AMJ, VQT. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: Long: ARBFX, KHYZXAll written content is for information purposes only. Opinions expressed herein are solely those of Right Blend Investing and our editorial staff. Material presented is believed to be from reliable sources, however, we make no representations as to its accuracy or completeness. All information and ideas should be discussed in detail with your individual adviser prior to implementation. The presence of this article shall in no way be construed or interpreted as a solicitation to sell or offer to sell investment advisory services to any residents of any State other than the State of New Jersey or where otherwise legally permitted. This is not a complete discussion of the information needed to make a decision to open an account with Right Blend Investing, LLC. There are always risks in making investments, including the investment strategies described.