Why Most Retirement Investment Plans Are Wrong—Part I


The current financial crisis and market panic demonstrate why most of the investment plans for those in or near retirement are wrong. There are better ways to manage retirement money, but you will not learn about them from conventional advisors and sources.

Most retirement investment plans generally use one of three general strategies. Each of these strategies is flawed and puts retirement goals in danger. In this posting we will look into these strategies and why they are inappropriate. You can find more details in my book, Invest Like a Fox…Not Like a Hedgehog.


Traditional Investing

One strategy is to shift the portfolio from equities to bonds as one nears retirement. The idea is that an older person cannot take much of the risk in the equities markets, so should be in stable, income-producing bonds. A standard formula is to subtract your age from 100 and let that be the non-bond allocation of your portfolio.

One problem with this approach is inflation. Because people are living longer, their income needs to increase over time to maintain the same standard of living. Otherwise, even modest inflation will significantly reduce the purchasing power of bond income by 20% over five years and by half over 24 years. Two percent inflation reduces purchasing power by 20% after 10 years. Another way to look at the problem is that after 10 years of 3% inflation, a retiree needs 130% of the first year’s income to maintain purchasing power.

With longevity increasing, inflation is a major consideration. The average 65 year old man who retires today has about a 20-year life expectancy—and that means half today’s retirees should plan to live longer than another 20 years.

To combat inflation, a portfolio needs a growth component that will produce adequate growth for the future while providing enough income to meet current expenses.

Other problems with the income portfolio have become very obvious since the summer of 2007.

Yields on bonds and income-oriented stocks can sink to low levels. The yields might not be enough to generate adequate income unless the portfolio is quite valuable. Yields on safe treasury bonds are at or near historic lows, and the financial panic caused a flight to safety that pushed short-term treasury yields below the inflation rate. The disinflation that began in1982 pushed treasury yields from generous double digit levels to today’s inadequate yields. I regularly hear from investors who purchased certificates of deposit or treasury bonds years ago. When those investments mature, the investor has to reinvest the principal and will receive much lower yields if reinvesting in the same vehicles.

You could venture out of treasuries and CDs into higher-yielding alternatives such as investment grade or high-yield corporate bonds, preferred stock, emerging market bonds, and other alternatives. Unfortunately, these investments carry additional risks, and those risks can be much higher than for treasuries, especially during a recession or financial panic. The risk can be as serious as the bond issuer’s going bankrupt and the bond’s becoming worthless. Lesser risks are that the bonds lose favor with investors and their values decline. If the investor buys bonds directly (and not through a mutual fund) and holds them to maturity, sinking market values are not a problem as long as the issuer makes the income payments and repays the principal at maturity. But market prices are a problem when investments are through funds or the investor does not want to be locked in to holding until maturity.

Many investors cannot tolerate these risks. The income for retirement investment strategy was developed when the average retirement lasted five years. Today, the average retirement lasts 20 years or longer, and a different strategy is needed.


Modern Portfolio Strategies

A second investment approach, which has been used by most investors in recent years, is to buy and hold a diversified portfolio of investments. This strategy is based on academic work known as the efficient market theory and capital asset pricing model. This strategy is easy to implement using today’s technology, and most portfolios developed by investment professionals use it.

Implementing the strategy is fairly simple. You determine the level of risk you are willing to take and the rate of return you need. The software develops an “efficient frontier” that shows portfolios that generate the highest return for a given level of risk or the lowest level of risk for a given return. Investors select the portfolio with the risk-return combination they want and are told that while returns will vary from year to year, over time they are likely to achieve their return goals while taking the level of risk acceptable to them.

This strategy has significant problems, especially as it is usually is executed.

The risk and return trade off for different investments often is determined using historic data. Unfortunately, the markets change over time. Historic returns and risks do not determine future returns. One would think at least sophisticated investors would have learned that after the collapse of Long-Term Capital Management in 1998 in the last great liquidity crisis. Even so, Alan Greenspan admitted in his recent testimony to Congress about the financial crisis that professional investors used flawed models to evaluate the risk in their portfolios. They used only 20 years of data, which did not show what would happen in a crisis.

Likewise, the portfolio is supposed to be diversified by holding investments that have low correlations with each other. When part of the portfolio is down, other parts of the portfolio should be up. That prevents wide swings in the portfolio’s value from year to year though individual markets will have volatile prices. But correlations between investments change over time. Sometimes the changes are permanent; other times the changes are temporary. One adage among some investors is that diversification works until you need it most.

We saw that adage at work in most major market declines, including the one in September and October 2008. Virtually all investments except treasury bills declined. Even gold, a usual haven in a crisis, suffered. The only benefit of diversification in the panic was that some assets declined less than others.

Another problem with the CAPM approach is it treats volatility as risk. To most investors, volatility is not risk. Risk is the probability investment goals will not be reached, such as running out of money during retirement.

An even more important problem is returns over shorter periods differ from long-term averages. When long-term returns are presented as a list of numbers, one can have the impression they are earned more or less steadily. Examining a long-term chart of how the average was developed or computing returns for shorter periods paints a different picture. Returns can lag the long-term average by significant amounts and for significant periods.

A foundation or other institutional investor might be able to invest for the long-term using 70 or more years of market data. Individual investors, however, are not able to invest with a 70-year time frame that ignores return patterns over shorter periods. To most investors, returns over the next 10 years are what matters. An investment plan can fail if returns for the first 10 years are significantly below the long-term averages.

There are extended periods when returns are below the long-term average. From 1998 to the present is one such period. If one retires early in such a period and begins taking withdrawals based on long-term returns, the retirement portfolio is likely to run out of money before a new bull market can restore the portfolio. The long-term average return is a myth. There are very few years, much less extended periods, when markets return the long-term average. With U.S. stock indexes, the return in any year usually is significantly above or below the average.

The advice under CAPM is to develop a portfolio that over the long term provides the trade off between risk and return the investor seeks. Then, hold that portfolio allocation. Most followers of CAPM recommend investing only through index funds and similar passive investments. Over time, as markets perform as they have in the past, the investor will achieve his goals.

We have seen that is not the case over periods less than the long term. Asset classes can change their correlations, and that increases the risk in the portfolio. More importantly, asset classes have long-term bull and bear markets. They can underperform their long-term average not only in the short-term but for extended periods. Stocks and commodities earn less than their historic averages for periods of 10 to 20 years. Such extended periods of underperformance can be devastating to investors who do not have a 70-year time frame.

During these periods, a return that matches or beats an index is not useful to an investor. The investor wants assets in the portfolio that have no correlation with the major indexes and whose returns are not tied to the return patterns of the indexes.

The efficient market theory and investment strategies built upon it do not serve investors well, because they are flawed in theory and as practiced. Here are some key questions that are not answered by modern investment theory:

Ÿ Why do long-term bull and bear markets occur? Under efficient market theory, they should not occur.

Ÿ Why are prices of investments more volatile than the underlying fundamentals?

Ÿ Why do stock prices change without a change in fundamentals or without a change in fundamentals of the same magnitude as the price change?

Ÿ Why does the equity risk premium exist? If markets are as efficient as in the theory, stocks should not generate higher long-term returns than other investments.

Because these questions are not answered, the strategy produces flawed portfolios.


Adventures in Timing

The third strategy in common use is to change the portfolio based on market or economic signals. As we will see in the next posting, changing the portfolio to reflect fundamental changes is a good strategy. But too many investors use the wrong standard to make changes.

Many investors seek one or a small number of data points for guidance on when to make portfolio changes. They conduct a great deal of research to find correlations between changes in data and subsequent changes in an investment, an exercise known as data-mining.

The problem with this method is that the value of an indicator declines over time. Once-reliable indicators fail to generate the same results. Indicators that once were considered reliable but failed in recent years include the dividend yield, price to book value, price-earnings ratio, and q ratio. Technical analysis also has a spotty record. Some investors use data external to the market, such as interest rates and federal budget deficits, but these also are not reliable.

Data-mining fails to produce useful investment tools for several reasons:

The exercise confuses random events with causation. My favorite example of this comes from David Leinweber who wrote a paper for First Quadrant, an investment management firm, in 1997. Leinweber took a CD-ROM of data from the United Nations that contained detailed economic data from most countries in the world. He ran the data through computer programs to find the best predictor of the S&P 500. He concluded that the data with the closest correlation to that index is butter production in Bangladesh. He titled his paper, Stupid Data Mining Tricks.

As protection against such results, a piece of data should be used to make investment decisions only if a reliable theory can explain why it works. Otherwise, the correlation between the data and the markets is a coincidence and not cause and effect.

Even when a correlation is found, it never is a 100% correlation. There are times when the cause and effect do not both occur. Before putting his capital at risk, the investor has to be confident this will not be one of the outlier times or that the potential loss will not be too severe.

There are too many variables affecting a market to be able to capture the meaningful ones in a manageable amount of data.

Markets are dynamic. Markets are human beings acting together. People learn, or at least they change their behavior over time. A condition that produced a certain reaction in the past might not produce the same reaction today and probably won’t five or 10 years from now.

Beginning and end points affect the results of data mining. There are many examples of researchers who found a correlation between a market and certain data in a particular time period. Other researchers who explored the relationship over different time periods either did not find the correlation or found a much weaker correlation.

Another problem with both CAPM and using data to make portfolio changes is that markets have fat tails. This is a statistical term. A normal distribution of results has thin tails. That means extreme results occur very rarely. In the investment markets extreme results, both good and bad, happen with much greater frequency than under a normal distribution.

Fat tails have several effects. One effect is that a few extremely positive periods can make the long-term average return high. In fact, only investors who were invested during those few brief periods earned positive returns. Negative fat tails can have such high negative returns that they wipe out not only recent positive returns but also part of the investor’s capital. An investor who enters an investment with negative fat tails at the wrong time could lose all or most of his capital and not gain the benefit of the positive fat tails.

A timing tool that has good long-term average results but that has fat tails might have achieved those results based on one or a few good periods while underachieving most of the time.

Turning points are obvious only in hindsight. It is easy to look at long-term data and identify the turning points. It is much harder to be in a fast-moving market in real time and identify today’s confluence of events as a turning point. In the bull market of the 1990s many investors exited the markets several years before the peak. Their historic indicators showed the turning point was reached long before it actually was. In the 2007-2008 decline many investors re-entered the markets early, believing a bottom had been reached. To use price-earnings ratio as an example, only after the fact can one determine whether the low price-earnings ratio for the cycle was 16, 13, or lower.

Another reason to be wary of data as market indicators is that once an indicator is well-known, it stops working. Since investors learn, the knowledge of a correlation between the markets and some other data tends to alter behavior.

Most investors think like hedgehogs, as characterized by poet Archilochus in 7th century B.C. He wrote, “The fox knows many things but the hedgehog knows one big thing.” Investors want to learn one big thing and hold on to that. That is the downfall of many investors. Markets are dynamic. They are dynamic because markets are people acting together and people change and learn over time. Markets also are dynamic because the legal, economic, and financial structures of the markets change. Investors have to recognize the changes and be able to adjust their strategies.

An investor who thinks like a hedgehog can be successful if his prime investment years coincide with a market period that matches his “one big thing” insight. For example, an index fund investor did quite well from 1982 through 2000 but less well since then. A treasury bond investor did well during the same period, earning high yields and capital gains. Today, a treasury bond investor earns low yields and might see the bonds lose value as rates rise in the coming years.

A long time successful investor needs to think like a fox. The fox, as characterized by Archilochus and later by philosopher Isaiah Berlin, pursues many ends and thinks and acts on many levels. The fox uses many experiences to inform his beliefs and does not try to fit all experiences into one all-compassing principle. Instead, the fox is eclectic and even inconsistent. The fox is wary of big, central principles and simple historical analogies. Most importantly, the fox adapts and adjusts with conditions.

In the next posting we will learn how an investor can think like a fox to improve investment returns and how to develop a retirement investment strategy that is more likely to achieve its goals and reduce risks.

Posted 10-24-2008 1:58 PM by Bob Carlson