Advertisement

Retirement-specific behavioral finance and derivation of benchmark behavior for FP actions

Part of the Contributions to Economics book series (CE)

Abstract

This chapter represents the normative part of the study. It aims to derive benchmark behavior for FP actions based on an enhanced understanding of typical “normal”159 behavior. It builds on the fact that “people depart from rationality, but they do so in ways that can be predicted — and exploited.”160 To achieve this, findings related to “normal” behavior — as opposed to rational behavior — are introduced and illustrated based on three theories: the decision-making process, the behavioral life-cycle hypothesis and the behavioral portfolio theory.

Keywords

Loss Aversion Pension Plan Asset Allocation Ambiguity Aversion Hindsight Bias 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 160.
    N.N. 2006a, p. 69.Google Scholar
  2. 162.
    Shleifer 2000, p. 12.Google Scholar
  3. 163.
    Scott/ Stein 2004, p. 218.Google Scholar
  4. 164.
    Here a distinction between cognitive and emotional biases is also important. Since “cognitive biases stem from faulty reasoning, better information and advice can often correct them. Conversely, because emotional biases originate from impulsive feelings or intuition — rather than conscious reasoning — they are difficult to correct.” Pompian/Longo 2005, p. 59.Google Scholar
  5. 166.
    In this model, the individual updates his probability assumptions according to new information following the Bayes’rule and thus converts à priori probabilities into à posteriori probabilities. This dates back to the mathematical theorem of Bayes. Furthermore, the utility function follows other maxims of decision theory such as e.g., steady and constant preferences. See Eisenführ/Weber 2003, p. 169ff.Google Scholar
  6. 168.
    This means that he prefers the certain prospect x to any risky prospect with expected value x. Correspondingly, when choosing in a gamble between a sure gain of $1 and a risky strategy with an expected value of $1, he prefers the sure gain. In that case, the sure $1 gain can be considered as security equivalent. The difference between the security equivalent and the expected value of the risky strategy is called the risk premium. For a risk averse individual to be independent between a sure and a risky strategy, the expected value of the risky strategy would need to be higher by a positive risk premium. If however, the individual were risk neutral, then no risk premium is required and the individual would be independent between the two options described above. Furthermore, if the individual were to show risk friendly behavior, he would choose a strategy whose expected return is lower than the security equivalent of that decision. Thus, the risk premium is negative in that case. See Eisenführ/Weber 2003, p. 222ff.Google Scholar
  7. 169.
    Mitchell/ Utkus 2006, p. 90.Google Scholar
  8. 170.
    Kahneman/ Riepe 1998, p. 56.Google Scholar
  9. 171.
    See Shleifer 2000, p. 113 or Mitchell/Utkus 2006, p. 90.Google Scholar
  10. 172.
    Kahneman/ Riepe 1998, p. 55.Google Scholar
  11. 173.
    Shefrin 2002, p. 4.Google Scholar
  12. 174.
    Kahneman/ Riepe 1998, p. 56.Google Scholar
  13. 175.
    An interesting point can be noted with regard to this tendency to estimate future performance based on past performance of a stock: Performance reports are mostly accompanied by the disclaimer “Past performance is no guarantee of future returns.” But since the size of that disclaimer is very small compared to the illustration of past performance, the availability bias limits the impact of this disclaimer. See Mitchell/Utkus 2004a, p. 21.Google Scholar
  14. 176.
    Edwards 1968, p. 359.Google Scholar
  15. 177.
    When the movie “Jaws” was shown, the number of people swimming in open waters dropped substantially, despite the fact that the probability of encountering a shark had not changed on an objective level. Also see Harrison 1999, who analyzes the enduring fear caused by “Jaws” and other movies and finds that exposure to “frightening media” leads people to continuously avoid the situations portrayed.Google Scholar
  16. 178.
    See Shleifer 2000, p. 127f.Google Scholar
  17. 179.
    Kahneman/ Riepe 1998 find that 80% of individuals think that their driving skills are above the average.Google Scholar
  18. 180.
    See Kahneman et al. 1982.Google Scholar
  19. 181.
    Barber/ Odean 2001, p. 289.Google Scholar
  20. 182.
    Shefrin 2002, p. 48.Google Scholar
  21. 183.
    Yates 1990, p. 204.Google Scholar
  22. 184.
    See Mitchell/Utkus 2004a, p. 24.Google Scholar
  23. 185.
    See Dehm et al. 2000, p. 24f and Höllger/Sobull 2001, p. 16f.Google Scholar
  24. 186.
    Yates 1990, p. 100.Google Scholar
  25. 187.
    Kahneman/ Riepe 1998, p. 55.Google Scholar
  26. 188.
    “The editing phase consists of a preliminary analysis of the offered prospects, which often yields a simpler representation of these prospects.” Kahneman/Tversky 1979, p. 274. An example of this editing phase is the segregation of secure gains: In a gamble an individual either gets $30 with a probability of 30% or $70 with a probability of 70%. Typically, individuals frame these two options differently in that they consider the $30 as a sure gain and formulate the remaining cash flows as a gamble with a 30% probability of getting zero and a 70% probability of wining $40.Google Scholar
  27. 189.
    Mitchell/ Utkus 2006, p. 85. This point is further detailed below in the section on the lack of firm preferences.Google Scholar
  28. 190.
    Mitchell/ Utkus 2004a, p. 16.Google Scholar
  29. 191.
    See Mitchell/Utkus 2006, p. 87.Google Scholar
  30. 193.
    “Once households get used to a particular level of disposable income, they tend to view reductions in that level as a loss. Thus households may be reluctant to increase their contributions to the savings plan because they do not want to experience this cut in take-home pay.” Thaler/Benartzi 2004, p. 169.Google Scholar
  31. 194.
    The standard economic theory supposes that the rate of commodity substitution at a point on an indifference curve is the same for movements in either direction, meaning that the trade-off between A for B and B for A is identical over the same interval. However, empirical findings have shown that people make choices that differ depending on the direction of the proposed trade. See Knetsch 1989, p. 1277.Google Scholar
  32. 195.
    Knetsch 1989, p. 1277.Google Scholar
  33. 196.
    See Loewenstein et al. 2003, p. 1214.Google Scholar
  34. 197.
    While the average stock return from 1926 to 1995 was 7%, the return on treasury bonds amounted to less than one percent. See Benartzi/Thaler 1995, p. 73.Google Scholar
  35. 198.
    See Tversky 1990, p. 74.Google Scholar
  36. 199.
    Shiv et al. 2005 examine the effect of emotions on decision making and find that individuals with emotions (compared to those who were unemotional due to accident or sickness) showed sub-optimal behavior. They were too optimistic after a winning experience and too pessimistic in case of losing.Google Scholar
  37. 200.
    See Simon 1955, p. 99ff.Google Scholar
  38. 201.
    Thaler/ Benartzi 2004, p. 167.Google Scholar
  39. 202.
    Only 80% of all interviewees dared to make an estimate and of these, 45% thought the contribution rate was lower than 19.5% and 13% expected it to be higher. See Boeri et al. 2001, p. 23.Google Scholar
  40. 203.
    Reilly/ Brown 2000, p. 53.Google Scholar
  41. 204.
    Mitchell/ Utkus 2004a, p. 35.Google Scholar
  42. 205.
    See Thaler 1994, p. 187.Google Scholar
  43. 206.
    Practitioners for example have developed the rule of thumb to save 10% of the disposable income. See Kobliner 2000, p. 42. However, as the simulation of the required saving rate for the different groups of investors will show in subchapter 6.4, the ideal saving rates deviate substantially from this rule of thumb indication and are strongly influenced by the individual situation and preferences. Indeed, organizations specialized on individual financial advice such as UBS, Credit Suisse or Wegelin emphasize that it is impossible to set up generally applicable rules. See Fischer 2006, Gast 2005 and Orgland 2006.Google Scholar
  44. 207.
    See for example Lusardi 2004, Clark et al. 2004 and the OECD 2005.Google Scholar
  45. 208.
    Choi et al. 2002, p. 70.Google Scholar
  46. 209.
    Mullainathan/ Thaler 2000, p. 9.Google Scholar
  47. 210.
    Thaler/ Benartzi 2004, p. 167.Google Scholar
  48. 211.
    Angeletos et al. 2001, p. 48.Google Scholar
  49. 213.
    Angeletos et al. 2001, p. 52.Google Scholar
  50. 214.
    O’Donoghue/ Rabin 1999a, p. 118.Google Scholar
  51. 215.
    Angeletos et al. 2001, p. 52.Google Scholar
  52. 217.
    Angeletos et al. 2001, p. 48.Google Scholar
  53. 218.
    Loewenstein et al. 2003, p. 1209.Google Scholar
  54. 219.
    See Laibson 1997, p. 446.Google Scholar
  55. 221.
    See Warner/Pleeter 2001, p. 34f.Google Scholar
  56. 222.
    See Bodie et al. 2002.Google Scholar
  57. 223.
    Thaler/ Benartzi 2004, p. 169.Google Scholar
  58. 224.
    Shefrin 2002, p. 32. As an example the financial situation of two individuals who graduated from the same university one year apart and upon graduation took similar jobs is highlighted. While individual A encounters no inflation in the first year and gets a salary increase of 2% at the end of the year, individual B lives in an environment with 4% inflation and gets a salary increase of 5%. Correspondingly, individual A is better off in real terms. However, in most people’s mind, individual B tends to be happier while individual A will probably start to look for another job.Google Scholar
  59. 225.
    See Börsch-Supan et al. 2005, p. 13ff.Google Scholar
  60. 226.
    Loewenstein et al. 2003, p. 1211.Google Scholar
  61. 227.
    Sometimes, this tendency to the mean can also have rational reasons. Scharfstein/Stein 1990 have found that institutional portfolio managers sometimes ignore substantive private information and mimic the investment behavior of other asset managers. Even if this behavior looks inefficient, it can be rational from a social standpoint if the managers are concerned about their reputation in the market. This is due to the fact that a bad decision is not as bad for the reputation when others make the same mistake. In that sense herding and the acceptance of the mean strategy is even a rational attempt of managers to enhance their reputation. See Scharfstein/Stein 1990, p. 465ff.Google Scholar
  62. 228.
    Thaler et al. 1997, p. 648.Google Scholar
  63. 229.
    Thaler 1994, p. 188.Google Scholar
  64. 230.
    See Thaler 1995, p. 10ff.Google Scholar
  65. 231.
    Shefrin/ Statman 1993, p. 131.Google Scholar
  66. 232.
    See Shefrin 2002, p. 144.Google Scholar
  67. 233.
    On this basis, Shefrin 2002 proposes to motivate such a transaction by thinking of it as a reallocation of assets from one mental account to an other account (“Transfer your assets!”) rather than closing one account at a realized loss. Shefrin 2002, p. 27. The effect of why selling a loser is so painful has already been illustrated in the context of the prospect theory and the loss aversion highlighted above.Google Scholar
  68. 234.
    See Thaler et al. 1997, p. 648f or Kahneman/Riepe 1998, p. 62.Google Scholar
  69. 235.
    They conduct a test between two groups of employees covered by a DB plan. One group is shown the annual rates of return for two asset types; the other group is shown a distribution of simulated 30-year rates of return. They find that the latter group invests substantially more of their retirement funds in stocks than the first group (90% versus 40%). See Thaler et al. 1997, p. 659.Google Scholar
  70. 236.
    Benartzi/ Thaler 1995, p. 75.Google Scholar
  71. 238.
    See Eisenführ/Weber 2003, p. 361ff.Google Scholar
  72. 239.
    For example, in a drawing game of 90 red, black and yellow balls where 30 are red and 60 are black or yellow, people tend to prefer the game where they win if a red ball is drawn over the game where a black ball means victory. See Eisenführ/Weber 2003, p. 361.Google Scholar
  73. 240.
    Sethi-Jyengar et al. 2004, p. 86.Google Scholar
  74. 241.
    Benartzi/ Thaler 2002, p. 1594.Google Scholar
  75. 242.
    See Zweig 1998, p. 114f.Google Scholar
  76. 243.
    See Yates 1990, p. 198.Google Scholar
  77. 244.
    Thaler/ Benartzi 2004, p. 167.Google Scholar
  78. 245.
    Thaler/ Benartzi 2004, p. 168.Google Scholar
  79. 246.
    Samuelson/ Zeckhauser 1988, p. 7.Google Scholar
  80. 247.
    See Samuelson/Zeckhauser 1988, p. 33.Google Scholar
  81. 248.
    O’Donoghue/ Rabin 1999b, p. 131.Google Scholar
  82. 249.
    Madrian/ Shea 2001, p. 1177.Google Scholar
  83. 250.
    Choi et al. 2002, p. 70.Google Scholar
  84. 251.
    See Madrian/Shea 2001, p. 1184.Google Scholar
  85. 252.
    See Choi et al. 2002, p. 68.Google Scholar
  86. 253.
    Mullainathan/ Thaler 2000, p. 4.Google Scholar
  87. 254.
    See Mitchell/Utkus 2004a, p. 6.Google Scholar
  88. 255.
    Thaler/ Benartzi 2004, p. 167.Google Scholar
  89. 256.
    See Dehm et al. 2000, p. 1.Google Scholar
  90. 257.
    Choi et al. 2002, p. 70.Google Scholar
  91. 258.
    See Choi et al. 2002, p. 103.Google Scholar
  92. 259.
    Shefrin/ Thaler 1988, p. 622.Google Scholar
  93. 260.
    See Boeri et al. 2002, p. 397.Google Scholar
  94. 261.
    See Shefrin 2002, p. 150ff. In light of the above discussed inertia, dollar cost averaging does seem recommendable in connection with so-called life-cycle funds or target funds that automatically adjust the asset allocation to reduce the level of equity over time. See for example Fidelity’s target 2010 Euro fund (Fidelity 2006a) that will have only cash holdings in 2010 or the Swisscanto Life-cycle fund 2020 (Swisscanto 2006) with a similar investment strategy focusing on the year 2020 as payback time.Google Scholar
  95. 262.
    Mitchell/ Utkus 2004a, p. 8.Google Scholar
  96. 263.
    Shefrin 2002, p. 142.Google Scholar
  97. 264.
    See Mitchell/Utkus 2006, p. 88.Google Scholar
  98. 265.
    See Benartzi 2001, p. 1748ff.Google Scholar
  99. 266.
    French/ Poterba 1991, p. 223.Google Scholar
  100. 267.
    Shefrin 2002, p. 30.Google Scholar
  101. 268.
    See Kahneman/Riepe 1998, p. 63.Google Scholar
  102. 269.
    See Zweig 1998, p. 118.Google Scholar
  103. 270.
    See Shefrin 2002, p. 130ff.Google Scholar
  104. 271.
    Shefrin/ Thaler 1988, p. 621.Google Scholar
  105. 272.
    Shefrin/ Thaler 1988, p. 633.Google Scholar
  106. 273.
    Shefrin 2002, p. 122.Google Scholar
  107. 274.
    For example, consider also the distinction between lunch money and smart money brought forward by Spremann 1999, p. 22.Google Scholar
  108. 275.
    The interviews with Orgland 2006, Gast 2005 and Benenni 2006 confirmed these findings and point out the high level of individuality of each situation that makes it necessary to tailor the recommendations to the individual.Google Scholar
  109. 278.
    Weber 2004, p. 64.Google Scholar
  110. 279.
    See for example Thaler/Benartzi 2004: “The costs of actively joining the plan (typically filling out a short form) are trivial compared with the potential benefits of the tax-free accumulation of wealth, and in some cases a ‘match’ [that] is provided by the employer, in which the employer typically contributes 50 cents to the plan for every dollar the employee contributes, up to some maximum.” Thaler/Benartzi 2004, p. 169.Google Scholar
  111. 280.
    See OECD 2001, p. 27 for a table illustrating that the people living e.g., in the US, Canada, Germany, Japan or Sweden and aged 65 to 70 had between 80–100% of the income of individuals aged 18 and over.Google Scholar
  112. 282.
    According to Klöckner 1995 building-up a wealth position without taking inflation into consideration and without investment in inflation protected assets is more destruction of wealth and wealth illusion than creation of wealth. See Klöckner 1995, p. 49. According to Bodie et al. 2002 individuals are rather poorly informed about volatility in asset returns and inflation rates. This finding is based on historical circumstances. The nearly two decade-long equity bull markets that started in 1982 coincided with a period of tame inflation. Consequently the average retiree mistakenly assumed that equities provide an effective inflation hedge, leading to poorly constructed portfolios. See Bodie et al. 2002, p. 7.Google Scholar
  113. 283.
    See Börsch-Supan et al. 2005, p. 16ff.Google Scholar
  114. 284.
    Panis 2004, p. 270.Google Scholar
  115. 285.
    See Panis 2004, p. 267.Google Scholar
  116. 286.
    Hurd/ Panis 2006, p. 2226.Google Scholar
  117. 287.
    See Mitchell/Utkus 2006, p. 91 and Panis 2004, p. 270f.Google Scholar
  118. 288.
    See Choi et al. 2002, p. 82ff.Google Scholar
  119. 289.
    In MPT, Markowitz 1952 characterizes capital markets by distinguishing between risky assets such as securities and a risk free asset such as a AAA treasury bond. He defines them both in terms of their expected mean return and their volatility and thus derives the term mean-variance portfolio. The MPT calculates the efficient frontier, which explicitly takes into consideration the covariance between individual titles and represents combinations of securities that offer the highest return for a given level of risk. With regard to the individual investors, MPT makes two important assumptions: First, it follows standard economic theory in assuming that individuals are always risk averse and thus demand a higher compensation for accepting a higher level of risk. Second, MPT states that rational individuals choose a combination of securities that optimize their risk-return profile based on their utility function represented by their indifference curve. Flat indifference curves indicate that an investor has a high tolerance for risk and thus a less pronounced risk aversion, while very steep indifference curves are typical for highly risk-averse investors. The expected utility that an investor can expect from a portfolio is the risk-adjusted expected return, because the return is reduced by a risk penalty, consisting of the standard deviation of the portfolio, divided by the risk tolerance of the investor. This risk penalty increases, the higher the volatility of the investment and the smaller an investor’s risk tolerance is. For investors who are extremely risk averse, the risk tolerance is small and the risk penalty thus very large, leading to a low utility value. See Sharpe 1990, p. 7–9ff and Reilly/Brown 2000, p. 279.Google Scholar
  120. 290.
    Sharpe 1990, p. 7–3.Google Scholar
  121. 291.
    Brinson et al. 1986 and Brinson et al. 1991 conducted several studies where they examined the performance of pension plans and analyzed the extent to which the variation in return could be explained by the investment policy and the long-term asset allocation. In their 1991 study they conclude that 91.5% of the variance of actual returns of 82 pension plans could be explained by the returns of the tracking portfolio. See Brinson et al. 1991, p. 45.Google Scholar
  122. 292.
    As mentioned in Grünbichler 2004a, p. 67.Google Scholar
  123. 293.
    Polkovnichenko 2005 studied the holding of stocks among US households and found that the median family owns only two different titles whereas even the most affluent households only hold a median of 15 different stocks. With regard to diversification over international holdings, the evidence is also daunting: French/Poterba 1991 describe the domestic ownership of corporate equity and find that it amounts to 92% in the US, 96% in Japan and 92% in the UK. See French/Poterba 1991, p. 222.Google Scholar
  124. 295.
    Typically, these investors can be considered as being in their accumulation phase where the net worth is typically small relative to the liabilities (e.g., house mortgage). Assets in this phase tend to be non-diversified, with home equity representing the largest share. See Kaiser 1990, p. 3–18.Google Scholar
  125. 296.
    Reilly/ Brown 2000, p. 38.Google Scholar
  126. 297.
    Kaiser 1990, p. 3–19.Google Scholar
  127. 298.
    Bodie et al. 1992, p. 446.Google Scholar
  128. 299.
    See Spremann 1995, p. 123ff.Google Scholar
  129. 300.
    See Waggle/Englis 2000, p. 81 or Mitchell/Utkus 2006, p. 86.Google Scholar
  130. 302.
    Shim/ Siegel 1991, p. 238.Google Scholar
  131. 304.
    R. Shiller for example prefers inflation-indexed bonds because they “correspond to the ‘risk less asset’ in finance theory better than anything we have ever had, so people should use them to adjust the risk level of their portfolios.” Zweig 1998, p. 118.Google Scholar

Copyright information

© Physica-Verlag Heidelberg 2008

Personalised recommendations