The Neglected Variable Affecting Portfolio Choices in the 21st Century

Allianz Global Investors Center for Behavioral Finance onRisks The neglected variable affecting portfolio February 2016 choices in the 21st century Author In the digital age, people have far more access than ever before to their investment Shlomo Benartzi, Ph.D. accounts, allowing them to monitor the latest market changes with great ease. In this Professor, UCLA paper, we discuss how this variable—“monitoring frequency”— affects judgment Anderson School and portfolio choices. We argue that the ease of portfolio monitoring on smartphones of Management Chief Behavioral is likely to increase a behavioral tendency known as “myopic loss aversion,” where Economist, investors focus too much on short-term losses. We then ask an important question: Allianz Global Should financial advisors and plan sponsors incorporate the frequency of market Investors Center for monitoring into their portfolio recommendations? We conclude with suggestions Behavioral Finance for ways to use digital design to minimize myopic loss aversion. The current state we will look at how financial institutions can design of investment monitoring investment products that fit the monitoring We live in an age obsessed with investment patterns of their clients. In addition, we examine performance. Plan sponsors hire and fire fund how financial institutions can improve how they managers based on the returns they achieved in communicate investment performance to clients. the last couple of years. Individuals, meanwhile, keep close track of how much money they make or Let’s begin with the survey data. In October 2015, lose in the market. This white paper will investigate we commissioned a survey of 1,050 adults with a largely neglected variable of all this monitoring, retirement savings accounts and asked them which is how frequently people check their how, and how often, they checked their portfolios. investments. We will argue that, especially in the Not surprisingly, we found a wide variation in the digital age, the frequency of monitoring can have method and frequency of monitoring. As shown a large impact, influencing both how people feel in Figure 1 (see next page), while 15% of subjects about the market and their ensuing behavior. As a check once a year or less, roughly 20% check at result, it’s crucial that financial institutions monitor least once a week. The most common frequency the monitoring done by their clients. After is quarterly, with 34.5% of people getting updates reviewing new survey data on investment tracking, every few months. befi.allianzgi.com

The neglected variable affecting portfolio choices in the 21st century Figure 1: Portfolio monitoring frequency comes out sharply in favor of 34.5% acceptance.3 There is also little “risk” 29.7% in this bet for many people, as the maximum loss is no more than the cost of a nice dinner. Nevertheless, 16.4% the prevalence of loss aversion is 10.2% rooted in a simple psychological fact: Losses tend to hurt more—often 4.3% 2.9% 2.1% much more—than gains feel good. As a result, the pain of potentially Daily Weekly Monthly Quarterly Annually Every e ever losing $100 exceeds the pleasure of ear Source: Allianz Global Investors Center for Behavioral Finance survey, October 2015. gaining $200. And so people reject the bet on the flip of a coin, even Furthermore, it’s clear that the digital world is though they should probably say “yes.” reshaping the ways in which people monitor their We’ve already discussed Samuelson’s research and accounts. According to our survey, nearly 80% of loss aversion in the first OnRisks paper. However— people check their accounts on a computer, 30.6% and this is a big caveat—what happens if the bet check on a smartphone and 21.9% check on a tablet. can be played multiple times? While Samuelson’s By comparison, slightly over half review their paper gamble was a one-off affair—the coin was only statements and only 12.6% get updates in person flipped a single time—real-world evaluations of 1 from their financial advisors. The key question, risky choices are often replayed over many rounds. of course, is how these technological shifts might In this context, we argue that the appeal of the influence investor behavior. gamble depends in large part on how frequently 4 a person gets updates on the results. The digital world is reshaping how people monitor their accounts To watch this process at work, let’s play Samuelson’s bet twice. As you can see in Figure 2, there are four Does monitoring impact risk-taking? Figure 2: Possible hedonic outcomes when playing To understand why technology might affect Samuelson’s bet twice investor behavior, it’s useful to revisit a classic study by Paul Samuelson.2 More than half a century ago, First Second Hedonic Round Round Experience Samuelson proposed the following bet on the flip of +200 +200 a coin to one his MIT colleagues: If the coin landed on heads, they would win $200; however, if it landed +200 -100 on tails, they would lose $100. -100 +200 Samuelson’s colleague rejected the offer— as do most people—to take the bet, exhibiting -100 -100 a phenomenon known as loss aversion. However, given the 50-50 odds, the expected value of the bet Source: “Myopic Loss Aversion and the Equity Premium Puzzle” by Shlomo Benartzi and Richard H. Thaler, 1995.

The neglected variable affecting portfolio choices in the 21st century possible outcomes on the two coin flips: A player Figure 3: Possible hedonic outcomes if investors do can either win-win, win-lose, lose-win, or lose-lose. not check after each round Unfortunately, given the impact of loss aversion, a player who checks the results after each round First Second Hedonic Round Round Total Experience would only experience a positive hedonic ? ? +400 experience in one of the four scenarios. The reason is that in the win-lose and lose-win cases, our ? ? +100 loss-averse player focuses on the painful loss, even though the aggregate outcome is a net ? ? +100 gain of $100. ? ? -200 However, there is a very easy way to improve the hedonic experience of Samuelson’s gamble: Source: Benartzi and Thaler, 1995. Don’t check after each round. Instead, the gambler outcome. (To use academic terms, investors reset should wait until both coin flips are complete and their “reference points” too frequently.) only get an update on the aggregate outcome. As shown in Figure 3, by checking less frequently, Daniel Kahneman, the Nobel Prize-winning the player can triple the odds of having a positive psychologist, has a pithy description of human experience. The win-lose and lose-win cases are behavior that helps explain situations like this: What 5 no longer painful, as our player only observes the you see is all there is (WYSIATI). In the context of aggregate outcome, which is a net gain of $100. myopic loss aversion, WYSIATI can explain, at least in part, why people can be so influenced by hourly, The appeal of Samuelson’s gamble depends daily or weekly losses, even if their retirements are largely on how frequently a person gets updates decades away. Because losses are what they see, on results and losses loom larger than gains, losses assume a disproportionate weight in financial decision- The same lesson applies to the world of investing. making. Instead of focusing on their ultimate goals, When people get frequent updates on their people continually recount their gains and losses, portfolios, they increase the likelihood of getting and so they seek to avoid the stock market. news of a short-term loss—just like that gambler checking after every coin flip. However, because Because the loss is what people see, these losses have an exaggerated impact on our it assumes a disproportionate weight behavior, frequent checking can lead us to in financial decision-making abandon our long-term investing plans. This phenomenon is known as “myopic loss aversion,” The frequent recounting of gains and losses and it results from a mismatch between the time could have a dramatic impact on financial markets. horizon of an investment and the frequency of In our original paper on myopic loss aversion, our account evaluations. In general, the greater my colleague Richard Thaler and I explored the the mismatch, the more likely we are to overreact relationship between monitoring frequency and 6 to short-term losses, counting our recent gains the equity risk premium. As shown in Figure 4, if and losses instead of staying focused on the final investors were able to focus on 20-year returns,

The neglected variable affecting portfolio choices in the 21st century Figure 4: Implied Equity Premium for Different Evaluation Periods or annual basis. As a result, we 6.5% were relatively insulated from ephemeral market swings. The 4.7% digital age, however, has stripped away this insulation—investors are 3.0% hyperaware of every market bump, 2.0% swing and correction. 1.4% To understand the potential impact of 1 Year 2 Years 5 Years 10 Years 20 Years these new technological trends, let’s Source: Benartzi and Thaler, 1995, based on the real returns of stocks (using the CRSP return to Samuelson’s coin flip. In indices) and the real returns of five-year bonds for the 1926–1990 period. today’s digital age, we aren’t just they would expect equities to provide a risk getting updates on our investment “gambles” once premium of just 1.4% per year. However, as they a year—we might be getting updates 5,040 times monitor performance more frequently, the risk over the course of the account, assuming 20 years premium they demand goes up significantly. Those of checking the stock market every day. Or we might who focus on five-year returns expect a premium end up learning about the results of our market of 3% and those who recount their gains and losses “coin flip” a few hundred thousand times, which is annually expect a premium of 6.5%. 20 years of checking our accounts on mobile One way to think of it is that long-term investors phones several times a day. (The average American who can resist monitoring the stock market are checks their smartphone more than 150 times per 7 engaging in a “mental arbitrage.” They would have day; it seems reasonable to assume that a few of accepted a risk premium of just 1.4%, but because these glances might involve financial updates.) many investors are tempted to count their gains This suggests that technology might be altering the and losses frequently, they end up enjoying a much ways in which we perceive financial risk, as more higher risk premium. frequent updates lead to a parallel increase in the The impact of technology possibility of a short-term loss. And since “what we In the 21st century, the gap between our investment see is all there is,” these losses can lead us to seek horizons and our portfolio evaluations is likely out the safest investments. As Uri Gneezy and Jan growing; myopic loss aversion might be getting Potters have demonstrated, there is a direct and worse. This is for an obvious reason: We have far causal relationship between the frequency with more access to our financial information and are which investors evaluate their returns and their able to review our accounts—using all sorts of willingness to accept risk, with greater frequency leading to greater sensitivity to losses.8 gadgets and devices—at nearly any time. This represents a huge shift from only a decade Technology might be altering how we perceive or two ago, when most of our account updates financial risk arrived via paper statements issued on a quarterly

The neglected variable affecting portfolio choices in the 21st century But this increased access to financial accounts also ◽ his or her superficial monitoring frequency (often raises a new question, which is whether or not all an extremely short amount of time). financial evaluations are created equal. Put another It’s important to note that, while we can track how way, does every glance at our retirement accounts often people monitor the market, and we can lead us to count our gains and losses and reset our estimate their ideal investment horizons, it’s reference points? Or does the increased frequency extremely difficult to know when people mentally of market updates also lead to a dilution of their book gains and losses and reset their reference psychological impacts? points. Despite this uncertainty, it seems highly One hypothesis is that, in the 21st century, we likely that there is a correlation between the clocks, might “see” market performance but not fully as investors who get frequent market updates— process its details. In a sense, we habituate to the often using mobile devices—are more likely to constant updates, which means the bad news count their losses and reset their reference points, comes with reduced emotional impact. Let’s say, which can lead to increased myopic loss aversion. for example, that we’re playing Samuelson’s Helping investors focus on the long term gamble a few thousand times and that we get Here are a few action items for those who are trying updates on every coin flip. It seems unlikely that to help plan participants and individual investors we’d mentally record the outcome of every single think longer term: gamble. Rather, people might engage in a more superficial interaction with the results, neglecting to 1. Monitor the monitoring fully reset after each update. This principle might Measure how often people check on their portfolios, also apply to investment accounts. While there is especially as new apps and digital displays are likely a correlation between the frequency of introduced, such as the Apple Watch. Consider the evaluations and the resetting of reference points, time spent on the site or number of clicks within this correlation is no longer one-to-one. It’s a far the app to differentiate between people who are more complex relationship, possibly shaped by mentally booking their gains and losses versus recent market trends. We might not mentally book those who are engaging in superficial account a single loss, but what happens if several losses glancing. These are the neglected variables occur in a row? affecting portfolio choices in the 21st century. Of course, technology is not affecting everyone’s In the 21st century, there are likely three distinct monitoring the same way, so it is important clocks that influence financial decision-making: to slice the data by demographic groups. ◽ a person’s long-term investment horizon 2. Reevaluate the glidepath (potentially several years or decades); A mismatch between the risk preferences of ◽ a reference-point period in which he or investors and their portfolio allocations poses she mentally books gains and losses (often greater risk in the 21st century than ever before, significantly shorter than our investment as short-term losses are more noticeable. We horizon); and believe it is a valuable exercise for fiduciaries and

The neglected variable affecting portfolio choices in the 21st century investment committees to debate whether the just as easy. Ideally, the increased engagement glidepath should be adjusted to reflect the triggered by constant account monitoring should frequency of monitoring. In particular, does more actually be leveraged to encourage improved frequent monitoring call for a more conservative financial planning, and not investing paralysis. allocation to avoid investors panicking and selling Summary when markets sneeze? We don’t necessarily have In the 21st century, technology has transformed the answers, but we feel these are important the ways in which investors get information about debates for investment committees to start their retirement accounts. The task for financial engaging in. advisors and plan sponsors is to ensure that the 3. Foster longer-term thinking newfound convenience of mobile access doesn’t One of the crucial tasks of financial advisors and lead to increased myopic loss aversion. Whenever plan sponsors is to help individuals focus on their possible, investors should be reminded of their long-term goals, even if they are monitoring their true investment horizon. To paraphrase Daniel investments every hour on a smartphone. While Kahneman, what they see at the moment is not these individuals who engage in a high frequency all there is. of monitoring might prefer an all-cash portfolio, effective communication can help foster longer- The task for financial advisors and plan sponsors term thinking. From a behavioral perspective, the is to ensure that the newfound convenience of most intuitive and important information to convey mobile access doesn’t lead to increased myopic to individuals is their projected retirement income, loss aversion so this information should be prominently displayed.9 Because this figure is relatively stable, it will smooth the emotional reaction to market volatility, but it will also discourage clients from seeking constant updates. (Sometimes, making 10 information boring is a virtue. ) 4. Make the responsible choice easier It has never been easier for investors to monitor market performance. We should make more responsible actions, such as raising savings rates,

The neglected variable affecting portfolio choices in the 21st century Endnotes 1. The percentages add up to more than 100%, as individuals can use multiple methods to monitor their investments. 2. Paul A. Samuelson. “Risk and Uncertainty: a Fallacy of Large Numbers.” Scientia 98.612 (1963): 108-113. 3. Benartzi, Shlomo, and Richard H. Thaler. “Myopic Loss Aversion and the Equity Premium Puzzle.” The Quarterly Journal of Economics (1995): 73-92. Benartzi, Shlomo, and Richard H. Thaler. “Risk aversion or myopia? Choices in repeated gambles and retirement investments.” Management Science 45.3 (1999): 364-381. 4. Benartzi, Shlomo, and Richard H. Thaler. “Myopic Loss Aversion and the Equity Premium Puzzle.” The Quarterly Journal of Economics (1995): 73-92. Benartzi, Shlomo, and Richard H. Thaler. “Risk aversion or myopia? Choices in repeated gambles and retirement investments.” Management Science 45.3 (1999): 364-381. 5. Kahneman, Daniel. Thinking Fast and Slow. Farrar, Straus and Giroux (2013): p. 85-88. 6. Benartzi, Shlomo, and Richard H. Thaler. “Myopic Loss Aversion and the Equity Premium Puzzle.” The Quarterly Journal of Economics (1995): 73-92. 7. Kleiner Perkins Caufield Byers. Internet Trends D11 Conference (2013). 8. Gneezy, Uri, and Jan Potters. “An experiment on risk taking and evaluation periods.” The Quarterly Journal of Economics (1997): 631-645. 9. While Congress is still debating the Lifetime Income Disclosure Act, several record-keepers have already started displaying projected retirement income on participant statements. 10. Looney, Clayton Arlen, and Andrew M. Hardin. “Decision support for retirement portfolio management: overcoming myopic loss aversion via technology design.” Management Science 55.10 (2009): 1688-1703.

…out the …llian† loal In‚estors ‡enter for ˆeha‚ioral ‰inance Žhe Œllianž ‘lobal „nvestors šenter for œehavioral Finance is committed to empowering clients to make better financial decisions by offering them actionable insights and practical tools. efiŠallian†giŠcom ©… ƒ. Œllianž ‘lobal „nvestors is an asset management arm of Œllianž ­E. Žhe šenter is sponsored by Œllianž ‘lobal „nvestors U.­. ››š, a registered investment adviser, and Œllianž ‘lobal „nvestors istributors ››š. Œ‘„… ƒ Š…£…ƒ£ƒ ° †££