The Expectations Clock provides a framework for modeling performance, expectations of future performance, and human behaviors during phases of the performance cycle. It is a model of human over- and under-reactions to the environment.
Success in investments (and business in general) is highly dependent on effectively forecasting future fundamentals (e.g., sales, earnings, etc.); however, these projections are influenced by human interpretation of events which are impacted by behavioral biases. Further complicating forecasts, human reactions (such as investments) to current and recent fundamentals also influence the fundamentals being forecasted. Thus, to predict performance, one must project the underlying cycle and how people interpret and respond to it.
Over the past four decades, substantial research has identified how people under-react and over-react to events. Under-reaction is associated with momentum, or short-term continuation of trends, and over-reaction is the basis of reversion and the efficacy of value investing.
.
.
The Expectations Clock, depicted in figure 1, illustrates the performance and expectations cycle. Performance may be great (12:00), very poor (6:00), improving (6:00 to 12:00), and deteriorating (12:00 to 6:00). Performance can be of anything – GDP growth, sales growth, ROE, P/E, P/B, prior returns, credit spreads, and other economic, political, and corporate events. Expectations of the future are highly dependent on today’s and the recent past performance; however, given that expectations determine one’s reactions to performance, expectations can also influence future performance.
The remainder of this paper is divided into two sections that provide evidence supporting the Expectations Clock. The first section reviews evidence of performance cycles, and the second shows how level and change in performance influence expectations and how expectations and performance interact to exacerbate the volatility of cycles. In both sections, I refer to phases and times on the Expectations Clock.
Performance Cycles
The Expectations Clock suggests that performance reverses. If performance is great and improving (phase III), future performance is likely to deteriorate but still be strong (phase IV). Likewise, if performance is poor and deteriorating (phase I), the next phase is probably below average but improving (phase II).
Figures 2 and 3 show evidence of reversion. In each year from the end of 1998 through 2007, companies in the S&P 500 were sorted on ROE into five quintiles and then tracked for the next nine years. As predicted by the Expectations Clock, firms with high ROE declined over time, and firms with low ROE rose (figure 2). The difference between the high (29.9%) and low (1.5%) ROE companies was 28.4% at the beginning of the nine-year period, but only 9.3% at the end of the period. Great performing companies deteriorate, and poor performing companies improve. Furthermore, as predicted by the Expectations Clock, the low ROE companies declined (phase I) before rising (phase II), and the high ROE companies improved (phase III) before declining (phase IV) (figure 3).
.
However, if you look closely, even after nine years, the low performing companies do not become high performing and the high performing companies do not deteriorate to bottom half performers. The high ROE companies still have a median ROE of 21.1% after nine years while the low ROE companies only improved to 11.8%.
.
This continuing advantage of high over low performing companies may be partly because the original (at time -9 years) sorting of ROE was across the entire market and not within sectors, and some industries and sectors have above average or below average ROE over the long-term. For instance, sectors with value-added products where risk of development is high, such as health care, normally have above-average ROE, whereas commodity sectors such as financials and utilities typically have low ROEs due to their mature businesses and regulated status. Although, over the last 20 years, all sectors experienced periods when ROE advanced and declined on a relative basis to the S&P 500 and all experienced periods when they were above and below average (except healthcare and consumer staples which have always had above average ROE). In addition, the analysis focused on large firms. These firms may be more established and old, and it may be easier to change the fortunes of a small company than a large firm.
.
While there may be a sector and size bias impacting the results, it probably does not fully explain why top performers tend to stay above average and bottom performers remain below average. Given this, it appears that overall level of performance is “sticky.” While change in performance is volatile, overall level of performance is not. Thus, the Expectations Clock can be modified as shown in figure 4.
.
.
Figures 2 and 3 indicate that performance reverses, or cycles, over time. These cycles occur due to gyrations in economic conditions and because of strong forces such as life cycles, the competitive landscape, and corporate manager errors.
.
Economic Cycles (figures 5-6)
.
A bad company may simply get better as the economy recovers from a recession (phase II). The tide lifts all ships, but those that are sinking may be lifted higher. Cyclical firms in sectors such as industrials, materials, and energy just need an improving economy to move from above-average to great, or from poor to below-average.
.
Cyclical firms are often capital-intensive, have significant fixed costs, have high priced products with elastic demand, and may be laden with debt. The auto industry, which like other cyclical sectors, is severely impacted by recessions. As sales deteriorate, cash flow and earnings may turn negative since many costs, such as depreciation, interest, and labor, remain relatively unchanged. Interest coverage ratios decline and paying interest on debt, let alone repayments of maturing debt, is a challenge. An economic recovery that brings a slight increase in sales leads to a huge sigh of relief. High operating and financial leverage magnifies the impact of a small sales increase and leads to a large rise in cash flow and earnings. Default on debt is avoided.
.
.
On the other hand, stable firms in the utilities, health care, and consumer staples sectors underperform when the economy recovers (phases II and III) and outperform during risk-off environments (phases I and IV) as investors gravitate to safety.
.
Forces of Change are Great
.
Besides economic cycles, performance may be influenced by corporate life cycles, competition, and even errors by management.
.
It is difficult for great companies (high performance, 12:00 on the Expectations Clock) to stay at the top of their game. The forces of change, or reversion, are high. A great company may falter as its products mature, competition enters its (highly profitable) business, or as external forces such as technology, government regulations, social changes, demographics, and foreign influences (e.g., subsidized businesses) impact results. A bad company may also be acquired (for a low cost) by another firm that believes it can fix its problems.
.
C
onsider the case of Apple from 1980 to today.[1] [2] It went from great, to lagging competitors, to rejuvenation, to near bankruptcy, to back to great (or fantastic), to perhaps, at least temporarily, peaking sales. Over this period, the firm had several periods of improving fundamentals (phases II and II) and deterioration (phases IV and I). It also experienced periods of greatness (12:00), near failure (6:00), and times in between.
.
Early Years: Apple was first with a PC and launched an IPO in 1980, but its system was closed while IBM had an open Microsoft/Intel system that was easy to clone. Steve Jobs was out a job by 1985. From 1985-1993, Skulley pushed the Mac to new markets (education and desktop publishing) and by 1990 the firm was very profitable and people loved their Macs. However, the Mac was premium priced and costs for computers continued to drop so IBM-compatible machines took market share. Skulley was out in 1993. The next four years were very volatile as the firm reversed several of its business strategies (alliance with IBM for OS, work on Mac OS to run on Intel chips, etc.) and slashed costs. The stock floundered, and investors who purchased the stock were considered crazy to expect a rebound (I was one of those investors as Apple moved to our #2 position in a $7 billion fund) as Apple racked up losses ($1.6 billion in 1997). Then, Steve Jobs was asked to return as CEO.
.
Middle Years: The firm stopped making Apple clones in 1997 and Microsoft invested to develop core products (Office). Increased innovation resulted in a $1,299 iMac in 1998 and non-Apple peripherals. They also slashed costs and made a profit in 1998. Then, the company stepped toward success by introducing a new operating system in 2001 (first since 1984) and an upgrade cycle every 12-18 months (faster than Microsoft). Apple then entered the consumer market with an iPod in 2001 that worked with the Mac, and iTunes, iPhoto, and Safari (2003). The Apple store was opened in 2001.
.
Later Years: The 2002 iPod syncs with Windows. In 2003, iTunes music was only $0.99 which led to market share gains. Moving iTunes to Windows increased market share 20 times and by 2010 Apple had 70% of the MP3 market. In 2006, the Mac ran faster on less power-hungry Intel chips that enabled laptops, and the Mac could run Windows. The 2007 iPhone, which was 2.5 years in development, became Time magazine’s “Invention of the Year.” The app store was introduced in 2008. Finally, in 2010 the firm introduce the iPad. The firm was a leader again.
.
Today: After many years of being dominant and creating new revolutionary products for consumers and businesses, Apple’s sales appear to have peaked in 2015 and are down over the last year. Today, the firm competes with a number of competitors with strong smart-phone products. Apple is a large firm, so it is difficult to grow off its base, especially since some product categories, such as PCs, have matured and may be declining. Could the firm once again reinvigorate growth with another great invention, expansion into China, or through other opportunities? Possibly. Or, this could be the start of its next phase downward? There has been plenty of controversy about growth for Apple over the last three years. The P/E bottomed below 10 in mid-year 2013 as growth slowed, before rising to about 20 in early 2015 as growth picked up, and then it fell again to about 10 in early-2016 as sales peaked. However, it has since risen to 17, or about 13% less than the market’s multiple.
.
Management Error
.
Capital expenditure decisions are some of the most important decisions management can make. Management uses of one of a firm’s most important resources – cash flow. Unfortunately, management tends to make these decisions to grow through expansion based on current and recent performance. This pro-cyclical orientation to capital spending makes the peaks of cycles higher and exacerbates the depths of recessions.
.
Figure 9 shows that fixed expenditures growth is highest when economic growth is highest, and vice versa. Wouldn’t it be best to expand (divest) when the cycle is at its lowest (highest) when the cost of expansion is cheapest (most expensive) and just before the cycle rebounds (deteriorates)? Of course, if management heeded this advice, then economic volatility may be muted or nonexistent. We discuss the biases that cause people to make errors in investment in the next section.
.
.
Figure 10 shows that the managers of the largest companies (firms in the S&P 500) are no better than the economy at large (figure 9). They expand after experiencing periods of high cash flow (EBITDA) growth, and reduce growth expenditures after periods of cash flow deterioration. They expand production at peaks, only to discover that expansion leads to over-capacity. They contract at the bottom of cycles, instead of growing by buying other companies (M&A activity declines during recessions) at low prices. You may claim that corporations cannot buy other firms at bottom of cycles since funding is not available. However, if they saved their cash flow at peaks, then cash and debt capacity would be available to buy assets cheaply at troughs. You may also assert that corporate managers cannot invest counter-cyclically since owners (investors) of public firms are short-term – expanding when times are weak is unconventional and CEOs may face job risk if they pursue this strategy (see my later discussion on rational irrationality). To this, I respond that they have an obligation as leaders to act responsibly even if it may cost them their jobs.
.
Expectation Cycles
.
The Expectations Clock indicates that expectations are related to performance.
.
Figures 9-10 showed that corporate managers do not accurately predict the future. If they did, then they would not invest at peaks and cut off investments at bottoms of cycles. They tend to believe that whatever trend, high and improving (phase III) or low and declining (phase I), will continue even if it tends to reverse. Corporate managers are not alone. Investor expectations of the future and actions which are reflected in the value of the stocks also vary coincident with – are not predictive of – the performance cycle.
.
Figures 11-14 reproduce figures 2-3 (on the left) and include valuation (P/B) on the right. Price to book measures investors’ expectations of future ROE, growth, risk, and payout. If ROE is expected to rise, growth is expected to improve, and risk is expected to fall, then P/B should rise, and vice versa.
.
- Equation 1 (the dividend growth model): Po = Eo * (1 + g) * pay / (r – g)
- Divide both sides of equation 1 by B
- Equation 2: Po / B0 = Eo / B0 * (1 + g) * pay / (r – g)
- Equation 3: Po / B0 = ROE0 * (1 + g) * pay / (r – g)
- Where P is price, B is book value of equity, g is growth, r is risk, and pay is payout.
.
Below, one can see that P/B is lowest when ROE is the lowest, and highest when ROE is the highest. However, ROE reverses. This implies that investors do not predict the change. They over-react (and drive P/B down) to whatever negative events drove ROE down for low ROE companies, and under-react to signs that good times may reverse for high ROE stocks which command the highest P/B ratios.
.
.
Investors also miss signs of reversal in above-average and below average rates of sales growth. Figure 15 shows that low ROE companies, at the beginning of the first period of ROE sorting (time -9), had below average sales growth. However, these firms had the highest sales growth over the next eight years. If investors expected this, then the P/B for these firms may not be so low. The opposite can be said for the high ROE companies that experienced deteriorating sales growth over time (below average by year -1).
.
.
Finally, missing signs of change results in missed opportunities to make money. Figure 16 shows that the median cumulative performance of the top quintile ROE stocks over the three years before the sorting was 77.6% versus 28.9% for the bottom quintile (for an advantage of 48.7%); however, by three years after the sorting the top ROE stocks underperform the bottom ROE stocks by 9.9% (23.0% versus 33.0%). The advantage of the low ROE stocks continues to grow over the eight years post the sorting as they rise 177.2% versus 102.8% for the high ROE firms (figure 17). The P/B of the initially low ROE companies remained low (it only rose modestly – see figures 11-14) during the entire period, which implies that investors never fully appreciated (or only slowly adjusted their expectations of) the ensuing above average sales growth and improving ROE. Hence, the cumulative performance of low ROE stocks rose more quickly than the high ROE stocks throughout the period under investigation.
.
Extrapolation, Inertia, and Duress
.
People tend to extrapolate current and recent changes in performance when forming their expectations of future performance.
.
People tend to have the lowest expectations of future performance when current performance has been deteriorating and is low (phase I). Here, inertia is low. Stress is high. People may be willing to “do anything” to change, and this attitude often results in over-reacting to negative events (e.g. selling stocks at the low, selling business units at the bottom, firing managers just before their strategies start to work, etc.) and under-reacting to positive signs of change. While the cycle often improves (phase II), it may improve further and more quickly if it were not for these over- and under-reactions.
.
On the other hand, people have the highest expectations of future performance when current performance is strong and has been improving (phase III). Inertia to the status quo is high. Stress is low. People believe “don’t rock the boat” or “if it ain’t broken, don’t fix it.” This attitude results in over-reacting to the current trend (investing more in winning stocks, investing more in capital when times are good (see figures 9-10), etc.), and under-reacting to forces of change (products maturing, competitors entering markets, etc.). Under-reacting to signs of weaknesses results in eventual deterioration (phase IV), and possibly more so than if one did not under- and over-react in phase III.
.
.
Causes of Expectations Blunders – Behavioral Biases
.
Human behavioral biases and short-cuts cause investors to miss clues of change and extrapolate expectations of the future from past changes in and current levels of performance. This exacerbates cycles. These biases cause stocks and corporate fundamentals to trend (the basis of momentum investing), but trends that move to extremes result in reversals over the long-term (the basis of value investing). Sir John Templeton said, “The four most expensive words in the English language are ‘this time is different’.”[3] He warns that trends cannot go on forever, which contrasts with typical economic reality where reversals arise as conditions change. A few biases that cause people to miss signs of change are discussed below.
.
- Overconfidence: Overconfidence is one of the most common and impactful biases. It causes people to miss signs of changes. These changes could result in a firm abandoning business strategies that are about to flounder, but a CEO may stay the course if he/she is overconfident in the current direction and cannot accept that he/she could be wrong. As a result, during good times (phase III), people under-react to problems, such as maturing products and intensifying competition.
.
- Confirmation Bias: People like to be right, so they often overlook and do not seek information on why they can be wrong. Second guessing one’s viewpoint, which normally cycles with performance, could open one’s eyes to signs of change; however, we tend to only look for evidence that supports our views. The more quickly one recognizes mistakes in phases I and IV, the sooner one may be able to take action to return to phases II and III. In addition, if one properly questions the status quo and is aware of negative forces of change in phase III, he/she may react appropriately and perhaps change strategy and avoid entering phases IV and I.
.
- Representativeness and Anchoring: People tend to extrapolate the past into projections of the future. If the past is poor and declining, then the future is assumed to be the same. They also take action based on these expectations which results in poor decisions (overinvesting at tops of cycles (figures 9 and 10) and selling stocks at bottoms of their cycles (low P/B for low ROE stocks as shown in figures 11-14)). The past and recent trends provide a strong anchor, so people are slow to alter their views. As a result, stocks continue to outperform/underperform for long periods. If investors reacted quickly and sufficiently to signs of changing business conditions, then much of the outperformance should appear in early years and cumulative underperformance of the high ROE versus low ROE companies would disappear in later years (figure 17 shows that cumulative underperformance grows each year).
.
- Herding: People like companionship. They like to agree with others, which is a natural survival instinct. Imagine animals on the African savanna. They run in packs to avoid being attacked by lions. The animal that strays from the pack is easy prey for the lion. Similarly, people do not like being left out of the crowd during good times, and do not want to be caught holding a poor company in bad times. As a result, it is difficult to purchase underperforming low ROE companies (phase I), and many people cannot help but purchase high ROE stocks (phase III).
.
- Loss Aversion: When the environment begins to deteriorate (phase IV and early in phase I), people are slow to sell losers (change corporate strategy). A famous study by Kahneman and Tversky in 1979[4] showed that people have asymmetric feelings toward losses and gains. Recognizing a loss is quite difficult, as eliminating the loss brings more gratification than doubling the damage. Consequently, people are very slow to change strategy or sell losers when performance is deteriorating. They may only do so when the pain is greatest, and just before reversal. Yet, selling early can save a performance record. To recover from a 20% loss (100 to 80), one needs to make a 25% gain (80 to 100), but to recover a 50% loss (100 to 50), one needs a 100% gain (50 to 100).
.
Contrary to perhaps popular belief, to overcome these biases and help make good decisions during times of extreme duress (6:00) and bliss (12:00), one should consider keeping the current leader at the bottom of cycles and replacing the leader at the top. At the bottom, a new leader may change the firm too much as he/she has free rein to set a new strategy into motion (as he/she was likely hired to do so), while the current leader probably understands the issues that caused the problems and is quite worried about losing his/her job and is finally willing to make correct choices. At the top, a change agent may be needed to recognize potential problems. A new leader is probably biased against the status quo/inertia since he/she wants to create his/her own legacy. Thus, he/she will scrutinize everything and perhaps identify negative issues and react accordingly. Conversely, at the top, the current leader is more likely to ignore possible problems and perhaps bask in this period of temporary bliss. Thus, management turnover can reduce under-reaction to signs of deterioration at the top, and result in over-reaction to negative conditions at the bottom.
.
Reflexivity
.
Behavioral biases are tied to the performance cycle, and they also exacerbate the cycle and thereby influence the biases themselves. Consider the following example.
.
Imagine you are a manager of a retail clothing store. Things have been going well for you. You were just promoted, the economy is good, and you just received a raise. You are confident in the outlook so you order higher inventory this season than the prior season. A recession hits and you discover at the end of the season that you have too much inventory that needs to be put on sale. The owner of the store is not too happy with you since profits are down, but you believe in yourself (you are overconfident) and find little reason (confirmation bias) to believe that your performance will be poor in the future. You order less inventory in the next season, but only to the level you sol
d last season since you tend to anchor expectations of the future on recent trends (you believe the past is representative of times to come). You also have friends who manage other stores, and they are all ordering based on the last season as well (you feel comfortable herding with them). However, your and other retail managers’ cuts to inventory put people out of jobs along the supply, finance, and marketing chain (the people who make, ship, finance, advertise, etc. the clothing). These are some of your customers, and they now have less disposable income to buy your products. Your season is terrible, you have higher than normal inventory again, and even lower profits than last season as you have to heavily discount prices again to move inventory at the end of the period. The media abounds with discussions of how severe the recession is becoming. You become scared. While you believe in your products and expect that the economy could get better, you also realize that if you order too much clothing again that you will be out of a job since everyone is herding by cutting back and the owner will believe you are out of touch. Thus, you order much less than normal during the next season, which makes the economy even worse, and so on.
.
.
The cyclic nature in which biases and fundamentals feed on each other, as illustrated above, creates efficacy performance spirals or what Soros calls “reflexivity.”[5] Lindley, Brass, and Thomas (1995) defined efficacy performance spirals as “deviation-amplifying loops in which the positive, cyclical relationship between perceived efficacy and performance builds upon itself.”[6]
.
Learning
.
Learning is most hampered during times of turmoil (6:00 on the Expectations Clock) and times of bliss (12:00) (Fiol and Lyles, 1985[7]) (figure 19). Without learning, or being able to think clearly, one may make poor decisions during extreme environments. Change at 6:00 may result in selling losers, firing management, and disposing of businesses at the bottom of cycles; thus, too much change is detrimental during times of turmoil. Lack of learning during phase III may result in one missing signs of when it is best to adapt. Without learning, it is not surprising that people over-react to signs of positive change as we move down phase 1 and under-react to negative information when conditions are best (near the top of phase III).
.
.
Rational Irrationality
.
Rational irrationality is when someone behaves irrationally for rational reasons. Specifically, one could behave rationally with one’s job by doing an irrational activity (selling a loser at the bottom (6:00) and going with the trend at the top (12:00)). Rational irrationality may be just as powerful a behavioral force as biases in causing people to make poor business decisions.
.
As noted earlier, going with the crowd, or herding, provides protection. Keynes said “Worldy wisdom teaches that it is better for reputation to fail conventionally than to succeed unconventionally.”[8] This means that it is better to sell a loser at the bottom than to potentially hang on to it and hopefully see it rise and outperform. Holding the loser and underperforming could cost one his or her job, but selling it and losing the opportunity of return will not be penalized (or at least as much).
.
Rational irrationality may result in a person not admitting to his/her mistakes since admitting errors may result in being fired. Instead, a person with a loss (in a project, stock, etc.) may double down in hopes of being proven right even if he/she believes the extra investment is likely to only compound the error. Of course, if one is proved wrong after the new investment, he/she may be fired. However, this is no worse than the result of admitting the error earlier, so it may be rational to bet more in hopes of success even if the hope is unfounded. Rational irrationality may drive loss aversion.
.
Examples of rational irrationality abound, but perhaps one of the most infamous example is a quote by Charles Prince, CEO of Citigroup in 2017, just before the financial bubble burst. “‘When the music stops, in terms of liquidity, things will be complicated,’ Mr. Prince said, when asked about problems in the US sub-prime market…‘But as long as the music is playing, you’ve got to get up and dance. We’re still dancing.’”[9] Was Prince extremely overconfident, and/or did he have to continue “dancing” or face the possibility of losing his job if he “sat out” and the bubble continued a few more years before exploding?
.
A common example of rational irrationality impacts many, if not most, professional fund managers. For example, a large-cap growth mutual fund manager may “have to” own at least some positions in stocks such as Apple, Amazon, and Alphabet even if he/she believes they are overvalued. Not owning them risks the possibility of underperforming since these behemoths make up a substantial percentage of large-growth benchmarks. Most clients (individuals) probably believe these are great companies, and therefore, should represent great investments and be part of the large-cap growth portfolio. If they rise substantially and the manager does not own them and he/she underperforms then the underperformance is unconventional (outside the herd) and this could cost one his/her job.
.
Unfortunately, examples of people being irrationally rational – going against the crowd and being rational – are few. This is partly because it is difficult to survive in business long enough (since clients flee) to tell stories about going against the herd and succeeding during significant bubbles and bursts. For example, many value managers who stuck to their discipline lost their jobs or went out of business during the internet bubble. The Big Short is a movie and a book (by Michael Lewis[10]) about a group of individuals who bet against the housing market. They were early, but it is very difficult to predict when bubbles will burst. They nearly lost their jobs! While being irrationally rational may cost one his or her job, it is the right thing to do.
.
I encourage you to do the right thing. When people are fearful (6:00), be greedy and look for signs of improvement, and when people are greedy (12:00), be fearful and look for signs of deterioration. If you do so, you will improve your chances of success while also helping to allocate capital to the best investments which ultimately leads to improvements for society.
.
.
[1] Yoffie, David, and Renee Kim, March 21, 2011, Apple Inc. in 2010, Harvard Business School Case: 9-710-467.
[2] Collins, Jim, and Morten Hansen, 2011, Chapter 4: Fire Bullets, Then Cannonballs, Great by Choice (HarperCollins Publishers, Inc., New York, NY).
[3] http://www.worth.com/why-are-the-words-this-time-its-different-called-the-four-most-expensive-in-the-english-language/.
[4] Kahneman, Daniel, and Amos Tversky, 1979, Prospect Theory: An Analysis of Decision Under Risk, Econometrica 47: 263-191.
[5] https://en.wikipedia.org/wiki/Reflexivity_(social_theory).
[6] Lindsley, Dana, Daniel Brass, and James Thomas, 1995, Efficacy-Performance Spirals: A Multilevel Perspective, Academy of Management Review 20, 645-678.
[7] Fiol, C. Marlene, and Marjorie Lyles, 1985, Organizational Learning, Academy of Management Review 10, 803-813.
[8] https://www.marxists.org/reference/subject/economics/keynes/general-theory/ch12.htm.
[9] Freeland, Chrystia, October 8, 2009, Investors Had Little Choice But to Keep on Dancing, Financial Times.
[10] Lewis, Michael, 2010, The Big Short: Inside the Doomsday Machine (W. W. Norton & Company, Inc., New York, NY).
.
.