Last year finished up being a real surprise for many investors and commentators alike. After dipping 15 per cent in the early part of the year, the Australian sharemarket looked set to repeat 2008’s 38 per cent fall. But how different the year turned out to be, with the market staging an incredible recovery from its March lows to finish up 37 per cent for the year – its best calendar year performance since 1993.
The recovery in the sharemarket was accompanied by a rebound in investor sentiment in 2009, according to the IFSA-CoreData Investor Sentiment Index (refer to chart 1). With the economy and sharemarket improving, many Australian investors may be asking, “What Global Financial Crisis?” Those of you who have read the previous articles on investor behaviour in Professional Planner may not be surprised that the recovery in investor sentiment has corresponded with the bounce in the sharemarket. Both articles highlighted the powerful role emotions and investor psychology can play in driving sharemarket returns. Past experiences, pre-conceived ideas and fear of regret are just a few of the biases and behaviours that can lead investors to make emotional and often irrational decisions.
Often these are powered by two very strong underlying emotions – “greed” and “fear”. And as we move into more positive and optimistic times, we may well see investor greed begin to creep back in and dominate once again – replacing two years of pessimism and fear. In this article we look at two more types of behaviour: “anchoring and adjustment”; and “aversion to ambiguity”. Both of these heuristics (or “rules of thumb”, often gained from trial and error) have a strong influence on investor behaviour and the choices they make, and in many instances can lead to overly conservative investment decisions and inadequate diversification.
Anchoring and adjustment
Anchoring and adjustment describes the situation where investors rely on one piece of initial information and fail to adequately adjust their view to take into account new information as it comes to hand. The natural tendency is to place too much emphasis on the initial piece of information. Amos Tversky and Daniel Kahneman, two highly-regarded behavioural finance experts responsible for identifying this anchoring and adjustment heuristic, first highlighted this behaviour in 1974 in a well-known experiment. They asked two separate groups of people what percentage of the United Nations comprised African countries. The first group (Group A) was asked, “Is it above or below 10 per cent?” (The number “10” was “randomly” selected from the spin of a wheel; however, the wheel was actually rigged to ensure it landed on 10.)
The second group (Group B) was asked, “Is it above or below 65 per cent?” (Again, the number “65” was “randomly” selected from the spin of a wheel.) Group A’s average answer was 25 per cent and Group B’s average answer was 45 per cent. This was a marked difference between the two groups. Each group had essentially used their respective “randomly generated number” as an anchor for their decision, rather than answering with an independent frame of mind. This is despite the numbers 10 and 65 being apparently random. This conservatism in adjusting forecasts was also explored by psychologist Ward Edwards (1964). He conducted research on a number of people using black and red poker chips. He asked his respondents to imagine 100 bags, each containing 1000 poker chips, whereby: • 45 bags contained 700 black chips and 300 red chips; and • 55 bags contained 300 black chips and 700 red chips. Imagine that one of the bags was selected randomly.
Ward then asked the respondents two questions in two stages (and they were not allowed to look inside the bags). Question 1: What is the probability that the selected bag contains mostly black chips? Most of the respondents answered 45 per cent. He then asked the respondents to imagine that 12 chips were randomly selected from the bag – eight black chips and four red chips, whereby each ball was subsequently replaced. He then asked another question. Question 2: Would they change their first answer in response to the new information? A number of respondents remained with their original answer of 45 per cent, while others adjusted their original answer to incorporate the new information. In answer to the second question, most respondents said, less than 75 per cent (that is, they believed there was a 75 per cent probability that the selected bag contained mostly black chips).
The correct answer is actually 96.04 per cent (calculated using probability theory). The fact that most of the respondents’ answers were well below this number showed that many of them had underreacted to the new information. This outcome suggests that they may have been too anchored to their original answer of 45 per cent and did not adjust their assessment sufficiently. Ultimately they were too conservative in their revised assessment when presented with the new information. This same research has also been conducted on analysts’ reactions to company earnings announcements. They often don’t revise their earnings estimates enough following the release of new information.
As a result, positive earnings surprises often can be followed by more positive earnings surprises, and negative earnings surprises can be followed by more negative earnings surprises. Behavioural finance theory suggests this conservatism in analysts’ earnings predictions can create mispricing. Work by Victor Bernard and Jacob Thomas (1989) showed that those stocks that had positive earnings surprises outperformed those with negative earnings surprises. These unexpected surprises can also be a function of over-confidence (another heuristic covered in our first article) whereby investors’ overly narrow confidence bands mean they are more surprised than they initially expect to be.
Anchoring and adjustment is something we all do, mostly without realising it, and retail investors may well be a victim of it in 2010. The market’s recent strong rally has factored in favourable company earnings forecasts – however, if these numbers don’t live up to expectations, the market will fall sharply. The question is: will investors adjust their own stock/market return forecasts sufficiently to reflect the downwardly revised earnings? It’s highly unlikely – and they could therefore be very surprised and disappointed with the outcome. Basing investment decisions on past returns of asset classes is another common example of anchoring behaviour.
A look at the historical returns of each asset class in Table 1 (next page) over the past 20 years shows that anchoring behaviour is a fruitless exercise. Last year was a classic example. After returning -38.4 per cent in 2008, those investors who shied away from Australian equities in 2009 and chased Australian fixed interest and cash, following their relatively stronger performance in 2008, would have paid the price, with Australian equities returning 37 per cent over the year versus 1.7 per cent for Australian fixed interest and 3.3 per cent for cash. As Table 1 illustrates, there are repeated examples of when investing in the previous year’s best-performing asset class has proven to be a poor strategy (often proving to be the worst-performing asset class the following year). While looking at the historical average return of an asset class can provide an initial guide or a benchmark, it’s not wise to use it as the sole basis for projecting future short-term returns. As with any investment strategy, rather than chasing or over-weighting too heavily towards last year’s best- or worst-performing asset classes, investors need to invest in a diversified portfolio and adopt a long-term asset allocation strategy.
Aversion to ambiguity – we prefer the known to the unknown
Another heuristic bias that investors are prone to is “aversion to ambiguity”. Essentially, when faced with a decision, investors fear the unknown, preferring the familiar to the unfamiliar. They want to know what they are dealing with and like to have a degree of confidence in the distribution or the odds of the outcome. According to Frank H Knight (1921), it is possible that people dislike uncertainty more than risk, particularly when the odds are unknown. Knight defined risk as a gamble with known odds and uncertainty as a gamble with unknown odds. Studies conducted by Hersh Shefrin, a professor and pioneer in behavioural finance theory, on his MBA students showed that if they were given the choice between a guaranteed $1000, or an even gamble where they could win either $0 or $2000, only 40 per cent accepted the gamble.
Shefrin then varied the example by telling them there was a bag containing 100 poker chips comprising 50 black chips and 50 red chips and the students had the option of choosing the guaranteed $1000, or a lottery ticket that pays $2000 if a black chip is drawn at random from the bag, but $0 if a red chip is drawn. To test their decision-making preferences, he also gave them the same option – but this time the mix between the red and black chips was unknown. He found that even those students who were prepared to gamble when the odds were even, preferred the safer option (that is, take the guaranteed $1000) when the odds were unknown. We have recently experienced aversion to ambiguity on a grand scale during the global financial crisis, which resulted in both the US and UK governments rescuing a number of “too-big-to-fail” US and UK banks.
At the peak of the crisis, there was tremendous fear about what would happen if the governments didn’t rescue these banks – the global financial system could have collapsed. Consequently the governments needed to step in, as they weren’t prepared to risk the unknown. This fear of the unknown was also pervasive in 1998 with the rescue of hedge fund manager, Long Term Capital Management (LTCM) by the US Federal Reserve. The then President of Merrill Lynch said of the rescue, “It was a very large unknown. It wasn’t worth a jump into the abyss to find out how deep it was”. One of the implications of aversion to ambiguity for investors is insufficient diversification – particularly across countries.
It is quite common for investors to have a higher weighting towards their home country than international markets – certainly much higher than portfolio models suggest. This “home country bias” has been well documented and is a global phenomenon. The reason? They are much more familiar with the domestic economy and local companies than international companies: they know them, work for them and have much higher exposure to them via the media. While the home country bias is understandable, it’s not necessarily the most sensible strategy. It is worth mentioning, though, not all home country bias can be attributed to aversion to ambiguity. Information theory may also play a hand, since investors, even mums and dads, in most cases have a higher level of awareness and understanding of the domestic economy and the companies that operate within it. Home countries can also offer tax benefits, such as franking credits, which can increase the appeal of investing locally.
Diversification and advice is key
“Anchoring and adjustment” and “aversion to ambiguity” are just two examples of behaviours that can plague investors, often leading to conservatism, disappointment and insufficient diversification. Identifying and understanding these behaviours can not only assist investors with making more informed and rational decisions, but can also provide them with an edge over other investors who are not aware of these behaviours – allowing them to identify and benefit from mispricing opportunities that can result from irrational investor behaviour. However, it’s difficult to shake these behaviours – many are entrenched. Seeking independent financial advice and investing with a professional fund manager are steps investors can take in order to reduce the role these behaviours play in their decision-making process.
Bob Van Munster is head of Australian equities for Tyndall Investement Management
