Biases in Investing: Hindsight Bias
I previously started to write about biases and how they might affect investment decisions and could lead to huge and costly mistakes. Psychology is playing an important role in investing and although there is extensive research on the subject of behavioral finance, many investors still seem to underestimate the importance psychology is playing in investing. Instead, they are often focusing on other areas of knowledge and expertise: economy, accounting, sector knowledge, mathematical and statistical skills.
In the following article I will try to present to you another biases — the so-called hindsight biases — but will start once again with a definition what a bias is.
Definition Bias
(If you are already familiar with biases, you can skip this section)
In general, a bias is the tendency towards a certain perspective (or a certain opinion or belief). And while biases can be conscious and unconscious, they are always influencing the way someone thinks or behaves. Especially in decision-making and problem-solving these biases can be a huge problem and a threat as they might lead to terrible decisions — for example terrible investment decisions.
We can try to group these biases in different categories, including:
Cognitive biases, which arise from the way our brains process information — like the confirmation bias (the tendency to look for and rather accept information that is in line with our already existing beliefs).
Memory biases affect the way we remember information and examples include the hindsight bias (the believe that events were more probably after they occurred).
Decision-making biases, which affect the way we make decisions — like the sunk-cost fallacy (when we continue to invest in a project because we already invested money before although further investments don’t make sense).
Hindsight Bias
Hindsight bias is describing the misconception that people are seeing events more predictable and more likely after these events occurred. It is describing the tendency of people to believe after the event has happened that they would also have predicted the event beforehand. People tend to believe they knew it all along — even if that is hardly ever true. And the hindsight bias is describing the fact that people are overestimating their own predictive powers when talking about past events.
When looking for reasons why the hindsight bias occurs, we can mention two different aspects:
It is influenced by our tendency to filter memories and we are remembering facts and our own actions and decisions differently after a few weeks or months have passed.
And it is also influenced by the fact that usually more information is available after the event. And with the available information after the event, it would have been predictable beforehand, but we often tend to ignore that the information was simply not available, and it would have taken visionary abilities to know the information beforehand.
And the hindsight bias can lead to different problems for investors and could be the reason for costly mistakes.
Backtesting Models
While backtesting is probably more common among traders than investors, I still like to mention it and explain why we should be cautious and that the hindsight bias can lead to problems. When already knowing what worked in the stock market in the last few years it is easy to generate a strategy based on that knowledge and we already identified the problem the hindsight bias is creating. Knowing what worked in the past does not mean we will know what will work in the future.
Additionally, we must be very careful on what database we are testing. In general, the database should be going as far back as possible and include as much securities as possible. But I would assume many people will only test on the last 10 years which is including no severe bear market, which is already a problem and a distortion. And even if we go back a long time, the starting point will make a huge difference (at least for long-term strategy). Results will be different when starting in 1929 instead of 1932.
A second problem considering backtesting is the so-called survivorship bias, which will also play a role. We will most likely use the data of securities still trading on the stock exchange and ignore those stocks that were delisted a long time ago (for example due to bankruptcy). And by looking only at the companies that survive, we are once again getting a distorted picture.
Being Overconfident
Another problem arising from the hindsight bias is investors becoming overconfident. If the hindsight bias is leading to the illusion we would have been able to predict events in the past, it is just a small step to assuming we will also be able to do so in the future. And being overconfident is always a huge problem in the stock market — as it can lead to costly mistakes. I have written in a past article, that we should always be humble in our approach to financial markets.
And it is easy in hindsight to see why everybody should have bought Amazon in 2002 or 2003 and Alphabet (or back then: Google) in 2004. Now it is easy to argue why Amazon was a great investment — but I assume only a handful of people will be able to identify the next Amazon.
The hindsight bias is especially brutal to decisions makers — like CEOs or money manager making investment decisions for others. If these people make mistakes (and they will make mistakes) everybody will claim they should have known better (and that is was an obvious and unnecessary mistake from the beginning). In hindsight, changing the recipe for Coke was a stupid idea and selling Instagram for only $1 billion was obviously a price tag way too low. And after the Dotcom bubble, everybody knew it was a bubble as well as everybody in 2009 knew the housing market in 2005 and 2006 was irrational and set up for disaster.
You should be aware of overconfidence as it will be punished hard. If you start to ignore basic risk management principles due to overconfidence, you risk losing everything.
Identifying Patterns
A third and final problem is that people tend towards identifying patterns. There are hundreds or thousands of books trying to identify patterns for a successful life and what habits one must “develop” to lead a happy and successful life. Especially when the topic is self-optimization those books are mostly just crap and not worth reading. First, many of these books fall victim to another bias I already mentioned above and that is hard to fight — the survivorship bias. When writing books about Jeff Bezos, Bill Gates or Warren Buffett and claiming you just need to do the same these people did (maybe reading a lot or starting early like Warren Buffett) they are ignoring the millions of people that did the same and did not become the next Warren Buffett. In our case, these multi-billionaires are the survivors we are always looking at.
And especially when writing about self-optimization it is almost impossible not to fall victim to the survivorship bias as we usually don’t have enough data on those that failed. The situation is a bit better when talking about businesses as good, scientific studies can also include those businesses that failed in the past (as we have more data). But as Kahneman is pointing out, even great books like “Built to last” (which I enjoyed reading very much and quoted several times) are still flawed. Many of the successful companies identified in these books underperformed after the publication. And I also focus on the concept of wide economic moats and trying to identify patterns that lead to the outperformance of a businesses over long timeframes. But Kahneman writes:
Knowing the importance of luck, you should be particularly suspicious when highly consistent patterns emerge from the comparison of successful and less successful firms. In the presence of randomness, regular patterns can only be mirages (Kahneman 2011, p. 207).
It is a major problem that humans tend to look for patterns we can follow and in our search for these patterns we identify patterns where none are. We seem to have extreme difficulty to identify that luck plays an important role in life:
Consumers have a hunger for a clear message about the determinants of success and failure in business, and they need stories that offer a sense of understanding, however illusionary (Kahneman 2011, p. 206)
Fighting The Hindsight Bias
I already mentioned a few ways you can fight the hindsight bias. Just accepting that luck is playing an important role and that some events are out of our control is a good starting point. When accepting that we can’t time the market and knowing that nobody rings a bell on the market top is also a good starting point.
And we should always keep in mind that the economy and financial markets are complex, non-linear systems and predicting a certain outcome when dealing with such systems is (almost) impossible. When making predictions about simple, linear systems with high predictability — for example: the sun will rise in the morning as it does every day, or: the trading volume is rather high at the beginning of each trading day — it is rather easy and possible to make predictions.
But hindsight bias is almost always occurring when dealing with complex, non-linear systems. And while predicting the sun will rise tomorrow is easy — identifying when a meteor might hit the earth is not easy as there are no simple patterns and not a lot of past data available. And financial markets as well as the economy are extremely complex systems with thousands or million of participants influencing the system, plus having a memory and the possibility of unintended consequences arises. And therefore, we should stay humble and always include a margin of safety in our calculations.
Another way to fight the hindsight bias is by writing an investing journal which enables us to go back in time — for example to the years 2007 or early 2020 — and see if we really were able to forecast these recessions (and the pandemic) or if we are just lying to ourselves and were blind like almost everybody else (and we should also look if we did not forecast recessions in other years as well that did not happen). And the more information you include about your reasoning and motives for certain investments, the more helpful it becomes. One way for such an investment journal might be writing on Medium (or in my case: Seeking Alpha) — these articles are available, and I can go back in time and see what I wrote in early 2020 or can identify why I thought we might be close to a top in late 2016 (see here and here) which was clearly a mistake.
Conclusion
Similar to anchoring and priming which I mentioned in my last article, the hindsight bias is also a problem for investors. It is problematic as it makes us identify patterns where obviously none exist and makes us overconfident about being able to predict the future as we are suddenly under the impression that we were able to foresee past events — even though this is not true. But as long as we are trying to be humble about the stock market and always remember the stock market (and economy) is a complex, non-linear system an important first step is made.
Literature:
Kahneman, Daniel (2011): Thinking, Fast and Slow. Penguin Random House UK
Note: Article was originally published on Medium in March 2023 (can be found here)