Attentional Biases And Financial Markets

Attentional Biases are operative in everyone’s psychology; they can affect performance in financial markets because they control what information sources we consider

Are happy people better at picking up information that will make them happen?  Do sad people do the opposite?  Have you wondered how your mood can affect your behaviour in ways you don’t know about?  All of this is true and can be explained by considering one form of a Cognitive Bias called Attentional Bias.

We are subject to approximately 150 Cognitive Biases, at the last count.  All of them affect our thinking without us necessarily knowing too much about when they are at work or what the results are.  My project initially is to list and describe these mental subroutines before critically examining them and assessing how they work in a market environment.  The objective is to allow market participants to look out for the operation of Cognitive Biases in their own thinking and trade on the expectation that they will also figure prominently in the thinking of other players.

One of the most important Cognitive Biases is known as Attentional Bias.  It comes in several forms, but all of them have in common that they systematically slant which information we pay attention to.  Obviously this can be expected to have dramatic effects on thinking and market outcomes.  In this post, I will first describe Attentional Bias and then outline how it might play out in a market setting.

Much of the psychological literature on Attentional Bias looks at what we can term mood congruency.  The basic idea here is that we are more likely to look at information which fits our mood.  So, anxious subjects are more likely to look at anxiety-inducing information and depressed subjects are more likely to consider depressing information.  Clearly this is already rather unhelpful for such subjects, but my aims here are only to look at what this might do in markets.

This is widely important because generalised anxiety affects a significant proportion (estimated at between 5% and 30%) of the population.  This is people who are more-or-less anxious more-or-less all of the time. Since it is a significant  minority, it is likely that some of these subjects participate in financial markets, although it is possible that some anxious individuals will self-select out of stock markets.

Depression of sufficient gravity to merit a psychiatric diagnosis affects about 1% of the population; many more people will experience a less severe depression or a more episodic form.  Again, we can expect plenty of market participants to be depressed when trading.

Experimental investigations of mood-disorder linked Attentional Biases have focused on reaction time studies.  A pair of words was briefly presented to experimental subjects on a computer screen.  Sometimes, one of the words was replaced with a dot, which was the signal that a button should be pressed.  The time it took for subjects to press the button was recorded.  It would typically be in the range of several hundred milliseconds.

Sometimes, the other word presented on the other side of the screen to the dot was a threatening word.  The word could be socially threatening (‘humiliated’) or physically threatening (‘injury.’)  The experimenters found what is known in psychology as an RT spike — or a delay in reaction time.  People took longer to see and react to the dot if a threatening word appeared on the other side of the screen.  These effects were quite large.

Perhaps most interestingly, the RT spikes were larger for anxious or depressed subjects, especially if the threat word was specifically related to either anxiety or depression.

What Effects Of Attentional Bias Should Such Individuals Be Aware Of?

It is obvious that such effects could impair traders on a trading floor who are making rapid trade decisions themselves.  Information near their field of vision which is threatening — such as a negative Bloomberg headline — could grab the trader’s attention and cause a delay in response time even if it is unrelated to the trade under consideration at the time.

While this is a real issue, I want to consider non-professional traders as well. In general, day-trading is best avoided as 85% of day traders lose money.  (Day-trading is popular among people new to investing.  It is called that because the aim is to minimise risk by not holding any positions over-night.  However, the necessarily short-term nature of this approach means that one can really only benefit from ‘noise’ in stock movements and there is no way to rationally forecast noise.  Relying on luck is even worse in markets than elsewhere because the punishment is swift.) It is better to be a buy-and-hold investor.  What effects of Attentional Bias should such individuals be aware of?

If one is episodically depressed or anxious, then these are not times to be trading.  Negative mood-congruent information will grab attentional resources and make traders much more likely to exit positions.  This may or may not be the right decision to make; what is clear is that such a decision should be made rationally and with a fair and open consideration of the relevant data.  Often this will not be what everyone else is doing, so my approach lends itself naturally to a contrarian investment stance.  There are other good reasons to be a contrarian investor, including that it fits with a long-term approach — so it is not something much engaged in by day-traders.

If someone is permanently depressed or anxious, then treatment should be sought and one should abstain from trading until an improvement is seen.  If no such improvement can be achieved, then I am sympathetic, but I would suggest hiring financial advisers in that circumstance.  It would be one thing less to be concerned about and would likely have more optimal outcomes, despite the extra fees involved.

I discuss in much more detail the important effects in financial markets of Cognitive Biases like Attentional Bias in my new book:

Email me at

Send Mail

Opposition to Gun Control is not “Superstition” 

There have been suggestions recently that some voters are immune to evidence and argument. They rely instead on gut feelings and instinct. These individuals are described as being “intuitionists” or “superstitious.”  I will suggest that while this is the correct direction of travel, that we can in fact be more precise and locate the issue as a consequence of cognitive bias known as Status Quo Bias.

The Onion puts the strangeness of the gun control debate best with its satirical headline: “No Way to Prevent This, Says Only Country Where This Regularly Happens.” To rational observers, it seems obvious that if you have 310m guns in a country and 93 people a day are killed by them, you should reduce the second number by reducing the first. Almost every other country in the world does this and it works. Yet a majority of Republicans and even 25% of Democrats disagree. This is ignoring data on a massive and deadly scale.

I think this is not best explained by appealing to superstition as The Economist did recently. Superstitious beliefs, to be sure, are not based on data and do not often result in true claims. If they do, it is a coincidence: superstition is not a reliable method of arriving at true claims. There is no such thing as bad luck and walking under a ladder wil not bring it. Opponents of gun control do not believe that firearms are lucky charms.

The appeal to intuition, by contrast, can I think throw light on the topic if precisified in the right way. I would understand intuition as being a collection of cognitive biases. These operate to slant and indeed direct our decision making, largely unbeknownst to us.

The Bias I have in mind here is called Status Quo Bias. For my U.K. readers, I should immediately clarify that this has nothing to do with Francis Rossi. The Bias is also known as the familiarity effect. I will introduce it by asking you to make a quick choice.

Would you prefer to meet a friend for lunch somewhere you have been before or would you rather go to see a stranger to pursue a novel activity in an unknown location?  Most people most of the time would choose the first option.

As with all Biases, this one has its origins in being valuable. Most of the time, it will produce the right result. This is because of a very approximate risk assessment heuristic. We assume that things we have done before which have not harmed us visibly are safe activities. This is wrong but better than nothing. It is in fact I believe an outgrowth of another Bias called the Availabilty Heuristic, but I will set that aside for now.

Status Quo Bias though can be rephrased as the idea that any change is more risky. This can produce conclusions which are as uncongenial to the left as to the right. It is not true, for example, that because countries have borrowed heavily in the past, they can continue to to do indefinitely. As to the topic at hand, for the “intuitionists,” changing to a scenario of tighter gun control is risky because it is new, rather than safer because all of the countries that do it are safer.

If you are wondering what this means practically, I suppose that depends on whether Cognitive Biases are hardwired in to us. That’s unclear, but I think it is at least a start for me to list the mental subroutines we are running. If they are hardwired, then they will be impervious to data, which would explain why the debate is sterile: proponents of gun control continue to say “if we changed this fewer people would die because that is what happens when you have gun control” and opponents would continue not to listen.  If they are not hardwired, then telling people they have these biases might be a start in the direction of changing them.

My new book focusses on these biases and their effects in financial markets:

I recommend it if you want to be aware of the subconscious processes which guide your behaviour in markets and elsewhere and if you want to become less dependent on your own autopilot which will not be optimising your outcomes.

83% of Millennials Regret How They Handle Their Finances: Why?

Hyperbolic discounting explains why we fail to plan early enough, which is the most common financial regret among Americans

It is reported that 71% of Americans express regret about their ability to handle their finances; the percentage rises to 83% among Millennials.  The most common regret, expressed by 48%, is failure to plan early enough.  Why does this happen?  I will explain it using an element of our psychology called Hyperbolic Discounting.  This is one of 150+ different sorts of Cognitive Bias.  In my new book, I discuss the most important Cognitive Biases and how they will affect your investment performance (see link below).

So what is Hyperbolic Discounting and how does it explain our failure to plan early enough?

As with many Cognitive Biases, they have a kernel of value within them.  This has to be true because otherwise we would not have them.  They can be seen as “quick and dirty” heuristics which in most everyday situations are good enough to allow us to get by.  If they are mostly right and avoid any scenarios of catastrophic error, then they probably do enough to pay their way in our mental architecture.

As an example, think of the widespread fear of snakes.  Evolution could have aimed to give us only a fear of venomous snakes, but that would have been difficult to achieve and would have involved a risk of missing some snakes that could kill us.  Better than this is to make people afraid of all snakes.  The cost of that is that people will sometimes run away from some snakes that are harmless.  But that’s fine.  That cost is greatly outweighed by the benefits of avoiding the venomous snakes.

Hyperbolic Discounting is one of these sorts of mostly useful bias.  It is founded on something like the common and accurate idea that “a bird in the hand is worth two in the bush.”  In other words, something I have now is more valuable than something of equal value that I will have in a year from now.

This then raises the question of how to compare the present value of an item I now own and the present value of something I will own a year from now.  This means applying a discount to the future item to account for the delay between now and when I will own it.  In a world of low interest rates, it is easy to forget this.  But when they return to 5% a year, it will be a lot more clear that $100 now is worth 5% more than $100 in a year from now. Because I could save the $100 now and it would be worth $105 in a year from now.

Turning this around, I can work out what the present value of the future $100 by discounting  it.  This means multiplying it by (100/105).  This comes to $95.24.  (You can check this by adding 5% to it and getting back to $100.)

So 5% is the Discount Rate here.  This is how you should discount a future certain $100 by under circumstances where the risk means that is appropriate.  Now we come to the problem with Hyperbolic Discounting.  It seems that our default discount rates are set way too high.  We set far too much store by what we have in our hands now.  This is also perhaps reflected in the way we are prepared to smoke and not go to the gym today.  Those things are easy to do and carry only minor immediate costs. Smoking of course carries an infinite cost at some point in the future because it will kill you.  It is only through Hyperbolic Discounting that anyone can manage to smoke. Similarly, not going to the gym will kill you.  But not today.

These and many other Cognitive Biases are who we are and explain much of our decision making.  The approach I take in the book is to describe some of the financially significant ones and then explain how they play out in financial markets.  Thus, by reading the book, you can obtain two key benefits.  Firstly, you can look out for biases like Hyperbolic Discounting in your own thinking and correct for them.  Secondly, and even more valuable, you can expect them in other market participants and trade accordingly.

If you want to discuss these and other concepts mentioned in the book, or for more information about the book, you can Send Mail

Cognitive Biases And How They Affect Stock Markets

There are an extraordinarily large number of cognitive biases which change the way we think — it is important to know about these in stock markets both in order to look for them in one’s own thinking and expect them in that of other market participants

A Cognitive Bias is an element in our psychology which makes decisions for us.  In fact, given there are so many of them, you might even say that our psychology is just Cognitive Biases.  I discuss about 20 of them and give market context in my new book:

One example which I discussed in a previous post is Hindsight Bias.  This makes us think that everything which has happened was inevitable.  This in turn makes us very bad at knowing what level of probability we assigned to past events.   We seem to remember that we said that a prior event was going to occur with 90% probability when in fact, we gave it a probability of 30% before that event occurred.  This will obviously have unhelpful effects in the context of markets.  It will make us greatly overestimate our chances of knowing what is going to happen next, which may make us trade with over-confidence.  There is nothing quite like over-confidence to kill market performance.

As I said, the book covers around 20 biases and their effects in markets in much more detail than I can give above.  The thing is though, that there are way more biases than that.  In fact, from the chart below, which is due to Bence Nanay, you can see just how many there are.

Every one of these biases will affect your market thinking and that of everyone else involved.  So it looks like we all have some work to do!

Email me:

Send Mail

Empathy Is Useless

Empathy is popular but I argue not helpful in financial markets or elsewhere

I have recently seen an interview with a poker champion in which he claims that “empathy” is one of the strengths of his game.  I will suggest this is false for a couple of reasons.  Firstly, we don’t know what it is.  Secondly, what is more likely to be useful is the related but distinct concept of “Theory of Mind.”  I think there are many parallels between poker and financial markets, so my position amounts to an argument that you don’t want empathy to make your stock portfolio perform.

I discuss at greater length the sort of psychology which is more likely to be useful in my new book:

There are several things that empathy might be.

  • I feel the same emotion as you
  • I can say what emotion it is that you are feeling
  • I feel a similar emotion to you
  • I feel a weakened form of the emotion that you feel
  • I “sympathise” with your situation
  • I can imagine what I would feel in your situation

This is already complicated enough, without asking difficult philosophical questions like “what does ‘same’ mean here?” or “what is the effect of similar?”  And we haven’t got on to the main point yet.

In poker and in markets, what you want to be able to do is predict and explain the behaviour of others.  That is what is known in psychology as a Theory of Mind task.  I have argued in my first book that the way we do this is by simulating others.  We imagine that we are in the situation that the others are in and see how we would feel.  We then predict that they will behave the same way as we would given the resulting emotions.  Note that this does not fit neatly into any of the categories above.

This I think shows the first problem with claiming that empathy is good for performance.  This isn’t a useful piece of advice without defining what empathy actually is.  And secondly, think about what it would do if you felt people’s emotions.  Do you want a surgeon who is about to operate on you to be paralysed with fear because you are?  Do you want a pilot to experience the emotions of passengers as he wrestles with the controls in a storm?

I think that when it comes down to it, what you want is actually a good Theory of Mind.  In fact, as I discuss in my book, there is experimental evidence that better traders have better Theory of Mind.  For that reason, I spend an early chapter explaining how we think Theory of Mind works.  It can be improved I think by taking account of cognitive biases and that will lead to better trading performance.  It will probably improve your poker game too!