In the Rethinking Treasury series, we have discussed at length on how behavioural biases can impact our decision making and hence the importance of relying less on our instincts but more on systematic and quantitative approaches to analysing risk.

    In this sub-series, we will take a step back and look at various sources of cognitive errors. Last month, we looked at the importance of stories to our brains. This month, we will look at something fundamentally contradictory to stories – randomness – also a fundamental feature of markets.

    Why randomness and stories are incompatible

    Stories rely on cause-and-effect relationships to piece together otherwise disparate elements. Randomness, by definition, has no such cause-and-effect.

    Our brains, being conditioned to understand the world around us through stories, are not very good at handling randomness, at least not in a rational way. In this article we will see how randomness can impact our risk management decision making in several ways, and what this can tell us about building a more robust treasury management process.

    The “Slot Machine” effect: why randomness is captivating and random payoffs addictive

    Imagine the last time you were walking on a street on the way to the office. Were you thinking of all the pavements, pedestrian lights and cars on the road? Most likely not. Our brains do not pay attention to things that are expected within our mental model, because if we did, we probably would not have any brain power left once we reached the office.

    Our brains are conditioned to turn our attention away from the regular and towards the unexpected, and be captivated by it.

    In one famous experiment (Skinner 1948), hungry pigeons were placed in a cage where they had to learn the “secret code” to peck the food lever in order to receive food. Sometimes it was just a tap, other times a few taps. Pigeons are quite intelligent – if there was a fixed, predictable pattern between the number of pecks and the amount of food that gets delivered, pigeons would figure this out pretty quickly.

    What’s more interesting was when the food lever delivered a random payout. A pigeon might get plenty of food tapping the lever now, but suddenly nothing for the next five hours. Guess what happened to the pigeons? Initially, they started to “learn” whatever response they were making before the food appeared and behave as if there were a causal relationship – where in fact there was none. Also, when compared to other schemes of rewards, the random payout “trained” these pigeons to be the most persistent and heaviest users of the food lever.

    Essentially, by subjecting the pigeons to “slot machines,” they were turned into superstitious addicts.

    Neuroscience would later suggest that a chemical in the brain, dopamine, was responsible for the addictive behavior when we face “variable-ratio schedule” payoffs. In a similar vein, many have suggested that the random intervals at which emails, social media posts and text messages arrive and the dopamine that they trigger can explain why we find them so addictive.

    In markets, treasury risk managers need to be wary of this “slot machine” effect. To a large extent, markets can be thought of as a giant slot machine with highly random payoffs that are bound to surprise us from time to time.

    If we are not careful, we can become like those pigeons who pick up “illusory correlations” and act superstitiously as if there were a causal relationship. With money on the line, the variable payoffs and the dopamine that they trigger could also turn us into compulsive gamblers rather than rational hedgers.

    Our brains are conditioned to turn our attention away from the regular and towards the unexpected, and be captivated by it.

    Hindsight bias: we are very good at back fitting stories to randomness

    Why are we so prone to picking up illusory correlations?

    Perhaps because the world is a much more random place than our brains are ready to accept.

    Historians use cause-and-effect relationships to weave together historical events, and with hindsight there often appears to be a sense of inevitability. For example, when a country wins a war, history books would be filled with reasons why this happened, but much less emphasis is given to ways in which the country might have lost. This can leave some with the impression that “history” is pre-ordained.

    “Hindsight bias” occurs when our brains under-appreciate the randomness in historical, realised outcomes, and is consistent with human natures’ urge to find narratives to explain events. (See more detail about hindsight bias in the sixth article of the Rethinking Treasury Series)

    “Resulting”: when we judge the quality of our decisions solely by outcomes

    The game of poker has a high degree of randomness. Unlike chess, which is a perfect information game (all there is to see is on the chess board), poker is an imperfect information game with high uncertainty due to the unseen or undealt cards. This makes poker a much closer model to what we experience in markets than chess does.

    Good poker players understand the difference between good decisions versus good outcomes.

    For example, a hand with 70 per cent chance of winning means that it could still lose, 3 out of 10 times.

    Let’s say the less probable happened. The 3-in-10 chance event has unfortunately transpired. Does it make the original decision to play the hand a bad one? Absolutely not. A rational player would have no problem handling the outcome, because he knew he had won “probabilistically.”

    That is why professional poker players learn to control their emotions and keep up their “poker face.” But not everyone is quite so rational. On an emotional level, it can be difficult to reconcile a bad outcome with a good decision.

    Key takeaways for CFOs and treasurers

    In this article, we have discussed some perils of randomness and how they can impact our decision making on a cognitive level.

    • Randomness, or simply noise, gets our attention when they should have been ignored, because our brains are conditioned to look for signs of change and the unexpected. As Nassim Taleb famously said in his book, Fooled by Randomness, “bad information is worse than no information at all.”
    • The “slot machine” effect demonstrates that random payoffs are the most addictive kinds as they trigger dopamine in our brains. We can erroneously see patterns in randomness that are not there and let this impact our behavior accordingly. This is because our hindsight bias helps us piece together stories and explanations that might not be there at all.
    • Randomness can cause bad outcomes to good decisions but also vice versa, i.e. good outcomes to bad decisions. The ability and willingness to objectively differentiate between the two are key.

    Data analytics can therefore be transformative and lead to better decision making, as data stands for objective evidence.

    For example, with data analytics, we can use historical data as evidence to objectively determine whether a “signal” you are seeing is repeatable across history, or whether it was simply “random noise.”

    However, data analytics cannot simply be a substitute, as human intuition is still vitally important to make sense of contextual information and foresee new possibilities that are not “included in the data.” (e.g. Black Swan scenarios that never happened before but can still logically happen in the future).

    To summarize, treasurers and CFOs should leverage both human and data capital to create a system that is stronger than the sum of its parts, with a robust and holistic evaluation framework that focusses on the quality of decisions made, rather than individual outcomes.

    HSBC Thought Leadership has launched a brand new series called “Digital Strategy Series” in which we will explore in more detail how this partnership between man and machine can help transform treasury risk management.

    Next month, we will talk about other cognitive errors related to our own emotional “selfishness.”

    Disclosure and disclaimer

    More, collapsed
    Demystifying Insurance Hybrid Capital
    There is steady flow of issuance from insurance companies that can complement an investor’s holdings in more traditional banks.
    Join the conversation?

    Join our Linkedin group to get an unparalleled view of macro and microeconomic events and trends from a bank that is a leader in both developed and emerging markets.