| The shocking financial consequences of how we think | | | | business unit, the more visible is the variability; groups |
| Niall Ferguson's best-selling and televised "Ascent of | | | | of partially-related businesses can appear easier to |
| Money" covers beautifully the evolution of the | | | | predict just because of the averaging of different |
| financial system from ancient Mesopotamia to today. | | | | under- and over-performing units. We often see |
| It is a superb book that relates the crucial role of | | | | business plans that overall are at or slightly below |
| financial tools in the growth and decline of empires | | | | target, but consist of component businesses that |
| and dynasties. Ironically, in amongst this excellence, | | | | show enormous variations from the original |
| the chapter that resonates most strongly is the | | | | projections. For some reason, we as managers |
| afterword. This is called, suitably and | | | | believe in our ability to perform within a tight range |
| contemporaneously, The Descent of Money. | | | | of projected expectations, despite this consistent |
| The theme of that chapter, and this article, is how | | | | evidence to the contrary. |
| our hard-wired thought processes cause us to make | | | | Trap 8: "The fallacy of conjunction" |
| decisions with a mindset that was useful for our | | | | This is a trap in which people overestimate the |
| evolution, but that in business and finance is | | | | likelihood that a series of highly likely events will all |
| destructive and irrational. He lists a series of cognitive | | | | occur, and conversely underestimate the likelihood |
| traps - ways in which we make poor decisions | | | | that at least one of a series of unlikely events will |
| without realising it. | | | | occur. |
| We at Latitude recognised every single one of these | | | | This leads management to believe that its mid-case |
| traps, both in ourselves and in the companies we | | | | scenario (which consists of all those highly likely |
| support and review. We could also relate to the | | | | events) is much more likely to happen that it actually |
| considerable damage that each one could cause if | | | | is. The corollary is that it is reasonably likely that at |
| unchecked. | | | | least one of the many highly improbable, left-field, |
| We illustrate some of these common traps in this | | | | events will occur, and management will |
| article. The terminology is complex, but the ideas are | | | | correspondingly be less likely to be ready for it. |
| simple and you will recognise every one. Even | | | | We see this fallacy most often, again, in business |
| becoming aware that they exist should help avoid | | | | planning, where a great deal of thought and |
| their destructive consequences. In the second half of | | | | preparation is given to the central scenario in the |
| this article, we attempt to go one step further in | | | | business plan, which from historic experience very |
| helping to steer clear of trouble and distress by | | | | rarely turns out to be true. It is why we at Latitude |
| proposing two well-tested answers that have been | | | | see business planning as a helpful process to prepare |
| used by science and sport from their outset. | | | | for possible futures, but see business plans as simply |
| Trap 1: "Extending the present" | | | | a means to this end. |
| In this trap, the individual assumes that the present is | | | | Trap 9: "Failure of invariance" |
| good guide to the future; much better than | | | | This trap recognises that people are risk averse |
| examination of previous experience illustrates. | | | | when prospects are positive but risk-seeking when |
| We see this at its most common in business planning | | | | they are negative. In a famous experiment, the vast |
| where future revenues and costs are based on the | | | | majority of participants preferred a 100% chance of |
| present, plus or minus a small percentage. The reliable | | | | winning 500 pounds versus a 50% chance of winning |
| rule that we apply to such plans is that they will | | | | 1,000 pounds. The same group preferred a 50% |
| usually be wrong, though the future profit out-turn | | | | chance of losing 1,000 pounds versus a 100% chance |
| may end up being more-or-less the same with good | | | | of losing 500 pounds. |
| management and a little luck. The alternative to | | | | Companies approaching distress or who have |
| extending the present - acknowledging that we are | | | | experienced the initial failings of an investment seem |
| much less knowledgeable about the future than we | | | | to follow this risk-seeking tendency by trying ever |
| think we are - is a more uncomfortable state of | | | | more unlikely approaches to getting back their original |
| affairs; but this more realistic mindset can lead us to | | | | money. It is almost always more rational and |
| adopt valuable approaches such as scenario planning, | | | | loss-minimising to write off sunk cost or a percent of |
| which make us much readier for when the | | | | equity, and to look at each decision on its own merits |
| unforeseen does happen. | | | | without the need to regain lost ground. Using share |
| Trap 2: "Hindsight bias" | | | | options as a reward mechanism can exacerbate this |
| Hindsight Bias is the trap that causes people to | | | | problem by actually making the risk-seeking rational |
| attach greater probabilities to events after they | | | | for the individual manager, incentivising to act against |
| happened than they did before they happened. | | | | the best interests of other stakeholders. |
| Whereas in Extending the Present, people assume | | | | The other side of this coin, risk aversion when the |
| the present is a better guide to the future than it | | | | company is ahead, is also very common and can lead |
| actually is, with Hindsight Bias, people over-rate the | | | | to tremendous lost opportunity in new areas of |
| past as a reliable guide to the future. | | | | business. This can be such a strong mindset in the |
| When we attempt to learn lessons from the past, | | | | team that created the company's success that |
| we must therefore make sure we cover the failures | | | | changing a winning team can sometimes be the only |
| as well as the successes. For example, we can | | | | solution when seeking continued growth. |
| blithely look at many successful companies and | | | | Trap 10: "Bystander apathy" |
| conclude that they were much more focused in the | | | | In this trap, people abdicate individual responsibility |
| services they offered than their more mediocre | | | | when they are in a crowd. Sometimes it is the |
| counterparts; that service or product focus is a | | | | apathy that stops anyone in a large crowd stopping a |
| pre-requisite for success in all market-leading | | | | mugging; sometimes it is abdicating individual |
| companies. We could use this to conclude that we | | | | judgement to the perceived wisdom of the crowd. It |
| should rid ourselves of all products, services or skills | | | | seems that the risk of taking the contrarian path and |
| except those that are part of this single core. | | | | being wrong is worse than being the anonymous |
| However, if we look at the failures, we see that | | | | lemming going over the cliff with all the others. |
| many of those companies were also very focused, | | | | We see this in the various fads and booms that we |
| but the successful ones were the minority that just | | | | fail to understand but cannot afford to miss out on. |
| happened to focus on the right thing. Looking at the | | | | The recent "arbitrage" profits experienced in the |
| full set of information, we would conclude that focus | | | | world of private equity from ever growing P/E ratios |
| with no contingency plan is a high risk strategy, which | | | | is an example. Every Investment Director we spoke |
| more often than not will fail. | | | | to when we surveyed them about this in 2006 knew |
| Hindsight Bias also leads us to project forward | | | | that the P/E growth would need to stop at some |
| assuming that the models and mechanisms that | | | | stage, and admitted to stretching beyond |
| worked in the past have a high probability of working | | | | managements' business plans to make the |
| in the future. We forget, or don't realise, that what | | | | investment case for purchase. But no-one felt they |
| actually happened was the one of a myriad of | | | | could afford not to keep investing. Everyone could |
| possibilities that happened to be supported by the | | | | see problems coming, and knew what would happen |
| circumstances of the time. Stepping into the present, | | | | if they were left holding the baby when P/Es |
| those myriad possibilities still exist and the chances of | | | | inevitably started shrinking, but to stop investing was |
| the future turning out as we project are much less | | | | to step out of the game. |
| likely than we think. | | | | There are numerous other cognitive traps, such as |
| Trap 3: "Availability bias" | | | | contamination effects from irrelevant data and scope |
| This trap causes people to base decisions on | | | | neglect where we don't minimise harm; but you |
| information that is to hand, usually in their memories, | | | | probably already get the drift - people aren't as |
| versus the information that they actually need, like | | | | rational as they think they are and they make |
| the car driver who loses his keys at night and only | | | | irrational and potentially harmful decisions without |
| looks for them under lamp posts. | | | | realising it. |
| We see this at its most dangerous in Board or | | | | Two old-fashioned ways to overcome cognitive |
| management workshops where the day is being run | | | | traps |
| on the basis that all of the important knowledge is in | | | | In the sections above, we described ten "cognitive |
| the room. We have even heard facilitators use this | | | | traps": ways in which we think that can cause us to |
| we-have-everything-in-our-heads-already as a key | | | | make damaging decisions without realising it. |
| premise for the entire strategy that emerges. | | | | The good news is that there are some decent tools |
| This cognitive trap also biases us to recency and | | | | to challenge these traps. We explain our two |
| proximity - we don't look back far enough for similar | | | | favourite approaches below, as lessons from science |
| patterns or warning signs, and we don't look far | | | | and sport. |
| afield enough for analogous evidence of failure or | | | | 1. A lesson from science - treat your beliefs as a |
| success. | | | | hypothesis to be challenged |
| Trap 4: "Confirmation bias" | | | | True science is not about test-tubes, double-blind |
| This trap causes people look for evidence to prove | | | | tests and professors with moon-shaped glasses and |
| what they believe to be true, rather than looking for | | | | speech impediments. It is about starting with a |
| evidence to challenge it: why Tories read the | | | | premise that you believe may be true - a hypothesis |
| Telegraph and Socialists read the Guardian. | | | | - and challenging it to see if you are right, or more |
| We see this bias at its most damaging in investment | | | | likely, where you are wrong. Under this definition, you |
| cases for acquisitions and in business cases for | | | | are more likely to see science from a good plumber |
| investment of money and time into new ventures or | | | | trying to work out why your central heating makes a |
| projects. Even when employing a third party | | | | knocking noise than you are from a PhD nutritionist |
| professional to assess the acquisition, venture or | | | | with research sponsored by High5, trying to persuade |
| project, the investor or business manager will actually | | | | you that High5 is better than Powerade. The plumber |
| ask for affirmation or substantiation - "I'm just | | | | is the scientist, challenging his hypothesis in search of |
| looking for confirmation of my hypothesis" - rather | | | | the truth; the nutritionist is no more than a |
| than "Challenge me and tell me where I'm wrong". | | | | fundamentalist seeking and selecting evidence to |
| Trap 5: "The affect heuristic" | | | | support his initial position. Unfortunately, when we get |
| This trap causes people to allow their beliefs and | | | | attached to our ideas, the cognitive traps make us |
| value judgements to interfere with a rational | | | | act more like the nutritionist than the plumber. |
| assessment of costs and benefits. | | | | This is why the true scientist needs to adopt a |
| We find this most dangerous at either of two | | | | mindset of challenging the hypothesis with data, and |
| extremes: on the one hand where the decision | | | | to have no belief that the hypothesis is true until the |
| maker is very passionate about a subject, or on the | | | | challenges show it to be so. Some practitioners even |
| other where he is once-bitten-twice-shy. | | | | go as far as setting up a formal challenge in the form |
| In the former case, whilst we find it critical that | | | | of an antithesis, an alternative hypothesis that is |
| managers be passionate about their products or | | | | posited as a more accurate or insightful version of |
| services, this passion can blind the person to reality, | | | | reality. This approach isn't confined to the material |
| and can be impossible to address without introducing | | | | and commercial - the Catholic Church appoints a |
| a very senior individual with authority to challenge | | | | devil's advocate to provide the rigour of challenging |
| assertions with information. | | | | its most important decision, the legal system applies |
| In the once-bitten-twice-shy case, we have seen | | | | the rigours of having separate representatives of |
| private equity companies abandon entire sectors | | | | both sides of the case. |
| following one painful loss, and refuse to entertain the | | | | So, how to apply this? Treat your belief as a |
| most solid business case that shares even the | | | | hypothesis and challenge it, if necessary with your |
| remotest common characteristics of historic | | | | own devil's advocate, to whom you give the |
| loss-makers. | | | | seniority and power to challenge your decisions. And |
| Trap 6: "The problem of induction" | | | | honestly expect your hypothesis to change as the |
| Induction is the process of generating a general rule | | | | evidence emerges. |
| from a series of observations. In the absence of | | | | Applying this lesson from science stops us being blind |
| clear indisputable deductive relationships, induction can | | | | to the evidence at hand, but it doesn't help us |
| be all a person has to go on. The problem of | | | | predict a future that is much more random that we |
| induction is that the brain will look for neat patterns | | | | think it is, or stop us being over-confident in our |
| and will try to create a general rule even if it is based | | | | ability to predict it. To prepare for this, we take a |
| on insufficient information. | | | | lesson from sport. |
| In acquisitions, people can over-estimate the | | | | 2. A lesson from sport - prepare for a range of |
| performance and prospects of a company by seeing | | | | scenarios |
| how satisfied its customers are. This creates an | | | | A lesson learned by those of us who have been on |
| overly-positive pattern from what is essentially a | | | | the wrong end of a drubbing on the sports field or, |
| self-selecting group: non-customers and disgruntled | | | | more seriously, have experienced military action, is |
| ex-customers need including for the full picture. | | | | that no plan survives contact with the enemy. |
| Another example of poor induction is where | | | | Whether you are a batsman facing a spin bowler |
| management projects forward on the basis of a new | | | | about to treat you to one of his box of tricks, or a |
| product's first year's sales and forget to consider | | | | tennis player trying to decide if your opponent is |
| that this first year was a golden year, where | | | | stretched enough for you to approach the net |
| everyone without the product bought one and would | | | | without being passed or lobbed, you have to be able |
| never need another. | | | | to cope with a range of scenarios. It doesn't mean |
| A very common area where induction knows no | | | | that your core game plan needs to be dictated by |
| bounds is in the practice of regression: correlating one | | | | the opponent and environment, but it does mean |
| factor against another to create what superficially | | | | that you need to be prepared for the range of |
| appears to be a causal relationship. For example, it is | | | | scenarios that might play out. If you can't deal with |
| possible to infer high price sensitivity when analysing | | | | the high ball, you can guarantee that a good |
| price-volume relationships, and miss the over-riding | | | | opponent will be sending up bombs for you to panic |
| effect of heavily marketed promotions that | | | | under all afternoon. |
| commonly coincide with lower prices. I was humbled | | | | In business, the normal corporate downside scenario |
| to the limitations of correlation when an analyst | | | | is maybe a 5% or 10% decline versus base case, |
| working for me at my former company determined | | | | which isn't really a scenario at all, but more of a |
| an almost 100% correlation between pallet demand | | | | smaller version of the base case. A more useful |
| and GDP. We started to doubt the causality when | | | | scenario is to work out how we would still thrive if |
| the analyst realised that she had used the wrong | | | | sales fell by 30% or 50%, or how we would grow if |
| source data and correlated UK pallet demand with | | | | competitive substitute product X gained critical mass. |
| Polish GDP numbers. Our confidence in the causality | | | | How would we deal with costs? Where would we still |
| was damaged further when the correlation with the | | | | invest, or even increase investment? Which divisions |
| "correct" driver, UK GDP, was about 30% lower. | | | | would we let go? What resources would we try to |
| Trap 7: "Overconfidence in calibration" | | | | acquire? What we are not doing here is trying to |
| In this trap, people under-estimate the potential | | | | create a plan for every single situation that might |
| range of possible outcomes. In particular, people are | | | | come about. What we are doing is stretching our |
| often not sufficiently pessimistic with downside | | | | thinking, in order to understand those common things |
| scenarios and/or attach too low a likelihood to major | | | | we need to do to thrive in whatever scenario might |
| problems and pitfalls. | | | | come about, and preparing ourselves to respond to |
| This issue is rife in business planning and financial | | | | the inevitable unpredictability. |
| projections. The more discrete and separate a | | | | |