Big data is a hot topic in business right now, and the suggestion is that it will help us to overcome the decision making biases that so afflict us. The promise of wiser, more rational decisions however rests on our willingness to actually listen to the data.
A recent study highlights how forlorn this hope may be. It reveals how when making investment related decisions, executives frequently ignore data in favor of decisions based upon gut instinct.
The paper, written by a German team and published in the Journal of Finance reveals what really underpins executive decision making.
“CEOs systematically put too much money in projects with high potential upside”, the authors say. “Many CEOs themselves say that ‘gut feeling’ is important for their investment decisions. The problem is, that there is by now overwhelming evidence from psychology and economics suggesting that intuitive reasoning in financial matters frequently leads to biased and therefore suboptimal decisions. Our paper shows that investment decisions biased towards long shots may indeed be a serious problem in many firms.”
How we decide
The paper saw around 1,000 conglomerates studied over a 20 year period to try and understand how they invest and which segments they invest in. These decisions were compared against rivals in those segments who specialized in that area and did nothing else.
The authors suggest that conglomerates are more likely to choose projects with an asymmetrical division of chances. In other words, the projects with lots of potential (no matter how remote).
For example, if an executive was given a choice between two projects:
- Project A, which offers a 50% chance of a 5% payoff, and 50% chance of an 8% payoff
- Project B, which offers a 75% chance of a 4% payoff, but a 25% chance of a 12% payoff
Despite project A offering, on average, a better chance of a good return, most executives chose project B.
Why are our decisions irrational?
The authors suggest that the prime culprit for this poor level of decision making is the CEO themselves, and especially among younger CEOs who they found were more inclined to go for long-shots.
There is also a tendency towards this form of decision making in companies with weak corporate governance, and also in ones whereby the CEO is extremely powerful.
I’ve written previously about the rise of automated traders that attempt to remove some of this irrationality from investment decisions. I suspect it would take a brave person to suggest the final say should be taken away from the boss and given to an algorithm however.
This doesn't surprise me at all. After all, their enormous salaries are predicated on them having supernatural abilities to make the right decisions. If an algorithm can do it better, we wouldn't need to pay them so much.
To assume otherwise would assume that decision making is a rational exercise, which I suspect is far from the case.
Not surprising when their compensaon is skewed towards taking higher risks with stock options and bonuse , etc. that reward the upside but doesn’t penalize the downside.
"Today’s CEOs face a host of complex business challenges, including how to harness disparate streams of data to generate pragmatic insights and deliver increased profitability. The difficulty CEOs face is not in obtaining the data, but in determining how to effectively use it in order to make intelligent business decisions and gain a competitive advantage in the marketplace. The challenge of harnessing the power of Big Data, as well as steps to unlock key insights to accelerate business performance, were the focus of a recent roundtable discussion at the 30th Annual Chief Executive of the Year Gala Reception"
http://www.marketwatch.com/story/the-data-enabled…
Good find Paul, thanks for sharing.
Probably! My (substantial) gut says yes!
Data still needs interpretation so the decision-maker role remains important. CEOs are undoubtedly missing a trick, however, if they aren't analysing and using the huge datasets that we previously didn't have computing power to exploit.
Data cannot be ignored unless it's a fraud.
Lets face it, business is seldom very rational, is it? Maybe the problem is that we ever expect it to be.
Aye, fair point Bill.
Data sometimes needs to be ignored. Data is past tense, you can speculate the future with data but your still are not 100 percent sure. It really depends on the situation.
Good point Jamal, and knowing when to trust and when to ignore is a key skill to possess.
I've seen plenty of examples where the data was selectively analyzed to support decisions that were already made. I've seen other examples where the data was disregarded because it ran too contrary to a ore-conceived comfort level. A big factor is the level of governance by a board. Is risk mitigation important to them, or are large returns important? What is the CEOs compensation and bonus structure set up to support? CEOs that face little downside risk are more likely to pursue larger returns, regardless of the amount of data available.
What about using datas only to check that what has been already decided with guts is actually the right thing? Save time! And gives opportunity to stop execution of gut based decision in case data analysis eventually shows that the decision was not good
Gut decisions absolve the CEO from scrutiny. Who is going to question a CEO's gut? It also allows a CEO (or other business leader) to incorporate personal bias with no oversight. When we fall into the trap of trusting our gut or trusting our instincts, what we’re really giving validity to is our unconscious biases toward the comfortable, the similar, or the familiar.
Confirmation bias, In Group bias, and Status Quo bias all creep in and affect our decisions when we rely solely on instincts. ‘Trusting our gut’ means we’re relying on past experiences which may or may not be indicative of how the future will play out. Yet our human nature is to constantly make associations predicting the future. Managers should focus instead on measurable metrics not just an instinct, intuition, or ‘feeling.’
One of the most egregious errors made using 'gut instincts' revolves around recruiting and talent selection. EVERYONE thinks they're able to read people, especially those in upper management but nothing could be further from the truth.
Several years ago Capital One’s CEO Richard Fairbank put it this way, “At most companies, people spend 2% of their time recruiting, and 75% [of their time] managing their recruiting mistakes.” Most of those mistakes were probably made on the basis of ‘trusting our gut.’
Big Data is an excuse to be too manipulated by those who can manipulate data. The end game is that we don't need anything but the data and a computer analyzing it. It's useful, but I'll never let Big Data dominate any of my decision making in and of itself.
The CEOs in this example who make the bad decisions care more about their bonus checks than the long term health of the company.
Data should guide, not dictate. Relying solely on data will give you consistent, but not extraordinary, insights. True industry / culture / thought leaders pull away from tbe pack by having some amount of willingness to act on gut instinct
Poor decision-making is an issue. However, research shows that experts make excellent 'gut' decisions. The real problem is having non-expert board members who rely on 'data' (big or small) whose limitations they do not understand. What are those limitations? The Big 5 in my experience: placing importance on what you can easily measure rather than understanding what is really important; mistaking numbers for knowledge; mistaking statistics for science; mistaking rationalizing for rational; having no-one on your board who understands customer psychology and makes the right 'gut' decisions informed by data.
Poor decisions in the context that you present them have to do with forgoing fact-based analysis and decision-making, not with big data (or any other size of data for that matter.) The sweeping generalization that "big data" is necessary for correct and informed financial decision-making is incorrect: each problem requires a proper set of metrics, data and analytical approach; some of these – alas not all – may entail big data and its appurtenant analysis methodologies.
I think Big Data is too big to ignore. For Progressive Organizations, however, Big Data is anything but bunk. On the contrary, it represents an enormous opportunity for one simple reason : It's Optimal utilization can yield superior information and unparalleled insight into a range of behaviors and even predict a few. With that information, everyone-employees, departments, organizations, scientists & politicians-can make vastly better decisions.
Intuition and experience must work hand in hand with "Big Data" to put the common sense factor into the "equation" …the problem with BD and its associated analytics is how the analytics is fashioned to help connect the dots and thus, enable humans to gain insight (insight comes from experience and intuition not mathematical equations scampering over a data landscape seeking to discover insight). So there is and must be a strong partnership betwixt the geek world and the practitioner world in order to get full advantage from Big Data.
Very few decisions are that clear cut with risk/reward defined. In that context, a ceo rose to that rank making great decisions consistently. Luck will run out. Other business decisions can and should rely on data, but there is intuition that needs to factor intangibles like public perception, political cost, and other immediate needs.
Data is only info, so one needs to learn to go somewhere with that info to make a decision, and that comes from trusting yourself to the full.
Very good article on this here
http://gbr.pepperdine.edu/2015/10/big-data-decisi…
An excellent read, thanks for sharing Wayne.