The Misconstrued Likelihood

Source: Lancashire Resiliency Forum

Source: Lancashire Resiliency Forum

 

Have you ever thought of stopping the use of “likelihood” in preparing a risk matrix? The shocked reaction is – “how can we calculate risk without likelihood?” But seriously, how competent are we in calculating the probability of each risk. As risk managers, don’t we just check the box based on our judgment? The thought process is – earthquake – rare, hurricane – rare, data theft – occasional, and we don’t need data to make these judgments.

 1. Inaccurate Calculation

My claim is that all this is hyperbole and we draw inferences from inaccurate information. To substantiate my argument, here are two statements of the EY 13 Global Fraud Survey 2014 and Kroll Global Fraud Survey 2013/2014.

EY 13 Global Fraud Survey 2014 quote:–

“More than 1 in 10 executives surveyed reported their company as having experienced a significant fraud in the past two years. In fact, the level of fraud reported by respondents has remained largely unchanged over the past six years: from 13% in 2008 to 12% in 2014.”

 Kroll Global Fraud Survey 2013/2014 quote:

 “The incidence of fraud has increased. Overall, 70% of companies reported suffering from at least one type of fraud in the last year, up from 61% in the previous poll”

The EY report does not define “significant fraud” .Kroll report does state that “the economic cost of these crimes mounted, increasing from an average of 0.9% of revenue to 1.4%, with one in 10 businesses reporting a cost of more than 4% of revenue.”

 Now assume you do not have historical data on incidence of fraud in your organization and have to infer the likelihood of fraud from the above-mentioned statements.

 

Please share the logic you used to determine the likelihood in the comments section.

 2. Unidentified Representative Bias

Implicit in our judgment is representative bias, which only a discerning eye can decipher. For instance, read the following lines from the EY 13 Global Fraud Survey 2014.

“The survey results show a correlation between executive roles and willingness to justify certain activity when under pressure to meet financial targets:

CFOs are more likely than other executives to justify changes to assumptions relating to valuations and reserves in order to meet financial targets.

General counsel are more likely than other executives to justify backdating contracts in order to meet financial targets.

► Sales and marketing executives are more likely than other executives to justify introducing flexible return policies in order to meet financial targets.”

How is this news worth reporting? Aren’t risk managers aware that employees are more likely to conduct frauds within their area of job responsibility and authority?

It would be interesting to know the probability of other departments (excluding sales and marketing personal) introducing fraudulent flexible return policies. Without that information, while conducting a fraud investigation we are likely to assume the fraud in sales department was conducted by sales personnel, whereas it is possible that another department personnel had done it.

Now if you want further proof of representative heuristic, here is a classic example of a study conducted on women’s propensity to conduct fraud by Steffensmeler. He concluded:

“There is reason to believe that over time increasing the number of female CEOs would reduce corporate corruption because women tend to promote a more ethical business climate rather than one that promotes personal and corporate profits at all costs, no matter what the potential societal costs or harms might be.”

Then he further states that lower rate of fraud might be because men do not conspire with women to conduct frauds and women may not have access to higher echelons of management to do big frauds.

However, it still does not explain how he has made the above statements. According to child psychology reports, both girls and boys in childhood have nearly equal tendency towards anti-social behavior though it reflects in different ways. For example, boys bully directly, girls bully indirectly.

So, are we saying nature and nurture have less impact on girls than boys because they are somehow hardwired to be more ethical? Alternatively, do you think that social conceptions are at play here because women are the weaker sex and therefore nicer. Wouldn’t it be interesting to study the tendency to commit fraud by giving equal opportunity to both genders to steal without fear of punishment and then find who is more likely to do so? It might show that women commit less fraud not because they are more ethical, but more fearful.

Closing Thoughts

Risk managers must ask themselves – “What is the worst that can happen if they do not check any box of likelihood? It is possible to create a bucket list of known risks, with undetermined likelihood and impact?” Adopt an alternative method or procedure, since inaccurate calculations lead to misguiding the management and implementation of wrong risk mitigation plans.

If we do not know something, why pretend to have a magic wand and claim knowledge. What is the harm in admitting that we do not have all the answers?

 

References:

  1.  EY 13 Global Fraud Survey 2014
  2. Kroll Global Fraud Survey 2013/2014
  3. Women still less likely to commit corporate fraud 

 

 

 

 

The Unreliable Expert Advisers

Let me start by asking you a question – How many times have you written the word “expert” in your resume. Were you, like me, obsessed about becoming a subject matter expert or a thought leader? Are you an ardent devotee of people who profess expertise and give expert opinion in the media?

1.  Experts Know Better

If the answer to the above questions is yes, then let me ask you another question. If a political scientist and an astrologer are giving predictions of the political scenario in India 20 years from now, whose judgment will you rely on?

Philip Tetlock studied 300 experts over a 20-year period and concluded that experts were just slightly better off than the novices were. The expert group consisted of economists, political scientists, academicians, journalists and Phd. Another study showed that a rat was able to predict the location of food in a cage better than Yale students were.

Started believing in astrologers yet?

2.  Experts Perform Better

The counter argument is Malcolm Gladwell’s study, that to become an expert 10,000 hours of practice is required while continuously enhancing the skills to become an expert. Against this backdrop, one would assume that expertise matters and experts would perform better.

Below is a question, choose one of the options.

 

A study conducted by Eric Schwitzgebel showed that old and rare ethics books in the library were missing at twice the rate of other subject’s old books. That meant ethics philosophers were more likely to steal books than other lecturers were. Their expert knowledge on ethics did not prevent them from doing something completely unethical when it was in their self-interest.

Hence, all the knowledge and teaching of ethics would go waste if it were not incorporated in the behavior and personal value system. Does it not raise the question in your mind as to what is the point of teaching ethics in the organization? Maybe it is worth conducting an organization survey to measure the results.

 3.   Risks Can Be Accurately Predicted

After the financial crises, the bamboozled regulators were grappling to figure out why the idyllic scenario collapsed. All the posturing and ping-pong battles by the self-proclaimed risk management experts could not explain how they were caught napping.

The reason is simple. Risk managers and business managers can only predict risks based on their experience and information. For instance, in David versus Goliath battle, Goliath was well prepared. He was strong; he had a shield and a headgear to protect him. However, he was not aware that a slingshot could kill him. Hence, he did not mitigate that risk.

The risk managers and business managers do not have complete information about the present and the future business environment. Business managers normally take decisions when 70% of the information is available. Complete information is not available and the future is unknown. Hence, the next big disaster is always waiting at the next blind turn.

4.  High Performers Take Lessor Risks

Another assumption is that under-performing managers who are at the risk of missing their targets take higher risks than high-performing managers do. A study conducted by Ping Hu on mutual fund managers showed that both low and high performing managers are likely to take higher risks. The high performing managers do not face any employment risks, are comfortable in their positions; hence, take more risks. The under-performing managers take risks to achieve their goals. Hence, risk taking of managers is a U-shaped curve and highly dependent on the employment risks of the manager.

Therefore, the question arises what is the benefit of all the training given to managers. It allows managers to do proper risk mitigation and enables them to take higher risks.

Closing Thoughts

The above-mentioned studies shatter the fallacies and assumptions of confidently relying of expertise. An expert’s opinion may not be worth the paper it is written. So what is the advantage of having trusted advisers within the company?

The key is not to rely on insiders only. Have a panel of outside experts to give impartial view. Moreover, the experts shouldn’t be attached to their own opinion or view. The higher the levels of constructive confrontation within the organization, the better are the chances of doing effective risk management. Additionally, incorporate risk management processes within the business processes, and integrate them in employee behavior, practices, and reward system.

References:

  1. Philip Tetlock’s book, “Expert Political Judgment: How Good Is It? How Can We Know?”
  2. Do Ethicists Steal More Books? – Eric Schwitzgebel, Department of Philosophy, University of California at Riverside
  3. Fund Flows, Performance, Managerial Career Concerns, and Risk Taking – Ping Hu Risk Analytics, Corporate Finance, Wells Fargo, Charlotte, North Carolina

 

 

 

 

Junk The Risk Assessments

Sorry folks for taking such a long break from blogging.  I was busy talking to a few angels who had entered my life all of a sudden. Now you are thinking that maybe I injured my head during the last five months. An adult talking about angels, absolutely insane! As kids we are happy to believe in Santa Claus. As we grow the social norms expect us to be more cynical, and we have to say – “We don’t believe in angels”. The question is –“have you seen any with your naked eye?” Off course not, but how does that prove that they don’t exist. In life, we have not seen many things, but we believe they exist.

So now, you are wondering what I am getting at.

As a risk manager has a business head ever told you – “You don’t have any idea of the business, this risk assessment is trash.” You wished to tell him that you did a proper job but he is absolutely is absolutely refusing to listen.

When business managers  submitted self-risk assessments, were you rubbing your eyes in disbelief? You could not figure out how they have rated the risks so high or so low, completely contrary to your expectations.

Is it possible that the risk assessments are frequently wrong and serve very little purpose except for completing the paper work? The idea of discarding risk assessments is scary as operational risk managers rely heavily on risk assessments matrix to assess the probability of occurrence of risk and the impact of the same. We advise business managers to complete self-risk assessments for their units. Organizations consider top twenty risks critical and depute resources to address the same.

Despite the risk assessments, unknown risks keep popping up. Risks rated low flare up into big issues. High impact low probability risks cause a whole lot of more damage than estimated.

Research on cognitive biases shows that subjective risk assessment done without data are prone to errors. Human beings have numerous biases in their thinking, due to which they tend to make incorrect decisions. Below is the list of biases I shall discuss in the next few posts:

1)      Representative Heuristic

2)      Availability

3)      Hindsight Bias

4)      Black Swans

5)      Conjunction Fallacy

6)      Confirmation Bias

7)      Anchoring, adjustment and contamination

8)      Affect Heuristic

9)      Scope Neglect

10)   Calibration and Overconfidence

11)   Bystander Apathy

You might be wondering whether the biases and heuristics really have any impact or is it just another aspect of psychology we can ignore. Let me ask you a question here:

 

Malcolm Gladwell did an analysis in his book David versus Goliath and stated that in 63% of the cases the smaller country defending its territories won the war. The powerful invader had to backtrack and generally lost the war despite its military strength. The small defending countries win because they use unconventional strategies for warfare, garner public support, and have higher commitment as they have more at stake if they lose. Then what percentage of the risk assessments of a war are incorrect? The loss of life and property are in vain.

Wait for the next few posts, as they might make you rethink on the conventional wisdom of risk assessment done by organizations.  

References: 1. Cognitive Biases Potentially Affecting Judgment of Global Risks – Eliezer Yudkowsky, Machine Intelligence Research Institute

2.  Probabilistic Reasoning by Amos Tversky and Daniel Kahneman