Moral Hazard The Silent Assassin in Trading Organizations

1
Moral Hazard The Silent Assassin in Trading
Organizations

• 2nd Annual UH-GEMI
• Energy Trading and Marketing Conference
• Houston, Texas
• January 22, 2004

Dunham L. Cobb
dunhamcobb_at_adelphia.net
2
The 2 Sides of Risk Control
The hard side of risk control is objective and
measurable.
The soft side of risk control is subjective and
difficult to measurable.
Over the years, thousands of man-hours and tens
of millions will be spent by an organization to
detect, manage and mitigate the hard side.
3
But Headline Events are Consistently on the
Soft Side
Institution Loss Related Activities Enron
Tens of Billions Corporate Fraud (shareholder
equity) Sumitomo 2.6 Billion Copper
futures Orange County 1.7 Billion Structured
notes Barings 1.4 Billion Nikkei
futures Metallgesellschaft 1.3 Billion Oil
futures Daiwa 1.1 Billion Unauthorized
bond transactions
4
Scandal Hall of Infamy -- Other
Institution Loss Related Activities HealthSout
h - Scrushy 1.4 Billion Corporate
Fraud 11 Wall St. Firms 1.4
Billion Corporate Fraud (analyst reports)
Tyco - Koslowski 600 million Theft
- Swartz NYSE - Grasso 190 Million ?
Comp package TYCO - Lord ?? millions
Undisclosed fees Brown IM Clone -
Waksal ?? CEO Insider Trading Rite Aid -
Glass ?? CEO Insider Trading
5
But Headline Events are Consistently on the
Soft Side
Institution Loss Related Activities Enron
Tens of Billions Corporate Fraud (shareholder
equity) Sumitomo 2.6 Billion Copper
futures Orange County 1.7 Billion Structured
notes Barings 1.4 Billion Nikkei
futures Metallgesellschaft 1.3 Billion Oil
futures Daiwa 1.1 Billion Unauthorized
bond transactions
Only instance NOT due to moral hazard.
Why are resources so misallocated to the hard
side?
6

• Understanding
• Moral Hazard

7
Understanding Moral Hazard

• Why have attention and resources been so
  misallocated?
• The answer is unfortunately simple Managing the
  soft side has been too elusive.
• Risk management is rooted in the three
  Msmeasure, monitor and managebut those are
  concepts not easily applied to behavior.

However, there are cost-effective ways to
mitigate moral hazard if a company is willing to
de -mystify the motivation behind it.
8
Understanding Moral Hazard

• Definition
• In a business context, moral hazard (MH) is
• The likelihood of the willful commission of an
  act that is prohibited by the rules of an
  organization, typically found in its Policies or
  Procedures handbooks or Code of Ethics.
• In a nutshell, MH is the risk that individuals
  will, by omission or commission, intentionally
  engage in bad behavior.

9
Understanding Moral Hazard

• Why are trading organizations so fertile for
  moral hazard?
• There is a LOT of money involved.
• By their very nature, empower employees to
  transact on the companys behalf and thus tempt
  traders to abuse that power
• What solutions have been considered up to now?
• One solution to moral hazard would be to act upon
  the assertion that sometimes, the best cure for
  a disease is to kill its host.
• Since it is generally accepted that energy
  companies should have some trading capability,
  any cure that would eliminate all trading
  arguably would be worse than the disease.
• Equally impractical, listen in on all deal-maker
  conversation.
• This would drain the coffers
• A more practical approach would be to leverage
  limited risk management resources in a way that
  optimizes protection to the organization.

Yet how can a company manage something that it
cant easily detect?
10
Understanding Moral Hazard

• First thing to recognize
• The underpinning of moral hazard is behavior, and
  behavior is grounded in the axiom
• People do exactly what they are motivated to
  do.
• To manage behavior, one must model it in a way
  that reveals its driving incentives and their
  interrelationships.

11

• Modeling Moral Hazard

12
Modeling the Risk of Moral Hazard

• A model for moral hazard already exists and was
  used to teach The Economics of Corruption at
  Yale University in the 1970s

Likelihood of Being Successful Likelihood of
Being Caught
MH
Reward Penalty
Component A Risk/Reward
Component B Probability of Success
Component C Personal Inclination
The goal in managing MH is to keep the numerator
from overwhelming the denominator.
13
Modeling the Risk of Moral Hazard

• In component A (risk/reward)
• Rewards act as a catalyst in favor of an act of
  MH
• Rewards include avoidance of the unwanted.
• Penalties are the consequences of being caught --
  reprimand, fine, suspension, dismissal, and/or
  prosecution thus are a deterrents to moral
  hazard
• The reaction an individual has to a penalty is
  influenced by two factors
• Personal Moral Ethic for some individuals,
  even a consequence as light as a reprimand is so
  distasteful that they will significantly increase
  their personal weighting of the penalty, greatly
  decreasing the risk of moral hazard.
• Consequences are not real enough - Why? Most
  employees of trading organizations have not
  suffered seriously as a result of behaving badly,
  nor have any of their friends. As a result they
  subconsciously underweight the penalty.

14
Modeling the Risk of Moral Hazard

• In component B (probability of success)
• The odds of succeeding and getting caught have an
  inverse relationship.
• Any steps that make the improves detection
  clearly makes it more difficult to get away with
  it.
• Ones perception of the likelihood of success is
  increased by a perpetrators familiarity with
  procedures and IT system capabilities and
  especially their gaps and weaknesses.
• Perpetrators are usually quite clever
• The likelihood of being caught is mostly a
  function of corporate controls.
• The likelihood of being caught decreases the
  higher a perpetrators level within an
  organization because there has traditionally been
  far less infrastructure providing scrutiny and,
  lets face it

They make the rules and face it, how many
whistle blowers are willing to sound off when the
police chief is the crook?
15
Modeling the Risk of Moral Hazard

• In component C (individual inclinations), two
  factors do battle -- a persons greed/need and
  moral ethic.
• Greed is universal, but more restrained in some
  people than in others. Need is also universal,
  but its urgency can be influenced by outside
  pressures like family tragedy or excessive debt.
• Balancing the forces of greed/need is the
  persons personal moral ethic, the wild card of
  the model because its range of influence can be
  so wide. For some people
• A persons word is their bond is absolute.
• For others, it is absolute drivel!

For most of us, the influence of personal
integrity falls somewhere in between.
16
Modeling the Risk of Moral Hazard

• Factors in the model have wider ranges of
  influence on ones behavior than others. The
  tornado-chart below provides a sense of the
  distribution to which these factors show up in
  a population.

Deterrent to Moral Hazard
Incentive for Moral Hazard
Personal Moral Ethic Reward (risk to
organization) Urgency of Need / Greed Penalty (if
caught) Likelihood of Being Caught Likelihood of
Being Successful
Essentially, if the lure of the prize exceeds the
sum of the other factors for a given individual,
they are likely to succumb
17
Modeling the Risk of Moral Hazard

• Some factors have impact on others
• Personal Moral Ethic has the greatest effect,
  impacting both ones Urgency of Greed as well
  as the weighting one puts on Penalty
• Likelihood of success has an inverse
  relationship to Likelihood of being caught

Component A Risk/Reward
Component B Probability of Success
Component C Personal Inclination
18
Modeling the Risk of Moral Hazard

• The factors of MH have differing degrees of
  influence on improper behavior.

Likelihood of Being Successful Likelihood of
Being Caught
? Penalty
Component A Risk/Reward
Component B Probability of Success
Component C Personal Inclination
19
Modeling the Risk of Moral Hazard

• The factors of MH have differing degrees of
  influence on improper behavior.
• Unfortunately, the transparency of these
  influences are inverse to their impact on
  behavior.

Transparency
Likelihood of Being Successful Likelihood of
Being Caught
? Penalty
Component A Risk/Reward
Component B Probability of Success
Component C Personal Inclination
20

• Managing Moral Hazard
• Where are we now?

21
Modeling the Risk of Moral Hazard
Odds of Improper Activity Occurring No
constraints
HIGH
D MH / D Reward
LOW
PILFERAGE
PILLAGE
Perpetrators Reward / Victims Risk
22
Modeling the Risk of Moral Hazard
Odds of Improper Activity Occurring Add Controls
and Penalties
Resources Spent in Prevention/ Penalty if Caught
HIGH
LOW
PILFERAGE
PILLAGE
Perpetrators Reward / Victims Risk
23
Modeling the Risk of Moral Hazard
Odds of Improper Activity Occurring More
Realistic
HIGH
LOW
Direct Performance Bonused Trader
C-Level Executive
Regular Trader
JOB LEVEL
Save Job
Bigger Bonus
Wealth, Power, Prestige
INCENTIVE
PILLAGE
PILFERAGE
Perpetrators Reward / Organizations Risk
24

• Managing Moral Hazard
• Is Sarbanes-Oxley
• the Answer?

25
Moral Hazard and Sarbanes-Oxley

• Five Components of COSO Internal Control
  Framework
• Control Environment
• Risk Assessment
• Control Activities
• Information and Communication
• Monitoring

26
Moral Hazard and Sarbanes-Oxley

• Five Components of COSO Internal Control
  Framework
• Control Environment
• Risk Assessment
• Control Activities
• Information and Communication
• Monitoring

Soft Side
Hard Side - Section 302 - Section 404
27

• Managing Moral Hazard
• Practical Steps

28
Managing Moral Hazard Practical Steps

• People do exactly what they are motivated to
  do.
• Its easier to ride a horse in the direction it
  is going.
• You cant herd cats but you can influence them
  with tuna fish

Moral Make DARN sure that what you want people
to do is aligned with what they are motivated to
do!
29
Modeling the Risk of Moral Hazard

• Which factors can we directly impact or
  indirectly influence?
• In what order should we attack these factors to
  achieve maximum leverage?

Deterrent to Moral Hazard
Incentive for Moral Hazard
1
Personal Moral Ethic Reward (risk to
organization) Urgency of Need / Greed Penalty (if
caught) Likelihood of Being Caught Likelihood of
Being Successful
30
Methods to Reduce Moral Hazard 1

• Do a better job of screening candidates for
  employment
• Companies should make a concerted effort to avoid
  hiring untrustworthy individuals
• A major financial company has statistically
  demonstrated that inaccuracies on
  resumesexaggerations or liesare very reliable
  indicators of a low moral ethic.
• Hold existing employees to a zero-tolerance
  standard on tell the truth
• There is never, ever a valid reason for an
  employee to withhold or be evasive about
  information -- much less overtly lie.
• But an organization must demonstrate reasonable
  tolerance for errors.

Anyone who avoids the truth when its not
important is guaranteed to do so when it is
important.
31
Modeling the Risk of Moral Hazard

• Which factors can we directly impact or
  indirectly influence?
• In what order should we attack these factors to
  achieve maximum leverage?

Deterrent to Moral Hazard
Incentive for Moral Hazard
Personal Moral Ethic Reward (risk to
organization) Urgency of Need / Greed Penalty (if
caught) Likelihood of Being Caught Likelihood of
Being Successful
2
32
Methods to Reduce Moral Hazard 2

• Make penalties seem less remote.
• The penalties listed in an employees handbook
  are not likely to change, but the employee can be
  made to take them more seriously.
• Do something improper and we will catch you
  maybe not today, maybe not even this month, but
  we will catch you. And when we do, you can count
  on the following you will be dismissed, if
  warranted you will be prosecuted, and to the
  extent we legally can, we will make it difficult
  for you to work anywhere else in the industry.
• Enforce the penalties you have.
• Give Risk Managers real power and authority!!
• Front-officentricity is a prevalent malaise
  within many organizations. Because trading is
  the revenue side the bread and butter of an
  organization, policies are often ignored.
• A wink and a nod by management is an invitation
  to subordinates to view rules as milestones to be
  passed on the way to an end justifies the means
  goal.
• Industry oversight with sharper teeth.
• The energy industry needs a mechanismlike the
  securities industrysfor making competitors and
  authorities more aware of a miscreants misdeeds.

33
Modeling the Risk of Moral Hazard

• Which factors can we directly impact or
  indirectly influence?
• In what order should we attack these factors to
  achieve maximum leverage?

Deterrent to Moral Hazard
Incentive for Moral Hazard
Personal Moral Ethic Reward (risk to
organization) Urgency of Need / Greed Penalty (if
caught) Likelihood of Being Caught Likelihood of
Being Successful
3
34
Methods to Reduce Moral Hazard 3

• Decrease errors
• Most acts of moral hazard start with a human
  error that leads the perpetrator to believe,
  fearing repercussions, that there is nothing to
  lose by trying to hide their mistake.
• Rarely are such individuals driven by visions of
  Ferraris or Caribbean villas, nor are they
  typically villainous.
• They are average people motivated by the fear of
  losing their job.
• Every error that goes undetected through the deal
  capture and confirmation process is a seed for
  moral hazard. Detecting and fixing these errors
  quickly nips incentives for MH in the bud.
• Death by a thousand cuts
• But even errors that do NOT lead to moral hazard
  are expensive
• Some organizations have estimated that cumulative
  cost of small errors can go as high 0.25 of
  revenues 2.5 million per billion.
• Rapid error detection decreases
• Prior Period Adjustments
• Trade disputes
• Accounting staff

35
Modeling the Risk of Moral Hazard

• Which factors can we directly impact or
  indirectly influence?
• In what order should we attack these factors to
  achieve maximum leverage?

Deterrent to Moral Hazard
Incentive for Moral Hazard
Personal Moral Ethic Reward (risk to
organization) Urgency of Need / Greed Penalty (if
caught) Likelihood of Being Caught Likelihood of
Being Successful
4
36
Methods to Reduce Moral Hazard 4

• Sufficient risk management resources,
• Senior management must determine the amount of
  risk they are really willing to accept and
  resource Risk Managers accordingly
• Single Version of the Truth. Invest in IT data
  management so that the organizations business is
  not being run from numerous un-audited
  spreadsheets.
• Easy to hide trouble in a sea of spreadsheets
• allocated better.
• Be smart with your resources as they are finite.
  Pay closer attention to the activity of those
  most motivated to misbehave (e.g., a trader whose
  bonus is based on trading profits).
• Streamline down to policies and procedures that
  work and stick to them
• They should be written in concert with the front
  office as their ultimate purpose is to increase
  bottom line.
• Poorly conceived procedures can promote less
  compliance!
• Substance over form always
• If a rule doesnt work, modify or get rid of it.

37
Methods to Reduce Moral Hazard Others

• Make sure incentives/objectives are not in
  conflict across an organization
• This is an unbelievably prevalent condition and
  begs people to act improperly
• Data quality
• Most organizations do not have a clue how good
  their data particularly deal data is.
• Completeness and accuracy
• Many organizations are more interested in timely
  risk reports than accurate risk reports.
• Senior Management often sets earnings objectives
  at unrealistic levels that almost force trading
  organizations to break the rules

38

• Managing Moral Hazard
• Getting Started

39
Moral Hazard Risk Assessment

• Perform a Rip-Off Assessment
• Looks for gaps/conflicts in
• Trading policies and process
• HR hiring and firing policy
• Planning and Budgeting
• Errors
• Systems
• Data
• Reporting
• Skills
• Governance
• Reporting lines
• Incentives / Objectives
• Compensation

40
Moral Hazard Risk Assessment

• Perform a Rip-Off Assessment
• Output Actionable
• List of Moral Hazard Flash-points
• People, process, system
• Flashpoint are ranked by
• Frequency
• Impact
• Cost of mitigation
• Business case
• Suggested remediation plan based on your own risk
  tolerance and resources to be committed.
• Many quick-hits are very inexpensive
• Ancillary benefits
• Fewer trade disputes with counterparties
• Fewer Prior Period Adjustments (PPAs)
• Lower overall accounting costs

41
Moral Hazard Risk Assessment

• Why use my group?
• To quote a member of the Association of
  Certified Fraud Examiners (ACFE)
• A person with experience is never at the mercy
  of a person with a theory. To catch a thief you
  must be able to think like one. You dont make a
  person a lead detective just because they
  graduated top-of-the-class from criminology
  school.
• We use people with both trading and risk
  management backgrounds and the battle scars to
  prove it.
• Every one of us understands the behavior-drivers
  for MH and how to influence them.
• We all know how to game the system.

42
Is it worth it to NOT address MH?
Institution Loss Impact to
Reputation? Enron Tens of Billions Priceless
Sumitomo 2.6 Billion Priceless Orange
County 1.7 Billion Priceless Barings 1.4
Billion Priceless Daiwa 1.1
Billion Priceless HealthSouth - Scrushy 1.4
Billion Priceless 11 Wall St. Firms 1.4
Billion Priceless Tyco - Koslowski
600 million Priceless - Swartz
you decide.