Friday, February 27, 2015

All Are Not Equal

All risks are not equal.  Some opportunities have bigger potential than others, and some losses are potentially more disastrous than others.  Treating all risks with the equal amount of time, energy and effort is not smart, nor is it effective.  All are not equal.

Wednesday, February 25, 2015

Risk Curation

We live in a time when we are awash with information.  The internet, our Twitter feeds, our email, our Facebook, our Linked-In feeds etc. etc. bombard us with a non-stop information torrent.  So much so that we intentionally ignore so much of our information flow that we actually might be receiving less information than we did in the pre-internet / pre-social media era.  Additionally so much of the information we receive is incomplete, biased, misleading, half-baked or simply wrong.  Many media pundits argue that the new role of respected media is not news reporting but news curation – sorting the wheat from the chaff and presenting the stream of information in the most useful manner possible.
In today’s business environment, the manager is also awash in an information torrent.  In certain industries such as financial services, a lot of the flow is risk based.  Without proper curation the manager is at risk for actually being less informed about the relevant risks of their unit than they were thirty years ago when such information was much rarer and much harder to sort.  Thus a key role of the risk department is to act as the curator of risk information. 
Risk information is key to good risk management, but it is the right information at the right time that is necessary.

Monday, February 23, 2015

Man and Machine

Big data is everywhere, and risk management is no exception.  Big data in a nutshell is nothing more than allowing computers to crunch numbers to see patterns that might escape the human.  It is data analysis that may or may not have been done by a human, now being done by computer.  Of course lots of other tasks formerly done by humans are being done by computers or by robots.  Indeed, many tasks in risk management are now done at least partially by computer – for instance credit analysis, fraud detection, human resources, stress testing, etc. etc.  This is not news, but the spread of computers into white-collar tasks is quite threatening to many.  However it must be realized, particularly in risk management, that there are things that a human can do, that a machine cannot do.  It is not man or machine, but man and machine. 

Friday, February 20, 2015

Warning Pollution

Almost everyone has heard the story of the boy who cried wolf.  The Sheppard boy repeatedly cries “wolf” and each time the townsfolk come running to save the herd from the wolf – but each time it is a false alarm and there is no wolf.  Finally a wolf does attack the sheep herd, but no one comes when the boy cries “wolf”.  The townsfolk are tired of the false warnings and thus do not take the “wolf” signal seriously.
I believe a similar thing may be happening with risk warnings.  For both legal and regulatory reasons there are so many risk warnings in our daily lives that we simply ignore all of the warnings just as the townsfolk started to ignore the cries of “wolf” from the Sheppard boy.  Warning pollution is real, and like all forms of pollution it is not good.

Wednesday, February 18, 2015

Connecting the Dots

Steve Jobs once famously said that “you cannot connect the dots going forward”.  Every risk manager at some level understands this.  The fact that the future unfolds in an unpredictable manner is in some way the reason that the discipline of risk management exists.  It is however a lot easier to connect the dots going backwards – but this exercise has some hidden traps.
Connecting the dots on what happened is a way to deconstruct a risk event and learn from it.  This (to coin a phrase from the ‘90s Martha Stewart) is a “good thing”.  However it often becomes a bad thing for several reasons.
Firstly, we are often shallow in connecting the dots going backwards.  We find a few major correlations, and we assume these correlations are the cause.  In reality it is often the much more subtle and more hidden catalysts that are the loosely connected dots that as risk managers (and regulators) we need to be looking for and learning from.  The reality is that we often superficially take the first few major dot connections as our answers.  Risk is much more nuanced than that.  (I will leave the obvious issue of correlation not necessarily being causation out of the discussion for now.)
Secondly we assume that the same dots will exist going forward.  This is frequently incorrect.  The dots, and their connections going forward are usually not the same.  Risks and how they arise evolve, adapt and change.  The same dots will not necessarily produce the same connections and in turn produce the same results.  The effect is compounded when we include the fact that we likely did too superficial of a job analyzing the dot connections from the previous risk event – as discussed in the previous paragraph.
Risk management is not an “if-then” management exercise.  Fortunately, or unfortunately (depending on your point of view), risk management is not a connect-the-dots coloring book.  (I really enjoyed connect-the-dot coloring books as a kid.)