It is now over four years since the publishing of what I personally consider a seminal book for our time, "The Black Swan" by Nassim Nicholas Taleb. In his uniquely engaging and brash manner, Taleb slices through the false confidence and human epistemic arrogance that was witnessed in the run up to the financial crisis. It is clear that in certain quarters, the risk management profession failed to curtail this epistemic arrogance and indeed failed to communicate in a sufficiently forceful and vigorous manner the risks that some banks were taking on. Taleb is particularly scathing of forecasters even going so far as to urge them to resign from their posts. As we peer into the black box of the future (which at the time of writing is largely framed by the potential outcomes of the Eurozone debt crisis) has the risk profession embraced Taleb's arguments and is it able to communicate its value by sufficiently differentiating itself from forecasters? Where does the credibility of risk modelling as a practice lie in the public perception after a financial crisis that has challenged even the most conservative of models?
It is extremely difficult to argue or indeed find fault with large parts of Taleb's arguments. We clearly cannot predict black swans such as the terrorist atrocity on September 11th and therefore are constantly surprised by the course of history. (For those not familiar with Taleb's terminology, a black swan is an unexpected outlier that carries an extreme impact). Using Taleb's terminology, the financial crisis would be a grey swan rather than a black swan. A number of commentators (including Taleb himself) warned of the dangers of a highly leveraged and concentrated banking sector and so it cannot technically be deemed a black swan though it could be argued that the magnitude of the crisis was greater than expected. (A grey swan would be an event that can be anticipated but would still have a significant impact if it occurred, or alternatively "known unknowns" courtesy of former U.S. Secretary of Defense Donald Rumsfeld). We clearly cannot project or forecast with sufficient accuracy or certainty oil prices in thirty years' time. Anybody who claims they can do so has either not tested the accuracy of their previous predictions, has not made a sufficient number of predictions or has hit lucky in the same manner as a gambler winning repeatedly at the roulette table at the casino.
https://www.wikipedia.org
In addition to Taleb's arguments, we have had senior people working in the finance industry comment publicly that they have suffered losses due to an event with a one in a million probability, which has occurred multiple times in that week. I personally think statements such as this impact negatively on everyone who cites risk management as one of their responsibilities. It helps to give credence to Taleb's inferred notion that the disadvantages of risk modelling outweigh the advantages. These statements can be classed in the same bracket as a former Prime Minister declaring that his country would never return to boom and bust. If you are serious about your profession's reputation and your industry's reputation it is completely paramount that you are able to communicate effectively the uncertainty around your models which is likely to be largely in the tails. If you fail to do this, you leave the door open for critics and sceptics to attack the credibility of your models, your profession and your organisation's understanding of risk.
We are currently living though a period of high uncertainty and in some markets, high volatility. The accuracy and robustness of many models during this time will be challenged and in some cases past models will be deemed unfit for purpose. The value of attempting to build a model to express a view of the future though is an exercise still worth pursuing. Multiple adjustments or future rebuilding of models may be required due to unexpected future events especially during uncertain times such as during sovereign debt crises. The alternative is not to construct your view of the future, not to sufficiently prepare for the future and purely be an organisation that is reactive rather than proactive. Crisis management, though in some cases necessary, can be extremely costly, time consuming and damaging to a company's reputation. To purely rely on crisis management rather than forward looking preparation is clearly the wrong strategy. It is also worth reminding that whenever we do emerge from this debt crisis, it will likely be the organisations that are able to communicate their risk strategy and utilise all the available tools for analysing risk that will be attractive to investors. In this new dawn, I suspect the risk tolerances of investors will be at the forefront of their minds and therefore should be at the forefront of management's minds.
A rather basic metaphor would be to compare risk modelling to building construction. Builders will typically construct a toy or a visual model of what they are attempting to build before they start building. Once they begin building, they will stumble upon problems and complexities that were not highlighted by the model. Plans will need to be updated and decisions will need to be made how to respond. However, the alternative would be to set out without any model in the first place and begin building from scratch. A high level view of the landscape would not have been formed. This does not strike me as prudent planning for the building project. Expressing a risk appetite and a risk tolerance without coming to a view on the current and future risk landscape also does not strike me as prudent planning for an organisation. Surely there is significant value in attempting to quantify risks and building distributions for those risks which feed into the decisions on risk appetite and risk tolerance? Taleb describes models as "potentially helpful medicines that carry random but very severe side effects". I would be stronger in my defence of models than that. I think models play a crucial role in the risk management process. They can frame an organisation's current view of the landscape before deciding on its risk appetite. However, the limitations of the model must always be at the forefront of everyone's minds. The challenge is to make our organisations robust to over-reliance on the models whilst still utilising the clear value in the process of constructing, monitoring and adjusting them.
https://exed.canvas.harvard.edu/eportfolios/3033/Dumps_PDF/How_Benificial_300901_Braindumps_to_Pass_300901_exam
https://exed.canvas.harvard.edu/eportfolios/3033/Dumps_PDF/Believing_In_200301_Braindumps_Myths_for_Far_better_Outcome_in_200301_Exam
https://exed.canvas.harvard.edu/eportfolios/3032/Braindumps/Updated_820602_Braindumps_To_Pass_820602_Exam
https://exed.canvas.harvard.edu/eportfolios/3032/Braindumps/Master_The_Art_Of_500174_Exam_With_Most_recent_500174_Braindumps
https://exed.canvas.harvard.edu/eportfolios/3032/Brain_Dumps/Updated_500173_Braindumps_To_Pass_500173_Exam
In "The Black Swan", Taleb classes accountants as experts and "risk experts" as non-experts. This sidesteps the fact that if accountants are going to sign off the accounts of banks, insurance companies and pension funds they are required to sign off on numbers that include large elements of risk. To determine this number requires a method and no method has yet been found that is perfect in this regard. Relying on market values leaves you exposed to huge volatilities such as those experienced in the run up to the Lehman's insolvency. We can only attempt to guess why market values have gone up or down peering into the black box which generates these numbers. It may be due to a significant piece of information coming to light or perhaps it may be due to a respected market observer stating an opinion. It may be due to a market fall that has been caused by herding due to a market rumour posted on an internet rumour site or as Taleb would be first to point out, perhaps there is no clear reason and it is just due to random fluctuations. Although market values are in some cases easily obtained and certainly a useful first glance of the underlying riskiness of that stock or instrument, purely relying on market values to assess the underlying risk of an investment also carries random but very severe side effects.
So what about using models instead? Clearly they are not a perfect method either. However, they have a number of significant advantages over the market value. Firstly, the number that is generated is due to the inputs, distributions and correlations that you have imposed. You are able to arrive at a valuation using your long-term view of the landscape rather than relying on the market's view. Secondly, a market value is arrived at using common supply and demand forces. Exposure to a risk may become unfashionable due to an infinite number of reasons. Perhaps management at other companies believe shareholders will punish them for being exposed to a particular risk. Perhaps changes in personnel have resulted in less expertise in that risk area and so they don't feel comfortable being exposed to a risk they don't fully understand. All circumstances will be different but it is by no means obvious that the market value will match the value your company, with its own individual characteristics and risk profile, would place on a particular risk. Thirdly, there is an argument that the market is excessively short-term focussed. Principles that successful investors such as Warren Buffett abide by such as focusing on the fundamentals of a stock, preferring steadily increasing streams of dividends to short-term trading in volatile markets and expressing a preference for companies with iconic brands that prudently invest in their risk management practices do not seem to be valued as highly by the modern day market. In times of turmoil, these principles are easily forgotten but at what cost to the stability of our industries and our national economies? Take pensioners as an example of a stakeholder who heavily rely on steady, consistent returns for their income needs.
Taleb also highlights the weaknesses and shortcomings in rationality of human beings when making decisions. This is a great concern when the decisions of key individuals can have a huge impact on the stability of the financial markets. They are explored in detail in the paper "Making actuaries less human - lessons from behavioural finance" by Nigel Taylor. One specific human bias is anchoring where decisions are often made by adjusting from an existing position even when that position is clearly not relevant to the present problem. Others include prospect theory where people are risk averse when facing gains but risk seeking when facing losses. Myopic loss aversion refers to when the frequency with which something is monitored can impact the decision. The modelling process clearly does not fully address these problems as human beings are required to decide on the methodology and assumptions of the model. However, I would propose that carrying out and collectively challenging the modelling process is more robust to these human deficiencies than arguing over numbers that have been purely plucked numbers from human minds. Experience of a risk area clearly brings intuition that needs to be channelled into a final estimate but complete reliance on individuals' intuition leaves an estimate highly exposed to this human risk which needs to be offset against model risk.
Another criticism of models is that they can be exploited for particular purposes. Just as an unregistered, unregulated (black-market) doctor can administer the most expensive medicine for his own financial gain, a model builder can set the assumptions to create an output that backs his or her particular view. However, this is not a case for discarding the use of models just as it is not the case for banning the use of all medicines. The datasets, the methodology and the assumptions which create these model outputs should be challenged by the management so they can get comfort that the model adequately reflects their view of the risk landscape. And of course, quantitative models are not the only solution. Risk controls, risk processes and other qualitative tools must be used to make your organisation as robust as possible to black swans, grey swans and indeed white swans. Although some black swans such as a meteorite wiping out the world's population will ultimately destroy your organisation, that is not an argument for not proactively addressing the risks you can control and mitigate using each and every tool at your disposal. All swans can pose both upside and downside risks to the future of your organisation and they all need to be looked squarely in the eye.
From a personal perspective, I was quite taken aback on my first read of "The Black Swan". As an actuarial student currently working in the banking industry I felt uneasy about embarking on a career which was potentially not able to offer clear value to society. Reading some of Taleb's arguments initially made me think the time studying for my mathematics degree and working towards an actuarial qualification had been wasted. However, the more I thought about the complexity and subtlety of the problems encountered by actuaries and risk professionals, the more I thought that there was more to this than met the eye. Quantifying and building distributions for risks is challenging enough but then communicating the uncertainty around these distributions to stakeholders makes it a very difficult business. I am also convinced that knowledge of the underlying mathematical techniques allowed me to understand risk a lot clearer than someone without a mathematical background picking up "The Black Swan" for the first time. Taleb himself explains in his previous book, "Fooled by Randomness" how his understanding of risk developed whilst experimenting with his stochastic model. I have encountered a number of professionals in my career thus far who have experienced past crises, assisted organisations in dealing with their risks and established valuations for certain assets and liabilities that were extremely difficult to value. It has been said that history doesn't repeat but it echoes. If this is the case it is possible that the banking industry can eventually emerge from this financial crisis wiser and more resilient, ushering in a sustained period of stability such as that experienced in the post-war period.
No comments:
Post a Comment