Wednesday, December 17, 2008


I write this listening to Sir Victor Blank on Radio4 blaming American sub-prime issues by Investment Banks for unforgivable irresponsible behaviour, not banks like his. He says his banks lending is only in the business of lending to those who can repay. He does not mention his bank's loans to American Special Purpose Vehicles who managed and issued the sub-prime and other securitisations. This is followed by William Haig and David Cameron breathlessly and wholly unrealistically attacking the Government for the chaos in the housing and financial markets. Victor Blank had said the first quarters of next year will be especially difficult and house prices are expected to fall another 10%. This sounds like a short and sharp recession. All these are matters examined and accounted for in banking regulations, especially Basel II. A long term friend and colleague, in Brussels, commented that FSA 'waivers' (regulatory licenses approving compliance with Basel II, CRD etc.) for banks, fund managers and insurers can be granted like backstage passes - nodded through by the theatre impresario, not perfect, much more work still needing to be done, risk still a silo culture, not yet fully imbedded in banks, for many a bridge too far etc. An underestimated issue is how everyone understands banking finance and risks graphically or pictorially. However well we define matters in words and mathematics, as every architect knows, ultimately we need pictures too.This got me thinking, beginning with a J.S.Mill view that may be obvious: conditions necessary for our everyday freedoms also facilitate private enterprise crime. We need our human, democratic and libertarian rights etc., to protect us from government and corporatist caprice. But, this is also a balance between public sector and private sector 'criminal behaviour'. David Cameron, Conservative Leader, is doing his best to be populist about credit crunch recession, most recently to lead the mob baying for criminal arrests and punishment of "irresponsible bankers". Madoff's pyramid of $50bn losses is a symbol of symbols for this. But he was not irresponsible; he was criminal. He banked his reputation, trust, confidence, status and created a discrete but massive business alongside his two (or more?) other alternative investment firms. The $50bn reputedly lost in his fradulent non-investment scheme ranks alongside Drexel BL, LTCM, Enron and Worldcom if few others as the world's biggest white-collar crimes, with numbers one associates with the losses of Citigroup, UBS, Lehmans and a host of other banks. MADOFF pretended he could operate in the top decile of hedge fund performance, but proposed instead to only deliver modest but steady returns, and not to charge a fixed fee. He was in fact far below the very bottom decile oh hedge fund performance? How the money was 'lost' remains a mystery. Was it all paid out back to investors as performance returns? Madoff falls into the category of rogue trader or fraudster or operational risk. He is not representative of so-called irresponsible bankers, that epithet belongs to the professionals who invested in him without full due diligence. They can however use the excuse that the SEC's regular due diligence did not uncover the truth either! Fraud and credit crunch do overlap when it comes to misprepresentation. There have been about 300 FBI arrests in the USA so far connected to mortgage fraud and credit crunch matters including at Bear Stearns and other substantial major firms. The main crime is insider trading fraud; saying one thing in public or to clients and something else to friends. There are hundreds of major lawsuits based on 'knowingly giving false assurances and advice'. Little approaching this scale is happening so far in UK or EU. Yet, the Capital Requirements Directive (BaselII regulations) in the EU are law. Not just firms are culpable for breaches, directors and other key individuals in banks and other financial firms are personally liable for breaches of the law. Checking lists of banks' risk officers from only a year or two ago, one can see that in the last year many have either been moved, fired or resigned. CEOs, Chairmen, FDs, other directors and heads of structured finance etc. who have 'gone' is not a short list. If hauled into court, as in the US, they may appeal on the grounds that the accounting systems didn't work or the regulatory supervisors gave them a green light or it was all too much for anyone to fully understand and be responsible for. Just as they transferred credit risk they know how to transfer blame. Solid prosecutions may require 'smoking guns' and hands caught in the cookie jar? We have yet to see anyone arraigned for not complying strictly with regulatory reporting rules. The pace too at which bank officers were moved around without sufficient desk time in any one job to be culpable, not least given the amount of discretionary interpretation allowed in key aspects of Basel II (Pillar II especially) means that prosecutions will be harder to achieve, costing cost a lot of time and money (halcyon years for corporate lawyers) and for this reason alone may not come to court.
The Bank of International Settlements, the author of Basel II, headquartered in Basel, Switzerland, heroically built a wonderful set of regulations (many thousands of pages, hundreds of papers) with a mind-boggling pedantic thorougnness and intellectual brilliance (I say this as one of very few who know and understand most of it). But, compromises, trust and confidence in the managerial and intellectual competence of banks and their accounting systems, have inevitably allowed many degrees of freedom, latitude to game the regulations. But gaming regulations is the elastic least of it. Gaming regulations is a complexity that exists in all markets and around all tradable instruments. Failures in the quality of markets is possibly more to blame for the present crisis than irresponsibility of financial firms, banks and hedge funds and so on. That nearly all of capital markets is over-the-counter i.e. not transacted through fully regulated data reporting exchnages, that equity markets have been shredded, that credit derivatives, stock lending, short-selling, dark pool, internal and external crossing networks, hedge funds and 'shadow banks' are all well hidden from public scrutiny. Basel II and its related regulations do not address regulatory rules and reform of how financial markets function, not directly, only the financial resilience of individual banks, retail fund managers and insurers. How market functions is the responsibility of exchanges, and or exchange members, and other 'stakeholder' authorities can have strong influence.An important aspect of all this is how to understand the extremities of risk. Last week at a gathering in Zurich, at the Swiss Federal University of Technology (ETH), Einstein's alma mater, we listened to mathematics professor Paul Embrechts who, in the regulatory context, is one of the most impressive risk experts in the world. His close colleague Alexander McNeil is professor at Heriot Watt, Edinburgh. Over the past decade they have progressively succeeded in convincing leading banks to rely less on mechanistic maths and fixed assumptions when analysing and calculating risk values. Their thinking and that of Claudio Borio at BIS will become more central central to the further evolution of Basel II (EU's CRD), Solvency II, and IFRS, rules and regulations - displacing Black-Scholes inspired formulae, discredited when LTCM collapsed, and direct descendants that assume a fixed or symmetric or Gaussian shape to increasing and decreasing risk. If you want the detail trying googling or wikipedia. What can be represented with clean two dimensional precision in mathematics is of course more complex in human-centred, stochastic and probabilistic, not like natural science or enginering realities, that are repeatedly reproducible. Essentially, the bell-shaped Gaussian image of risk ranges from highest return in normal expected conditions to embarassing losses in extremely unexpected conditions. When interdependence is important, the pattern represented as a Gaussian normal bell curve with thin tails may be contrasted with empirical observations of fatter tail risk events - at first glance, this may obtuse, abstract, and not very important, as the two curves shown here do not appear to be that different. But on closer scrutiny, there is an enormous difference. The “tails” of a fatter tail curve — the regions to either side that correspond to large fluctuations — fall off slowly in comparison with those of the Gaussian bell curve. These so-called fat tails imply that large events take place far more often than one would expect on the basis of “normal” statistics. In the case of market fluctuations, for example, the bell curve predicts a one-day drop of 10 percent in the valuation of a stock just about once every 500 years. The empirical fat tail statistics gives a very different and more reliable estimate: about once every five years, and 'unexpected' risk becomes 'expected' 100 times more often.
But, for the sake of factoring in guman subjectivity, let's go further. Replace the abstract curves with another more mesmerising set and you come closer to what bankers see when they focus on short term profitability attractions leading to irresponsibly ignoring the tail risks, the unpredictable risks of instinctive and reactive legs and feet, say? Banking is part of the economy. It is a machine and a system, but also red-blooded people-centred, and rarely especially politically-correct. There is only so far that risk analytics can go mathematically. Capital buffers, risk margins, are required to cope with the unexpected. But at the same time banks have to deal with competitive pressures and market prices. These can become dominated by short term profit-takers over long term investors - a an enormous conflict. Short term risk trusts in liquidity changing very little; the speed at which one can buy is the same speed at which one can sell. When the world was clamouring to buy asset backed securities (ABS) investors did not check whether selling would be just as easy. The ABS markets grew phenomenally market as a seller's, but were never tested by a buyers' market, by a downturn, a loss in confidence until suddenly in 2007. The rapidity with which what could be bought and pledged as collateral, was suddenly followed by a market in which no-one would buy except at enormous discounts. Markets can under-price risk. However sensibly banks manage their risks they are at the mercy too of markets behaving less than rationally about the totality of the quality of the maarket. Where major firms and reputable traders go,many others blindly follow. This too is what Madoff relied upon, his 50 year long reputation as a sharp cookie and a market insider's insider. The ethos of financial regulation in recent decades has been to say fix all we can but "let markets decide the rest"! This vests a lot of trust in those who are the hard-core of market traders. Can financial services firms be blamed for trusting in the ideology of markets and market-traders know best for igniting their own nuclear winter or are they victims too? It is one thing to understand banks as individual firms, with all their subsidiary business and profit-centres and individuals within them, quite another to understand them also as part of a total financial and economic system with all the interdependencies between banks and other financial service firms. Copula definitions, principles, warnings and prcepts can be simple or complex and are not yet commonplace in financial economics textbooks. The simplest kind critiques by Embrechts and McNeil and others are a stationery distribution of expected and unexpected risk along the bell-shaped curve that attempts to account graphically and fully for correlation covariances among risk factors determining the outcome of financial exposures. It also has the advantage of actually admitting that there are unexpectedly extreme shock events than (Gaussian normal) underlies Black-Scholes, by admitting the stylized fact that all asset markets exhibit "fat tails" or kurtosis, i.e. in an area beyond where normal risk cover is available - to produce suddenly far larger losses than any firm is prepared for (other than when market and economic conditions are near normal and not 'disorderly', panic-driven or 'turbulent' to use another now common expression). Copulas need an economic cycle dimension. It turns out, these are quite insufficient to deal with the multi-interdependency links driving the events of the last two years, with most funds and banks getting into trouble and losing on average twice their reserves to paper losses and at least one times total reserves to net economic losses. One of the Zurich conference participants was an econophysicist, Didier Sornette, who runs the "Observatory of Financial Crises." He made some not quite accurate forecasts (see "Econophysics and Economic Complexity," at, but described the problem of Gaussian bell-shape curves applied to individual transactions, and banks' portfolios: they don't take into account of 'herding' responses, which is where traders and investors follow the market in directions that do not fit with fundamntal risk analysis. Herding effects are why risk models, however used, did not predict the liquidity crisis or that a 10% fall in a market could happen not just on one day, but three days in a row. Crashes are predictable, but not long persistent repeating crashes. So again, another defence for the 'guilty' irresponsibles is that risk models, even the best, weren't up to the job of a 1 in 100 crash. Prof. Alexander McNeil makes some important corrections to the story I offered above:
1. The influence of the academic gnomes of Zurich in promoting the use of copulas is overstated. Paul Embrechts and I would consider our role to have been in showing the problems associated with applying Gaussian thinking in a financial world that is manifestly non-Gaussian. One problem of Gaussian thinking is an over-reliance on the concept of correlations for describing dependencies. In our hands copulas were tools for revealing the fallacies of Gaussian thinking and the pitfalls of a reliance on correlation.
2. They have a role in building models for particular financial modelling applications, but were never viewed as any kind of panacea.
3. I would say that we were surprised to find copulas being adopted in the world of
[Collateralized Debt Obligations = general term for banks' Asset Backed Securities]CDO valuation, particularly as the copula which came to dominate market practice was the Gaussian copula. In CDO valuation the use of copulas is purely for reasons of convenience.
Professor McNeil usefully also draws my attention to the difference between copulas and copulas, something I confess to having neglected hitherto:-
4. Best not to conflate copulas and cupolas. The former were originally
objects in grammar, verbs like 'to be' and 'to become' that perform a specific linking role. Likewise, in the theory of probability distributions, copulas perform a linking role - the maths is actually pretty simple. Cupolas don't need further elaboration but do carry unfortunate connotations of bell shapes and, by extension, Gaussianity.

This at last explains why I do not find cupola graphic charts in McNeil & Embrecht papers! Now back to Bernie. Bernard Madoff, former chairman of the world's second biggest stock exchange, NASDAQ. Bernie (bailed for $10m) was turned in by his sons, Mark and Andrew (also arrested) on Thursday, one of whom was the compliance officer. His hedge fund, Bernard L. Madoff Securities LLC, was declared insolvent with estimated losses of $50bn and bernie is charged with 'securities fraud'. This reflects badly on the Securities & Exchange Commission, SEC, which has u-turned a couple of times on hedge fund transparency. The SEC did seek more reporting oversight but was frustrated over a year ago in Federal Court. Madoff did not have to cheat anyone, least of all with so much funding (1-2% fee plus 20% of profit). He produced modest, steady returns for clients, claiming to be profiting by trading in S&P’s 500 Index options, and closing all positions prior to mandatory reporting dates so that investors had no window into the fund’s holdings. To professional examiners, the steadiness of the returns should have been a clear warning! Betting the index would have meant sizeable short term derivatives contracts on the Chicago Mercantile or other comparable options exchanges. The S&P500 has been volatile. It should have been easy for any financial professional exposed to Madoff's fund to check on that? Banks in UK, Spain, France, Switzerland, Italy, Belgium, Netherlands and other countries have potentially lost $billions. They loaned money on the strength of collateral in the form of Madoff certificates. Madoff also managed funds belonging to charities and was quite charitable to good causes himself, apart from just himself. Apart from individuals, charities and numerous "funds of funds" investing in hedge funds, banks such as HSBC Holdings PLC (ADR: HBC) and Banco Santander SA (ADR: STD), lent $billions to investors participating in Madoff's fund, secured only by holdings in th fund. SEC failed to protect investors. To some commentators it is shocking but not surprising that the losses could be so much as $50 billion, given the excessive growth of hedge funds. But, it is shocking and surprising to me. Everyone knows that hedge funds have not been reliably great performers. In 2002 returns were in small single figures and many have struggled ever since then. It is surprising that it was possible for any experienced investment manager to lose so much. It may be less surprising that major investors were trustingly naive. Madoff could be very convincing after over 50 years of winning other people's trust. Investors trusted him more than they would anyone else of their country or mid-town club set teeming with investment advisors. It is expected that charities should be unworldly and easily seduced by a great sales pitch. But, actually charities and their advisors can be as hard-nosed as any. The main investors were other hedge funds, who bought off screens analysing hedge fund performances using data not publicly available or full audit-checked. This raises a daunting question: how savvy are the big hedge funds? Modern-Day Ponzi (pyramid-selling) Schemes are a well-known (like the innumerable boiler-room scams) and the SEC has great experience unravelling them. But like all internet frauds, they are like Japanese knot weed; for every one closed down several more pop up. It is an endless and disheartening task to stamp them out. Ponzi schemes are fairly easy to detect by any reasonably suspicious professional. If offered an investment opportunity you simply keep asking questions of the promoter until you are absolutely confident about how the money is made. This may seem technically obscure, however, in the case of market indices. After all, Nick Leeson was able to fool his bosses at Barings and thought he could full the floor of the Simex about arbitraging the Nikkei index. I cannot believe that what Madoff was doing, if he ever did buy derivatives contracts would not have been known to others on the exchanges including the many Chicago 'own traders'. The rule should be if you can’t figure it out, you don’t invest, and take advice. Finance is an area in which there should be no impenetrable mysteries, and none to experienced and competent professionals. It is also worth noting that no-one has consistently profitable insight into a whole market index - why no-one on the floor of the Simex believed leeson and took him to cleaners. Leeson exposed his own bank's management incompetence. Madoff has exposed a swathe of similar incompetence among many banks, hedge funds, other investors and finance and regulatory professionals who ought to have been trained to know better! In the emerging capitalist markets of 1990s Eastern Europe, pyramid schemes were a known hazard. Victims were deluded about how capitalism worked and regulation was weak. The MMM scheme in Russia collected $1.5bn, Caritas in Romania collected $1bn, and in Albania in 1997, the banking system and government collapsed under a $1.2bn pyramid fraud. In the West, Enron and others operated in effect similar schemes insofar as they created circular ways of artificially inflating their profits and disguising their debts and losses. There have been thousands of cases of mis-selling and of dumping proprietary losses into client accounts and of preferring bigger clients over smaller clients. These can be described as partially-Ponzi. Two groups of investors appear vulnerable - rich club member types who trust in personal “connections” – and another type beloved of Swiss private banks whose dubious, criminal or tax evasion money is handed over on the strength of conspiratorial assurances by people who neither understand nor care about how investment returns are generated or whether they are relatively high or average or low. They avoid professional experts and specialists and buy 'mickey mouse' retail products from people whose “connections” appear to provide an “inside track” not unlike ractrack tipsters. Madoff, as a former chairman of NASDAQ with a charming personality, was qualified to appeal to gullible rich. Charities and others who don’t pay enough tax can be readily seduced. Ponzi schemes were enormously boosted by derivatives growth in the '80s and '90s. Even if clubbable investors, those with power of attorney and trustees have some idea how bonds or stocks work, many, even financial professionals, mentally cloud over when derivatives and trading strategies are described. Madoff was able to blow smoke'n mirrors. Private partnerships do not have disclosure rules comparable with public investment funds, and did not have to disclose trades or show accounts of any details about his trading methods. This does not explain or thereby excuse the gullibility, even culpability, of professionals managing hedge funds and “funds of funds,” or loanbooks who loaned to Madoff's victims or invested in his schemes. These people who are themselves inordinantly well paid perhaps believed excessive management fees buys superior investment, but they are no better than boiler-room scammers. They should be sued for failing in duty of care. It is not enough either to blame the “get-rich-quick” gold-rush fever or epidemic of 13 years of easy money and lax regulation. Professional investors and advisors failed in “due diligence”. They deserve to be sued for failing in fiduciary duty. If investing their own money, they deserve to be pilloried for crass stupidity and incompetence.

No comments:

Post a Comment