Why Big Capital Markets Players Are Unmanageable

Posted on by

John Kay comes perilously close to nailing a key issue in his current Financial Times comment, “Our banks are beyond the control of mere mortal” in that he very clearly articulates the problem very well but then draws the wrong conclusion:

At Oxford university, I often hear people say there is nothing wrong with the system: the problem is the vice-chancellor/master/bursar/ university officials. And, in a sense, they are right. If the vice-chancellor had the wisdom of Socrates, the political skills of Machiavelli and the leadership qualities of Winston Churchill, not to mention the patience of Job, he or she would be very likely to be able to fulfil the conflicting demands of the post. But such paragons are few and far between. In the meantime we must try to find structures that can be operated by ordinary mortals.

In the same way, the claim that the fault with the banking system lies not with the structure of banks but with the boards and executives that claimed to run them is both correct and absurd…if the failures are both as widespread and as persistent as it appears, the problem is in the job specification rather than with the incumbent. If you employ an alchemist who fails to turn base metal into gold, the alchemist is certainly a fool and a fraud but the greater fool is the patron.

The bank executives pilloried by the UK’s Treasury select committee of MPs were all exceptional people. The vilified Sir Fred Goodwin was an effective manager who had slashed through the National Westminster bureaucracy and revived a failing institution – a task that had defeated many able men before him. His chairman, Sir Tom McKillop, offered experience and ability that met every possible specification for such a role in a big international corporation. As chairman of HBOS, Lord Stevenson was Britain’s supreme networker. This skill is a particularly valuable attribute in an environment where the essence of banking is to extract very large sums of taxpayers’ money while giving as little as possible in return. His chief executive, Andy Hornby, was criticised for being a retailer. But Halifax, half of HBOS, needed retail expertise. The only thing it needed to know about complex securitised products was that there was no good reason to buy them.

Like Sir Fred, Sir Tom, Lord Stevenson and Mr Hornby, most of the people who sat on the boards of failed banks were individuals whose services other companies would have been delighted to attract…

The hapless four were criticised for their lack of banking expertise but it is, in fact, not clear what modern banking expertise is. The world of modern banking requires all the skills of these gentlemen, plus some others, and no one can expect to have all these attributes.

It has been said of Jamie Dimon (who does not have a banking qualification) that his dominance exists because at every meeting all the participants know that he could do each of their jobs better than they could. But the business world cannot operate at all if it can operate only with individuals of the calibre of Mr Dimon. Better, as so often, to follow an aphorism of Warren Buffett’s: invest only in businesses that an idiot can run, because sooner or later an idiot will.

Our banks were not run by idiots. They were run by able men who were out of their depth. If their aspirations were beyond their capacity it is because they were probably beyond anyone’s capacity. We could continue the search for Superman or Superwoman. But we would be wiser to look for a simpler world, more resilient to human error and the inevitable misjudgments. Great and enduringly successful organisations are not stages on which geniuses can strut. They are structures that make the most of the ordinary talents of ordinary people.

The problem is Kay is applying traditional managements structures to investment banking, Even though these entities may have substantial retail arms and bank charters, the area that poses the management challenge is the capital markets businesses. And he makes a dangerous, erroneous assumptions: that mere mortals, meaning generalists, can run these businesses. That is bogus.

What makes capital markets businesses different from any other form of enterprise I can think of is the amount of discretion given of necessity to non-managerial employees, meaning traders, salesmen, investment bankers, analysts. In pretty much any other large scale business, decisions that have a meaningful bottom line impact (pricing, new sales campaign, investment decision) are deliberate affairs, ultimately decided at a reasonably senior level. The discretion that customer-facing staff have in pretty much any business in limited. At what level does someone have the authority to negotiate a contract? And even then, how many degrees of freedom do they have?

By contrast, think how many decisions traders and salesmen in capital market firms make in a day, and their potential bottom line impact (though experiment: how much damage could a truly vindictive trader do in a day or a week, if he decided to blow up his employer?) Investment bankers work over longer time frames, and like many normal businesses, have a lot of things routinized so as to make them more efficient, but it also limits their latitude (standard forms for many types of client agreements, standard pitch book formats, etc). However, unlike “normal” businesses, a frequent activity in investment banking is creating new products, often in a very ad-hoc way, with teams with relevant skills thrown together to try to push something through. The politics are often sharp-elbowed, but people are too pragmatic to let turf issues interfere with getting a new deal launched).

The approach for managing these businesses in the days of partnerships, when the owners were personally liable for losses, was to have small units with partners running them who knew the business and could oversee it properly. Effectively you had four layers: associate/analyst (the college kids, the analysts, did pretty much the same stuff the associates did, who usually had MBAs, except the MBAs got to go to client meetings more often), VPs, and partners, but some of the more senior partners were department heads in units that also had partners (who’d manage either people on their desks, if traders of salesmen, or if in investment banking, had accounts and various VPs and associate types working on each client). But those department heads had also grown up in the business, and were still active in it. Heads of significant departments in turn would be on an executive committee, a part-time role.

The problem with this model is it starts to come under strain when the partner group gets too large. And OTC markets have strong network effects, so having bigger market share confers a competitive advantage. And now there are high minimum scale requirements for being in the business. You need to be in all major times zones with a pretty broad product array. all kinds of back office support, all kinds of IT for risk management, communications, position management…

So the scale of operation required to be competitive is too large for it to be managed by player-coaches who had deep expertise, and like the Dimon example, were more expert than the people working for them. But the normal corporate/commercial banking management structure, with more managerial layers, and the top brass having broader spans of control, was devised in earlier stages of industrial organization, when you had factories or service business with a great deal of routinization of worker and middle manager tasks. Traditional commercial banks are on the same factory format. They handle large volumes of very simple, standard transactions with a high degree of control and oversight. That’s a big reason why it took commercial banks over 15 years to make meaningful headway against investment banks. Although regulations were an issue, the bigger barrier was the radical difference between the two management cultures. There was no regulatory barrier to commercial banks offering mergers & acquisitions, for instance, but they were lousy at that for a very long time.

So Kay is effectively asking for a traditional commercial banking model, businesses “that make the most of the ordinary talents of ordinary people”. There are businesses like that in banking, but they are mainly in retail banking and corporate lending. If you want that world, you need a far more radical change in the industry than anyone is contemplating now. You’d need to go to the world that Taleb advocates, From a list of his ten suggestions:

4. Do not let someone making an “incentive” bonus manage a nuclear plant – or your financial risks. Odds are he would cut every corner on safety to show “profits” while claiming to be “conservative”. Bonuses do not accommodate the hidden risks of blow-ups. It is the asymmetry of the bonus system that got us here. No incentives without disincentives: capitalism is about rewards and punishments, not just rewards.

5. Counter-balance complexity with simplicity. Complexity from globalisation and highly networked economic life needs to be countered by simplicity in financial products. The complex economy is already a form of leverage: the leverage of efficiency. Such systems survive thanks to slack and redundancy; adding debt produces wild and dangerous gyrations and leaves no room for error. Capitalism cannot avoid fads and bubbles: equity bubbles (as in 2000) have proved to be mild; debt bubbles are vicious.

6. Do not give children sticks of dynamite, even if they come with a warning . Complex derivatives need to be banned because nobody understands them and few are rational enough to know it. Citizens must be protected from themselves, from bankers selling them “hedging” products, and from gullible regulators who listen to economic theorists.

If we can’t shut down credit default swaps, which the more I dig, the more I see they had a very direct role in the meltdown CDS on subrprime mortgages started in 2004, and there is a longer form gloss as to how that played a major role, if not the key role, in the superheated demand for “product” particularly subprime, in the manic phase of the credit bubble), we will never get to a world like the one Kay wants to see, or at least not until we hopelessly break the one we have now.

Print Friendly, PDF & Email


  1. tkr

    I can't agree with your case for I-banking exceptionalism. I can think of a few businesses where extraordinary responsibility is concentrated regularly in the hands of junior-level employees – medicine, law, and the military to name a few. And they seem to do a perfectly fine job of keeping themselves moving along without nuking the planet.

    I think the effective differences the above professions have from banking demonstrate the challenges clearly. First, they have high barriers to entry and a long indoctrination process that qualify and train people before they are allowed to make complex decisions. Second, they put the financial rewards (when they exist) at the medium-to-far end of the career path, so that quick buck artists are not encouraged to enter. Finally, they are united by a standard of moral/ethical behavior which guards against profiteering based on their unique positions.

    I think if you applied the lessons of real professions to banking, you might see an improvement. If you want to keep people from blowing up the world, you either take the black powder away from the bankers or put a real tight process in place on when and how they use it.

  2. Joe Costello

    Only time will tell, but Im not sure the system isn't hopelessly broken now, but again only time will tell.

  3. Hugh

    Vanilla banking should be segregated out as a government enterprise or tightly regulated private one. As for the rest, size and product constraints should be applied.

    As for the rest, I favor re-imposition of Glass-Steagall and tighter control of markets with a view to prevent companies and financial products that pose systemic risk.

    Other than a culture of greed, I still do not understand how the financial industry did not see the consequences of their activities from miles away. The bursting of the housing bubble and the financial meltdown were two of the most predictable events I can think of. That many economists and those who work in the markets maintain "no one could have predicted" is an indication that there is something fundamentally wrong about the structure of the markets. Most of the time I think they are lying and that they had to have known. But either way, markets need to be structured in such a way as to assume that on their own they will act in completely self-destructive ways.

  4. Doc Holiday

    The way I see this distorted situation is that when someone does come along (in banking or government) who can whip it (whip it good) they are slandered or screwed over like Brooksley E. Born (who could whip it): "As the financial crisis of 2008 gained momentum, newspapers began reporting on what might be some of its causes, including the adversarial relationship Greenspan, Rubin and Levitt had with Brooksley Born, [11] with Greenspan leading the opposition, and how Born's recommendations were suppressed.[9] She is retired from Arnold & Porter and until recently had declined to comment on the unfolding crisis and her efforts to rein in the growing market for derivatives. "The market grew so enormously, with so little oversight and regulation, that it made the financial crisis much deeper and more pervasive than it otherwise would have been." The disagreement has been described as a classic Washington turf war. She now laments the influence of Wall Street lobbyists on the process and the refusal of regulators to discuss even modest reforms"

    Praise be Wiki: http://en.wikipedia.org/wiki/Brooksley_Born

    We can also see this with that nameless women that is sort of a watchdog for TARP … yah know, the schoolmarm chick from Harvard with the thick glasses and short hair that seems so smart and nice, but who is now forgotten (I wanna say Leslie, but I'm drawing thin air). I like her, but where is she, whi is she?

    We see this also, in this: "Senators Hold Personal Stake in Bailed-Out Banks
    Conflict-of-Interest Rules Do Not Prevent Lawmakers From Buying Stock in Companies They Regulate": http://washingtonindependent.com/49451/senators-hold-personal-stake-in-bailed-out-banks

    Of the 74 upper-chamber lawmakers who supported the $700 billion financial rescue in October, at least 15 own direct shares in institutions receiving federal funds under the Troubled Assets Relief Program, according to financial disclosure forms filed by members of Congress last month. Combined, those 15 members hold between $1.2 million and $3.0 million worth of stock in TARP beneficiaries — firms that have received no less than $330 billion in TARP funds and loan guarantees since the program began.

    We can also hear this in the words of Bob Dylan:

    "Heard your songs of freedom and man forever stripped
    Acting out his folly while his back is being whipped
    Like a slave in orbit he's beaten 'til he's tame
    All for a moment's glory and it's a dirty, rotten shame."

    Re: "They were run by able men who were out of their depth. If their aspirations were beyond their capacity it is because they were probably beyond anyone’s capacity."

    We can also hear this here: http://www.youtube.com/watch?v=H3_1_0ZCCYw

    Full Disclosure: No bird found and I'm not sure what this was about.

  5. ThatGuyInTheBack

    The financial crisis is a direct result of a political crisis – one that happened silently.

    Financial industry lobbyists essentially bought legislation and legislators. Money became the silent, fourth branch of government, and remains so today. Until *that* power is controlled, we are doomed to lurch from one otherwise avoidable crisis to the next.

  6. Yves Smith


    With all due respect, you are missing the frequency and magnitude of the importance of the decisions rhat rest at comparatively low levels of the organization.

    Medicine is not a comparable. Medicine is still practiced in a very decentralized way. By contrast, capital markers businesses are highly concentrated. Do you have operations in medicine of such large scale that doctors are routinely working in very large organizations with 2-5 layers above the decision makers? How often does a doctor make decisions in, say, a hospital that if he is badly wrong could seriously damage the organization? He can kill people, which is a tragedy, but cause a hospital, which is the largest form of medical organization I can think of, to close? Perhaps only if he did something truly vicious, like released a horrid pathogen in the hospital.

    This isn't exceptionalism, this is a particular set of environmental requirements. They are peculiar to capital markets businesses as they have evolved.

    Ditto the military. They have much deeper hierarchies, much more training to instill obedience, much less routine discretion. Again, even in war (and front line combat is a small subset of the total manhours spent by the US armed forces), how many decisions does a soldier make in a day? How many have implications for his ARMY as a whole? Again, you are missing the significance of the discretion, magnitude, and frequency of decisions relative to the organization as a whole.

    In the law, in any real firm, nothing goes out the door unless a partner reviews it. I do not see junior level employees having much discretion. And the decisions made on any one matter do not have the same potential economic impact. Plus law firms are still of a scale, even all but the biggest, to operate much like the old partnership model I described, with senior staff more expert than juniors. And law firms do not have overwhelming scale advantages, plenty of practice areas where one partner or small group practices are viable.

    I think you are missing the key distinction: the decisions of the employees in the other contexts can and do have significant consequences for the individuals they interact with. I am talking about the consequences for the organizations within which they reside. No single junior or middle level person in the army, a hospital, or a law firm could bring down the entire organization, as did Nick Leeson at Barings. And there are other rogue trader or spectacular error cases where the business was mortally wounded, limped along a while, and then got sold.

    The rogue trader is an extreme case of the more general pattern, of the significance of the decisions made on a routine, day-to-day basis by middle level people to the organization as a whole.

    And these organizations have now been deemed systemically important and are backstopped by taxpayers.

  7. juan


    Then you do not favor decentralized decision making if such mode of organization at least potentially creates more, what I think I'll call, enterprise risk than not?

  8. Hugh

    Doc, I think you were thinking of Elizabeth Warren.

    Yves, would reduction in the kinds of financial instruments, their complexity, and the size and interconnectivity of the companies that dealt with them limit the damage that the current financial business structure could do? Would making corporate officers and/or partners personally liable be feasible?

  9. trelsco

    Thanks Yves

    I've been thinking for a while about why financial services seems "different" to other industries. In investment banking (on the trading side at least) we often have relatively junior people wielding outsized economic power. I believe this is a consequence of the high level of specific knowledge required and the very short decision-making timeframes. This creates a vulnerability but doesn't necessarily cause a crisis.

    There are a couple of other factors we need to consider.

    First, the opacity of financial services. We are dealing in leveraged future risk and returns which are difficult to understand and get a handle on. This creates a fog around those junior people that reduces the effective level of oversight.

    Second, the leveraged and asymmetric incentive structure for these people was not good for controlling firm or system risk.

    Finally, we do not seem to have worked out a way to apply an information-theoretic overlay (ie feedback, adverse selection , moral hazard issues, etc) to guide our decision-making and avoid systemic crises. At this point we do not even have the information to do this even if we had the framework.

    Personally, I think investment bank's boards and senior management did not show leadership in getting on top of these issues. From your comments and the above Finance has idiosyncracies, but I don't think that invalidates general approaches to running firms. We just need to acknowledge and adapt them.

    Great discussion, though.

  10. Yves Smith


    The problem is a bit different. There are limits to how many people one person can oversee effectively. The risks that line staff can and do take on a daily basis in capital markets firms are large. Using VaR and other tools to manage risk has limits; you need human oversight by someone who actually knows the business is a pretty granular way. The idea that you can have generralists oversee these businesses effectively is an illusion.

    So there are scale limits to effective supervision, but the economics favor larger scale in lots of businesses, particularly trading operations.

    Hugh, Doc did mean Born, who tried to get credit default swaps regulated.

    The best hope is probably increasing liability big time, and driving it below the officer level too. Aren't that many officers in a corp. Criminal prosecutions would help, but prosecuting complex financial matters is very resource intensive, they tend to go after simple frauds.

    I favor simpler products (you'd need to clamp down on fiduciaries too, so activity would be limited) but it is proving well nigh impossible to rein in credit default swaps. If we can't do that, I have little hope for other measures.

  11. skippy

    Ahh… the age old question, how to regulate human behavior umm lets see prison, nope they either get better at it, and reoffend… how about death, naw seems to only increase the ferocity of the act ie: if I'm going down I'm taking lots of others with me if I fail.

    For me at least, I think Yves earlier posts on psychiatric condition/ing is the key. In positions where MAD out come are possible screening is a must. Hay does the Military just hand over the nuke keys to the clever, nope its more of a psychological perspective with fail safes installed and even then accidents happen.

    So for persons of low or zilch empathy scores, you don't hand over the keys to the nukes or positions of high responsibility public or private, dam the GDP or P/E.

    skippy…always better to live than die, I say and if blood is the price of freedom, can you wait till I get a beer and a smoke, I want to settle in for a watch.

  12. Alan


    Ditto the military. They have much deeper hierarchies, much more training to instill obedience, much less routine discretion… Again, you are missing the significance of the discretion, magnitude, and frequency of decisions relative to the organization as a whole.

    "With all due respect" you are plain wrong. (we'll leave alone how and why you could develop this very bizarre idea of capital markets exceptionalism). I'll take one example. During supercarrier operations it's within the power of several hundred individuals to destroy the entire ship either through negligent mistake or intentional malice. Most of these people are pay grade E4 and below, 6-8 levels below the "CEO" level of ship captain.

    The control and operation of strategic nuclear forces for over 50 years now is another example. You may recall that increasing sloppiness was recently found to be creeping into daily USAF nuclear force operations. This led to the removal of the Secretary of the Air Force, the Chief of Staff and dozens of lower ranking people.

    How the personnel of "capital markets trading firms" were able to avoid similar accountability despite commensurately worse results remains the question. They outcome they delivered was the professional equivalent of an unauthorized launch against foreign targets.

    Maybe it's as simple as the fact that the Secretary of Defense wasn't an alumni of that dysfunctional USAF system. This is unlike Henry Paulson and the numerous other GS graduates teeming near the levers of control in all recent administrations.

    However, this manifest institutional corruption isn't an argument for an institutional capital markets exceptionalism to which the entire world should adapt itself, or for which unique structures must be created.

  13. Yves Smith


    Even the destruction of a supercarrier, while it would be very damaging, would not bring down the entire navy. The interdependencies are not as high.

    Moreover, I would assume that the military is able, IN THEORY, to supervise the activities of lower level operatives in most positions fairly effectively. In fog of war, front line settings, such supervision is very limited, usually impossible, but their ability to destroy an army (ex collaboration with the enemy, which I am not discussing in the IB case either, merely error or self motivated behavior) is just not there.

    We are also talking past each other on the nuclear issue. You are pointing to a failure of fail safes and supervision in nuclear ops. But those DO have tons of fail-safes, You mention "increasing sloppiness". This is not addressing my point. You are not saying the managerial risks CANNOT be contained, but WERE NOT. Those are two different statements.

    You are still missing the essential point, which I made earlier in the thread:

    So there are scale limits to effective supervision, but the economics favor larger scale in lots of businesses, particularly trading operations.

    You need people with a high level of expertise to properly supervise many IB activities. They involve specialized skills. The staff has huge incentives to game management, something also largely absent in the military. No one wants to create a nuclear accident, or have their unit or army perish, by contrast, while banking has "IBG, YBG", I'll Be Gone, You'll Be Gone.

    It is not exceptionalism to point out very peculiar requirements of an industry, given how it has evolved and succeeded in making itself critical to the functioning of the economy, to the degree that it is getting priority in the federal handouts and still has shocking power in K Street despite the damage it has done. "Exceptionalism" says the differences observed are inaccurate.

    You seem to have misconstrued my argument, First, the proportion of people making decisions in I-banking that have serious consequences to the bottom line of the organization on a day to day basis is far higher than any activity I can think of. Even in your supercarrier and nuke examples, how often do those decisions come up? And more important, are these decisions INHERENTLY difficult to manage (or train for) due to specialized skills needed to supervise adequately? Yes, you have chaos in battle, but that is a small percentage, as I said, of the man-hours spent in the US armed forces, and the bad decisions of a single mid level person cannot bring down an entire army. The one exception is nukes, but those can and I understand are designed with multiple fail-safes.

    Failure to supervise well doe not prove inherent difficulties in supervision.

    I am no fan of this industry, and you would know it if you have read this blog. But asserting "exceptionalism" and refusing to consider that the industry does indeed have qualities that make it difficult to tame in its current form GUARANTEES the preps will continue to have the upper hand,. You need to understand what you are up against in order to rein them in.

  14. Richard Kline

    So tkr, your examples—medicine, law, and the military—are in the first two instances not germane. Much as Yves has said, there are few instances where individuals have such discretion that their actions or failures to act could _commit_ the entire enterprise irretrievably to a course of action of a scale as to dissipate the whole.

    In medicine, the one example might be in epidemeology where are false proncement could precipitate widescale actions, for better or worse. No other medical professionals, great personal expertise notwithstanding, have such leverage on the actions of their institutions as a whole. In law, or one might take the example of accounting, in principle no one partner or group is supposed to be able to take 'enterprise at risk' positions without the backing of senior control.

    One might look at the example of Arthur Anderson as an exception, but it is the exception which supports the contention. There, the Enron partner a) engaged in actions which were definitely illegal and should have been aware that they were such, and b) very likely did so with the tacit understanding of at least some senior control. [cont.]

  15. Richard Kline

    There are issues in the example of the military, however, worth pursuing here. It is very much part of 21st century command-and-control to push considerable decision-making power down the chain. Partly, this is because tremendous levels of firepower can be concentrated at present very quickly. In part this is because the rapidity of combat action in highly mechanized and wired units is very high. So whereas the historical norm in many armies has been _highly_ top down, there is a good deal of 'commit authority' now further down the command chain. The various forms of the German Army not later than WW I were already moving in this direction. That will sound surprising to many out there used to thinking of those Pruoosians as highly inflexible, but there's a larger story. In WW II, much of the German unit superiority—and there is overwhelming evidence of pervasive unit superiority in combat relative to ALL other armies for the Germans—was due to pushing commit authority down the chain as far as immediate combat actions were considered. So more than most might think, modern military mid-level commanders have considerable operational 'commit' authority.

    . . . But even so, how much does that commit their enterprise? Say you have 100k men and are facing an unidentified number of threats/opportunities which may be of commensurate scale, though likely smaller and more dispersed than you are. You are an enterprise; those threats are typically dispersed rather than a counter enterprise [to mimic an ibank's position]. The effective equivalent of Yves' trader/mid-level operative is a colonel. Would you permit a single colonel to commit a third of your enterprise on their own assessment without running that by senior control/generals? That's a lot. If the colonel gets it right—they needed that level of resource, and are successful—then they should be a general. –But that colonel had better be right because if they lose or severely degrade a third of the total, what remains may be insufficient even for the enterprise to survive in some contexts, and the enterprise will be severely damaged under the best of circumstances. So giving the mid-level 'sole commit authority' over even a third of the whole is, well, excessively risky, even if it leads to repeated success. One failure, and the enterprise risks total failure.

    So extend the example. Would you limit the 'sole commit' of a colonel then to, say, a tenth of the whole, 10k men? That would seem wise. However, once their are committed, they may be insufficient for success and require the further commitment of much of the rest: that is 'engagement risk' problem. Getting in is a whole lot easier than getting out. So further more, the commit authority is context-specific. In some situations, yes, the colonel can commit because the opposition/opportunity is known, because they are in advance of a pre-determined operation, because the rest of the enterprise is concentrated and may not engage, etc. But the context must be _well-known_ for such situations to apply.

    How did the Germans solve this problem? The Staff. (I forget the Deutsch which really should be quoted.) Colonel-equivalents could commit, but they had to run what they were doing by the Staff, whose job it was to have the big picture handy: men, logistics, operations, intelligence. Those staff officers were NOT highly senior. Many doing the real work were junior to operationl 'deciders' in the field, to say nothing of generals. And they did not 'get the glory,' in this analogy the big bonuses. But the Wehrmacht ran well because the staff work was committed to enterprise-control and survivability. There's more to this, but that's enough. [cont.]

  16. Richard Kline

    Now take one further relationship more aligned with the ibank. That military enterprise just mentioned might have a dozen 'colonels.' But an ibank of international scale will have dozens and dozens. Moreover, these 'money colonels' are not necessarily limited to commensurately small chunks of their enterprises overall resource. In fact, I suspect that if you added up the operational 'commit authority' of those money colonels it would total several multiples of the actual resource base of their enterprise. In principle, risk management folks were supposed to keep track of that and rein in exposures at the individual decider level. But we know that this did not really happen. And moreover, with the leverage available individual deciders could effectively gamble with far more 'commit authority' than their senior control likely really intended to best at the mid-level. And of course everyone wanted to be a combat colonel, because that's where the glory and the bonus money was; no one wanted to be the staffer lackey saying "No." (It's an imperfect analogy, but one useful; others are invited to fine tune it.) [cont.]

  17. Richard Kline

    So what are the lessons? 1) Opportunity capture requires that commit-authority be pushed down to mid-level deciders directly engaged with the context. 2) Total commit-authority for any one mid-level decider needs to have a HARD CAP. Even a tenth is a lot. And not a farthing more, opportunity be damned. Unless senior control directly authorizes it in full understanding of the position, and even then they will be literally betting the firm on the outcome. 3) Operational _control_ has to be vested in a Staff. As far as I can tell, ibanks do not do this. But this is the organizational innovation which they should pursue. There needs to be an operational center whose sole charge is to know the whole picture and vet commitment. Its members do not need to be senior, but they must have all available information, and their decision has to be final subject to senior control intervention.

    And regarding 3) this _must_ be a Staff, not a General or two and a few analysts. There are simply too many little pieces to track, one has to have bodies to track them. In the era of gunpowder armies before the machine gun, there were never any two as good as Napoleon and Berthier working together. With 50k men, they and the talented 'commit authority' mid-level below them simply couldn't be beat. With 75k men it was highly unlikely that they could be beat, and still they adapted faster than anyone else. But at 100k, they started to lose control. The didn't know where all their resources were. And even when their commit authority folks made decisions at the interface, good or bad, Napoleon and Berthier increasingly were unable to capitalize: the scale was too big. At well above 100k, their control ability failed, and defeats, de facto and real, became increasingly frequent though little else had changed.

    Take the example to today and in finance. Soros may be able to keep track of all of his positions, or him together with one or two senior partners. And in that context, his talent triumphs. I doubt that Buffet could now of his enterprise. We may be sure that he is the general working off of a Staff model.

    Ibanks need a Staff model rather than a colonel model. It will take a massive cultural change to achieve this, but it is necessary.

  18. Richard Kline

    To complete the analogy above [sorry, had to take lunch], the culture of the ibanks was set in the 'Napoleonic Era,' the equivalent of the 50k to 100k enterprises. There, a tiny handful of senior controllers could, if they knew how to 'colonel' very well, keep tabs on all risks and opportunities engaged. In that model, it made sense then to push the power and the rewards down to the real 'colonels' and let they engage. But once the banks grew to be 500k enterprises, or twice that with some of the modern behomoths, this system is incapable. Colonel level deciders at the ibanks are no longer controlled, nor in that system controllable. Their are senior controllers, but they don't have the Staff level to really keep the chain of information solid with the colonel-deciders. At most, there's the equivalent of a tiny bean-counter quartermaster/risk assessment shop who's remit is to make sure that the colonels have enough beans and a road to retreat by (in principle). Not effective.

    The power has to be put in a Staff whose remit is _not_ to make book or dough for the enterprise but to know where the enterprise is, what it is doing, and who is doing it. And the Staff doesn't report to the colonel-deciders; it doesn't even work for them. The Staff reports to senior managment. In fact senior management gets its information _from the Staff_, and then checks this with the colonels, not the other way around. See the difference.

    Financial enterprises need a command systems redesign. Or they'll keep killing themselves whent he colonels committ their all for personal glory and reward.

  19. skippy


    With your leave, may I add a story of woe.

    On an E&E (escape and evasion maneuver) during peace time at 29 stumps USMC base, a Company of Rangers came under direct mortar fire. Well the Company Commander took exception to the predicament and decided to have a tanty with the Battalion Commander (rounds incoming whilst on the phone via PRC77), well long story short the Capt was told to get his ass out of there forth with, unfortunately the Capt was more interested in assigning blame to the event than the care of his troops.

    Never did see him again, gone as a vapor dispelled by a mighty wind. In that environment you lived or died, no fools were suffered.

    Skippy..whom is choking your wind eh.

  20. Alan


    Even in your supercarrier and nuke examples, how often do those decisions come up?

    Thousands of times daily during active operations. I'll add that these huge integrated systems of people and barely contained technology function by design at the extremes of human endurance.

    And it only takes one piece of ordnance from one aircraft to start a cataclysmic event.


    This incident began with one (1) five inch Zuni rocket getting loose from one aircraft. A carrier can fly hundreds of sorties (fuel/arm/launch/recover)over a 24 period with each aircraft carrying dozens of pieces of ordnance per sortie.

    After heroism and selfless devotion by the crewmen, the only reason the Forrestal didn't explode and sink was the application of previous lessons learned during World War II.

    However, these are ever rarer events. So are nuclear submarine losses following the Thresher and Scorpion incidents.

    What you had in your industry was an entire fleet of supercarriers going up at once as a result of what should have been routine operations. The carriers were named Bear Stearns, Lehman, AIG, GS, Morgan Stanley, Citi, BoA, Wachovia,…).

    Nor are the causes particularly obscure.

    And it's barely two decades since the S&L crisis began. The current "Black Swan" began manifesting itself only a decade after the RTC closed shop. iow, the present sheaves of rotten paper started being printed just a few years later.

    We could go back a few more years to the early 1970s and the near collapse of "Wall Street" then. The same combination of short term greed and institutional mismanagement manifested then, too. Perhaps you've read of the immense back-office problems in controlling transfer of securities.

    And this same bureaucratic negligence was repeated in the securitization process of mortgages, viz the frequent inability to locate original documents during foreclosures.

    Meanwhile Navy firemen recruits continue watching training films about the Forrestal fire.

    You do extremely good posts at times. But in my opinion this one arguing for some kind of investment banking exceptionalism isn't one of them. otoh I'm sure it will play well with much of your target market for your book. Wall Streeters are well known to suffer from chronic narcissism. Shameless flattery is a successful marketing technique with such groups.

  21. Richard Kline

    So Alan, you have the scale wrong, m'friend. At least. Bear Stearns wasn't a 'supercarrier.' It was at least a Fleet, even more accurately a theater command. The carrier equivalent would be a single trading desk, at best. And yes, 'blow ups happen': If one of those traders sets his stop-loss wrong or worse drops a decimal, he can blow up his position on a day. I'm sure this happens in reality more often then we outsiders are allowed to know. Ibanks have in principle invested a fair amount of time and resource to prevent this or limit the damage. This is the level that risk management is more focused on with modest success.

    The real problem is compounded by greed and avarice, yes, but not caused there in the first instance. The problem Yves is getting at here is a _command system problem_, where the theater commander has lost control of, even track of, his supercarriers, and many of his other vessels. And where a single below-decks error in one of them is, in principle, not contained to that vessel/desk alone. If some idjit punches the launch code on a flyboat, the whole theater/enterprise can receive a full fleet salvo in return. You see the difference?

    And to me, the reason why this command system problem has not been better addressed is not simply a function of the speed and volume of 'opportunity capture' on a given trading desk. That is a major problem, and part of the 'exceptionalism' which Yves correctly attributes to this problem. The deeper issue, to me, is a cultural one within the ibanks. There culture favors the colonels/buccaneers, not least because those are the folks generally promoted to senior control. They need to go to a Staff model where the Staff has enough eyeballs, talent, and mission to keep a close chain on the operations cowboys directly engaged.

  22. bob

    Alan, the difference as I see it is that people in the military are armed.

    Not only do you have to answer to the chain of command, you also have to answer to the guy next to you. If he is going to do something out of the chain of command, and put his men at risk also, he very well may be shot in the back.

    The solution, arm ibankers, and make sure they are all in the same room. Problem would be solved pretty quickly.

Comments are closed.