Monday, 9 July 2012

On Banking 'Culture'

Within the press and blogosphere much has been made of Bob Diamond’s note which implies senior officials at the Bank of England and in Whitehall gave Barclays implicit consent to under-report their LIBOR submissions. Such an event, if proven, would be devastating to popular trust in our political and regulatory establishment. For decades our elected leaders granted finance unprecedented privileges and freedom to maraud. The outcome of that freedom was a growing sense of invulnerability that still pervades the trading rooms of major banks today. So it is important, with the mute sense of collusion between regulators and regulated still lingering, to understand these internal cultures and behaviours within the structures that nurtured them.

Last Wednesday’s timid Treasury Select Committee meeting with Bob Diamond was a typical affair, with lots of self-righteous indignation but few probing questions which got into any level of analytical detail about the structural pressures and context of moral laxity. The refrain of the committee, and most likely any Parliamentary Inquiry that may follow, is that banks have a ‘cultural’ problem, and that this needs to change. Like giddy, overexcited children who knock things over at parties, it is assumed bankers can be taught to calm down and behave simply by giving them a stern talking and getting them to admit fault; only then can they be pushed back safely into the melee. ‘Bankers have got carried away’ we are told ‘and so need to take a long, hard look at themselves’.

Cultures set the boundaries of what is and is not permissible. Those rules are not necessarily spoken, they are rooted in behaviour and patterns of reward. When a CEO says ‘we want no more of that sort of thing’ and then increases derivatives trader bonuses, it is the reward rather than the words which reproduce that culture. These are the structures of culture. True, it is important to remember there are different ‘tribes’ with different cultures within banks (there is good journalistic and academic work on this), but it is also true that banks are hierarchies which impose a broader and more potent disciplinary logic: to maximise income for the senior traders. Donald MacKenzie in his most recent revisionist work, refers to this as the principle of ‘maximising Day One P&L’ (for the elite workforce), which often engulfs these silo cultures.

This ‘engulfing’ is what we really saw in this LIBOR fixing scandal: the normalising force of a culture which puts individual bonuses (not institutional profits) at its centre. What is perhaps most startling about the findings of the FSA, is not just the cavalier and open way with which corrupt practices were conducted, but the finding outlined in paragraph 8 of the proceedings which states:

 “Barclays acted inappropriately and breached Principle 5 on numerous occasions between January 2005 and July 2008 by making US dollar LIBOR and EURIBOR submissions which took into account requests made by its interest rate derivatives traders (“Derivatives Traders”). At times these included requests made on behalf of derivatives traders at other banks. The Derivatives Traders were motivated by profit and sought to benefit Barclays’ trading positions.” (my emphasis)

That employees from Barclays were willing to manipulate their own institution’s reported borrowing rates, (with uncertain effects for their employers), to benefit an individual at a competitor company says something profound about banking culture and the structures which support it. It tells us that the boundaries within and between institutions are fluid and that banking culture promotes, perhaps even exalts, the maximisation of personal gain, even if that means building and maintaining interpersonal networks across institutions.

So what is the purpose the institution under such circumstances? Classically in the ‘theory of the firm’ literature, firm boundaries are rigid: they exist to minimise transaction costs (the Coasian perspective), or to manage agency problems (the Jensen and Meckling perspective) or to protect and coordinate certain skills and competences (the Teece et al perspective). All of these theories presume that the firm ‘contains’ activity and in doing so performs some functional purpose for the broader economic good. None of them adequately explain banks. Banks represent something different: an institution captured by its elite workforce, where the purpose of the firm is to act as both shield and sword for its captors. The institution provides cover for the manipulation of prices and markets, for the proliferation of asymmetric information from which private gains are made. The current banking firm allows privacy and insiderism to flourish and incentivises strategies of position, disguise and deception. The outcome is akin to what Akerlof and Romer (1994) described as ‘looting’: maximising individual rewards at the expense of the institution when accounting is poor, regulation is lax and there are few penalties for abuse. The individuals take from the trading profits and the institution is left with the liability.

Psychologically, this is a difficult thing to grasp for the many politicians and regulators who for most of the 1990s and 2000s were absorbed by discussions about financial innovations that improved capital allocation and market efficiency and new whizzbang models that distributed risk away from the financial core, which all justified the need for light touch regulation. In doing so they forgot the crucial point that banks are run by people, and that markets involve a series of intermediated formal and informal agreements between individuals. The problem with banking is that the interests of the individual and of the institution are not necessarily aligned. Modern banking is particularly prone to looting, most obviously because the looting can only go on as long as the institution remains solvent, which is a long time when there is a State bailout guarantee. But more elaborately looting is possible because the presence or absence of profit on a particular trade or at the firm aggregate relies on the quality of the numerical inputs and valuation models used when assigning a price to often complex derivative assets on the balance sheet. Banks are conversion centres: they turn assets into income streams and income streams into bonuses. When derivatives traders ask someone on the cash desk to fix LIBOR, that affects the value of an asset which flatters the return on that trade. It produces, not profit, but the simulacrum of profit. And that is looting.

Of course Barclays no longer uses LIBOR to price many of its interest rate products. The game moves on. It now uses much more complex calculations to value its assets. On its valuation of collateralised interest rates it uses:

“Overnight Index Swap (OIS) rates…to reflect the impact of cheapest to deliver collateral on discounting curves, where counterparty CSA (Credit Support Annex) agreements specify the right of the counterparty to choose the currency of collateral posted”

And on interest rate derivative cash flows it uses:

“…interest rate yield curves whereby observable market data is used to construct the term structure of forward rates. This is then used to project and discount future cash flows based on the parameters of the trade. Instruments and optionality are valued using a volatility surface constructed from market observable inputs. Exotic interest rates derivatives are valued using industry standard and bespoke models based on observable market parameters which are determined separately for each parameter and underlying instrument. Where unobservable a parameter will be set with reference to an observable proxy. Inflation forward curves and interest rate yield curves are extrapolated beyond observable tenors”.
Barclays Annual Report & Accounts, year ending 2011, p234

Given what we know from the past week in banking, we are entitled to ask ‘what does that mean’? Do these accounting values from which revenues are booked and bonuses paid reflect the underlying reality of their financial position?

When you price assets using inputs of uncertain verity which are run through models of great complexity, it is tempting to assume that such convoluted calculations are designed to avoid writing down the value of assets. This would have an income effect and reduce the pot from which bonuses are paid. Do these accounting numbers present the simulacrum of solvency, which avoids the day of reckoning and allows the continuing payment of high incomes to senior bankers? It is something of a curiosity that just as the economics profession has become more and more concerned with finessing ever more complex models, so the financial sector has become acutely aware of its own sociology: a reflexive understanding of the signalling power of numbers. These are active numbers reported to elicit an effect in the future, rather than passive numbers designed to faithfully depict an image of the present. Thus LIBOR is not the rate at which you access unsecured funds, LIBOR is the signal to investors that everything is under control, or that a trade has been particularly profitable.

For many years here at CRESC we have had a growing sense that the instruments of finance have been put to different use from that originally intended. It is quite another thing to think that the most fundamental institution of the capitalist world, the public limited company, is itself being arbitraged in the interests of its elite workforce. But when a bank becomes a tower of assets built on the quicksand of confidence, the incentives to signal good news through numbers is great. It may well be that the more fundamental crisis going forward is not whether there was collusion between regulators and regulated, but whether – again – there is a collective loss of faith in the quality of the accounting data produced by these institutions.

This takes us back to structures. The culture that arises within banks is the product of particular structures, or rather the absence of structures, at three levels: accounting systems, organisation and activity level.

It is clear that the accounting numbers need to better represent the financial position of these organisations. There is too much leeway granted to these institutions to value their own derivative products – it gives rise to huge conflicts of interest. Much of this problem emanates from the pursuit of Day One P&L, which is largely an accounting phenomenon. As a basic start we should reconsider accounting rule HKAS39 which allows this practice to continue. As we have argued elsewhere, if costs were booked upfront and risks calculated non-normally it would mean most products would book immediate losses and only produce profits later in the product cycle. It would therefore tie in bonus pay to the long term performance of the particular products created. Similarly the structurers of those products would have to contemplate counterparty risks going forward, and thus think reflexively about whether they were passing on ‘too much’ risk to others – rather than current practice which is to maximize volume and pass off risk via a swap and consider it ‘somebody else’s problem’.

At the level of the organisation and activity, it is naïve to assume that banking culture will embrace restraint and responsibility without significant structural reform. That reform must take as its core organisational principle that banks must become public utilities with the duty to serve the wider economy. This is a considerable political challenge, and one that must begin by uncovering the patterns of patronage that fostered ‘light touch’ and nurtured the culture of invincibility in the City. Only then can an honest review take place of what needs to be done. Our recommendations are outlined here, in our Deep Stall Paper.