Article published in « Revue d’Économie Financière » n° 94 – june 2009
Today, two obvious facts clash. In the first instance, financial markets are not self-regulating. In global and deregulated finance, they lead unavoidably to crises which are violent to a greater or lesser degree and which intensify, or even spark cycles in the real economy, as much in their euphoric stages as in periods of depression. The second fact is the essential character of these same markets which allow for the reallocation of risks (interest or exchange rates, for example) and which allow, in tandem with the banks, the adaptation of the needs and capacities of global finance. Today the banks alone cannot assure the sum total of financing the economy.
That is why one conclusion is abundantly clear : the need for adequate rules and reforms of a diverse nature enabling us to limit the intrinsic instability of finance, as we cannot make it disappear. We must take care, with the subsequent return to a new period of euphoria, to ensure that these reforms have begun before we rush to forget the recurring lessons that each financial crisis shows us.
The reasons for the inherent instability of finance are increasingly well analysed. They reside in the underlying nature of a financial or property asset, for which the price is not determined by its production cost, following the example of a reproducible good or service. In fact its value corresponds to the estimation of a promise of future revenues that the asset in question will bring. And yet, this forecast is very uncertain in as much as the future is hard to predict in a decentralised and monetary economy. In fact, the evolution in household savings levels as the multiple interactions of private competition and complementary economic agents renders their respective successes or failures very difficult to foresee, and even more so to quantify.
Here economists talk about a situation of radical or fundamental incertitude, but the forecast is equally uncertain because the exact risk to the issuer of the asset in question (for example shares or bonds) is not known to its holder. In effect they possess accurate information on neither the issuer’s current situation nor what their future actions will be, therefore ultimately deeply altering their risk profile. This information asymmetry and fundamental incertitude leads to a profound difficulty in knowing the equilibrium prices, that is to say the “normal” prices, of the financial assets for all circumstances in all likelihood, so facilitating mimetic behaviours in them and creating bubbles. Add to this the capacity to forget the effects of previous crises during euphoric periods and you have economic agents increasing their debt levels, thus pushing the leveraging effect to such a level that it threatens their financial situations, and during periods of depression they seek desperately to reduce this debt, thus considerably worsening the economic reversal. In other words, this phenomenon increases further when borrowers no longer gauge the solvability of lenders by the yardstick of their likely future returns, but by the expected evolution in the asset prices (for example shares or property) which are so financed or which act as a guarantee.
In addition, exogenously there are rules on the remuneration of the involved parties and accounting and prudential standards which can end up adding to the instability and procyclicity of finance.
All reform projects must therefore aim to combat the endogeneous as much as the exogeneous causes of financial instability. To tackle the exogenous causes is without question the least arduous task.
Tackling the exogenous causes.
The IFRS (International Financial Reporting Standards) accounting standards have given priority to the assessment of assets in terms of their fair value, essentially based on the market price. This decision is based on the hypothesis that at any given moment the market price is the best indication of the “real” value of a given asset. And yet, today’s major credit crisis has shown, as if proof was needed, that while the market is fading under pressure from sellers, in the absence of buyers, prices are falling beyond all fundamental reality. In the same way and symmetrically, while we are in the middle of a speculative bubble the market price is totally dissociated from all equilibrium value. It is therefore necessary, as was the case in 2008, to be able to reasonably estimate value in an asset assessment, when the market does not allow it. Without this, the accounts depreciation leads to additional sales strung together one after the other in a self-maintaining flow towards low prices, and symmetrically in the event of a rise. In the event of market failure it is then necessary to use other methods than fair value to evaluate an asset. Avoiding returning to the method of accounting through historical value, which can be misleading in the case of assets held in trading, it may be useful to move to mark to model, provided there is external control of said methods, or the simple updating of reasonably expected future cash flows.
Furthermore, in contrast with the effect of the IFRS standards, in order to reduce the procyclicity of credit it is first highly desirable to encourage the banks’ supply. If they can accountably fund in advance as yet unproven future risks to their credit, they are less obliged to reduce their credit production during the occurrence of a major economic downturn. The impact of their accrued losses due to the increase of the cost of credit risk to their shareholders’ equity is in fact then compensated for, at least partially, by their provision write offs. Finally, it we should to re-examine the virtues of the old accounting framework of the banks on one point: that which would allow for the accumulation and discretionary provision write offs for general bank risks.
In the same way, the Basel 2 prudential standards are themselves procyclical. The bank shareholders’ equity required by Basel 2 is proportional to their credit liabilities in particular, themselves weighted by their associated risk. At the same time the positions in the financial markets are also considered according to their own risks (the value at risk method).
With evaluation models for these risks being essentially based on the data from a few previous years and hardly taking into account extreme risks, as much to credit as to market, a euphoric economic period leads little by little to more credits and speculative positions for the same amount of equity capital, and so further heightening the euphoria. While a period of depression forces banks to slow their credit rhythm or reduce their market positions for a given amount of equity capital, thus reinforcing the depression itself. It is also essential to modify the risk evaluation methods and the length of past time that they take into account, or moreover to apply stress scenarios to the models, enabling them to take into account more extreme cases (decomposition of risk factors and application of independent shocks). In the end, with identical models following the example set by Spain, the solution is probably to adapt the ratio of required equity capital itself, according to the economic phase in progress, enabling it to be raised during a boom period in the cycle and lowered when there is a reversal permitting it to play a contra-cyclic role. In theory the second pillar of Basel 2 allows monetary authorities to proceed in this way but in practice, in the absence of clearer and better shared rules, it does not function in a satisfactory and coordinated way.
The question of the way in which the rating agencies work is also at the heart of the issue. Their procyclical nature is also obvious here. Furthermore, the CDO (collateralised debt obligation) rating is not the same as the corporate. The evaluation models for ranges of securitisation have failed, and not just because they did not integrate liquidity risk. In addition the fact that these models used data collected over too short a time period, they took little or no notice of the non-linear effects linked to the threshold effects, themselves due to the successive bringing into play of risk in different ranges of securitisation. Moreover, they have not appreciated the correlations in the flaws of the different components of the supports of securitisation.
In short, it is crucial to enforce that the marking agencies be obliged to show or make shown due diligence in the underlying securitisation, which is not the case at present (for example the cheating on sub-prime credit documents stems from this).
On another level we should add that these agencies are paid by the issuers who need their rating, which could lead us to doubt their impartiality. However, because its users are spread out and of very unequal size, it is impossible to conceive a viable system based on a payment from these users, so the choice is either to nationalize these agencies, claiming that they provide a service for the common good, or more likely we put them under a supervisory organisation which checks the quality of the methods used and the results after the event, so respecting proper ethics.
Likewise, as this has been done with external auditors, it would be prudent to establish their civil responsibility in case of an error in their rating process in counting on the jurisprudential control to further assure that their method of payment does not influence their decisions. In the end, in the same sense it seems absolutely necessary to separate their rating and advice functions (advice in terms of preparing for a rating).
The question of trader compensations is also decisive even if we cannot in any way make them the principal cause of the current chaos. Bonuses, paid annually, represent extraordinary amounts on an individual scale and are in principle based on the achieved earnings thanks to their trading positions. This compensation system is totally asymmetric because it does not erase the previous bonuses in the case of a final loss. Thus, it is a strong incentive to take significant risks. At the very least it would be essential to only calculate and transfer the bonuses once the positions are finally released. But above all, because more and less favorable phases in the market follow one another over time, even disastrous phases as is the case now, we might say that the main part of the bonus may only be paid at the end of three or five year cycles for example, thus encouraging more long-term behaviour in traders. Without doubt it would be equally wise to limit these same bonuses to a multiple of their fixed salaries, not only as a question of social equality but also and above all to avoid unreasonable professional behaviour induced by abnormal sums.
Lastly we can remark that these compensation systems could be examined by supervisory bodies when looking at prudential solvency ratios. In effect it is likely that only banking self-regulation cannot manage to settle down the necessary new system of compensations once inter-bank competition is once again strong in this area.
Facing up to the endogenous causes.
To face up to the endogenous causes of financial instability is, less comfortable. A certain number of trails must therefore be followed.
Let us begin with the easiest path to apply to this end; monetary policy. As many central banks tell us, in the first place it is extremely difficult if not impossible to use interest rates as a weapon to slow or stop the emergence of euphoric phases in asset markets, because it is also the level of intervention of central banks that enables them to influence the rate of economic growth. And yet, slowing growth by an increase in rates is not often desirable even if it would be useful to prevent a euphoric state developing in the markets.
Secondly, the central banks cannot determine fundamental values with certainty and so cannot be sure to spot the beginnings of a speculative bubble. On the other hand there is no doubt that the monetary authorities could manipulate prudential solvency ratios better than they do presently, depending on the phase in progress. In effect, more often than not speculative bubbles on the stock market, as property bubble, come with a development of credit which is too fast in terms of the levels of debt and of leverage. If the debt was not able to increase in an abnormal fashion, than with a tighter control on bank solvency ratios, the bubbles would have less oxygen with which to develop. Following the same objective, the setting of an obligatory rate of reserves can be seen as the perfect companion.
However, the risk remains for the central banks to act at the wrong time.
As we have seen, these bubbles stem from the actor’s ability to develop a strong mimicry – rational in individual examples, but which lead to a collective irrationality – in the absence of reliable bearings as far as their fundamental values.
The central banks then try to speak out regularly, when it is necessary, in order to clarify to the market that prices seem to them to be some way off normal levels corresponding to a fair appreciation of the fundamentals. However, in general these warnings have little effect. Thus, Alan Greenspan spoke about an irrational exuberance in the stock market in 1996. This has did not avoid the creation and bursting of one of the strongest bubbles in 2000.
It may be possible to imagine an independent watchdog, a scientific panel of renowned experts, perhaps linked to the FMI or the Bank of International Settlements (BIS), that is capable of producing public reports on a quarterly basis for example, and which measures speculative tensions in the different asset markets.
Economists at the BIS have updated the fairly reliable predictive indicators of coming financial and banking crises. Essentially they are based on the measurement of the gap between the instantaneous evolution of property prices and stock prices and their long term tendency, along with the level of credits on the GDP and their long-term benchmark level. It is feasible to hope that if such relationships were regularly made and public, with suitable effect, little by little they could influence the creation of agent expectations on the markets. They could also enable a reduction in the capacity for markets’ disaster myopia. This largely shared cognitive bias fits with the progressive desensitisation that everyone has to the risk they are running, which grows little by little as the memory of the most recent of these rare yet violent events (in this case the financial crisis) fades with time, so encouraging behaviours which will ease the advent of the next disaster.
Short termism is inherent in finance since it is rational for fund managers or bank management to adopt or to have adopted a very short-term view in the management of their positions, taking into account the uncertainty concerning the fundamental value of assets. When we do not know what the “true” price is, we don’t bet for long on a convergence of market price towards an uncertain estimation of the fundamental value.
Thus, it is difficult to reduce this short-term view.
Some options, then. Aside from sovereign funds which are not restricted by the short term, it would be possible to aim certain funds towards the long term, for example pension funds, because their outflows can be predicted far in advance. Their accounting should be adapted in order to avoid needing an accounting of short term market fluctuations. In the same way, the rules for the outflows of funds could be revised according to their nature so that, for example, asset funds don’t have a daily liquidity thus lengthening the view of investors and managers.
Limiting this short termism could also be done by publishing fund values and their benchmarking at a reduced frequency so as not to enhance the mimicry of their managers.
Furthermore, in order to encourage individual investors to buy certain funds, for example, an attractive taxation should be implemented. In effect it is more sensible to impose low or nil tax rates, not to wrapped products or funds such as PEA’s in France, but to the direct or indirect holding of funds which are long-term in nature. In fact, even within a PEA it is perfectly possible to buy or sell funds listed daily and benchmarked monthly. Whether held within a PEA or not, these funds are led rationally to adopt very short term views and very mimetic behaviours in order to be sold as soon as they no longer have a high profitability, which could be offered by other funds. A beneficial taxation reserved for long term funds could therefore be a useful tool in limiting the inherent short termism in financial markets.
To tighten supervision is a necessity agreed upon by everyone. It comes through the supervision of up to now loosely or uncontrolled bodies, notably with hedge funds and securitisation vehicles. In effect they behave like banks but have an uncontrolled leverage and risks which are not scrutinised by supervisors. Evidently the same goes for investment banks in the US which for the most part have been helped by the Fed, although they were not supervised by the Fed itself. Added to that, supervision in the US is very broken up. And so for example, organisations which distributed sub-prime credits were not supervised by the Fed.
Besides, it could certainly be useful to envisage a pooling or at least an active cooperation on the part of bank and insurance supervisors. In effect, the circulation of credit risks between insurers and the banks is intense, for example due to the CDS market (credit default swaps).
In addition, a single supervisor, or at the very least federal, would be of great benefit in the euro zone, or even the European Union. With the supervisory bodies being national and the phenomena of financial crisis being worldwide since the advent of financial globalisation, a more forceful coordination in these bodies has become necessary.
More generally speaking, the regulators must pay extreme attention to the capacity of economic agents to raise leverages to unacceptable levels during euphoric phases, sometimes by circumventing the rules or using their shortages (distribution of sub-prime credit in the US, securitisation allowing the banks a strong leveraging effect despite Basel 2…).
A lot of ink has been spilled over securitisation and CDS market. They are certainly the specific and aggravating factors of the current financial crisis. They are, however, necessary since they enable banks to grant more credits than if they didn’t exist. Banks would not be able to finance the global economy solely through their equity capital. It merely remains that they be rethought so that they are at the same time more efficient and more “moral”. First and foremost, the supports of securitisation, such as the CDS, must be standardised. Their current heterogeneity added considerably to the mix-up of markets and their lack of liquidity. For CDS’s, it is furthermore essential to put them into organised markets, with guardianship of the market and clearing house, in order to guarantee a good end to contracts thanks to calls for a daily margin and deposits, which almost allows the elimination of risk of compensation.
For securitisation it is necessary to lessen the moral hazard which comes with them, since a bank which securitises its debts no longer bears the risk of credit it has granted, nor the obligation to monitor the borrower for the duration of the credit. All things which nevertheless normally define the role of banks in the credit process, from selection for the allocation of credit up to its’ reimbursement. To obstruct the possibility for banks not taking an interest in the reimbursement of credit – so as not to play the role of bank – as was seen at a preposterous level with the sub-prime market in the US, an obligation should be imposed on them to keep their risk liability at around, for example, 10% of securitised credits, by clearly indicating in their prospectus the exact risk held by the bank and enforcing a precise reporting next to the subscribers. Finally, it is unreasonable to have allowed securitised bank debts hold by structures (conduits) bearing credit which are often long term, with a very short term refinancing. These conduits are in a sense the ersatz of uncontrolled banks, de facto allowing them to increase their leverage without the same regulatory control. And yet these same banks should give guarantees for the refinancing of their conduits, callable in cases of liquidity problems, so obliging them to retake previously securitised risks. In addition, in that case the disappearance of market refinancing with which the conduits were faced reveals a deep uncertainty as to the quality of the so securitised debts.
Upstream and more fundamentally, major financial instability is often facilitated by global macroeconomic and macro financial imbalances. This is certainly the case with the current financial crisis which has unfolded in the context of very high American current external deficits. These deficits are financed with no limits through official Chinese reserves, themselves due to imposing current account surplus, and are made possible by a long undervalued Chinese currency. It is from here that the discussion on a new Bretton Woods agreement stems, in particular regulating the value of currencies between themselves in a more harmonious fashion. It is unfortunately unlikely that such an attempt will succeed since the national interests in question cannot agree with one another. However it is not useless to look for possible arrangements, even temporary, which could eventually establish the resolution methods for such a divergence of interest in a more coordinated way.
To reduce financial instability is not easy, but, as we have seen, serious and pragmatic paths are open to us. The ideas presented here are certainly not exhaustive. But it is necessary to study them and, if needs be, to accomplish them as quickly as possible. Financial stability is a collective good which contributes to growth and the well-being of all. A demonstration from the absurd is in the process of being given.
Essay written in January 2009.
By Olivier Klein
Professor of Financial Economics at HEC
Member of the Scientific Committee of the Doctoral School of Management Sciences at HEC – Paris 1