Home | Topics | stress test


Printer-friendly versionPDF version
basel carnival.jpg
Stress-tests are supposedly harmonised, at least for banks in the banking sectors of each country, but not across all EU member states, to test for shocks to a country's banking system when current risk assessments are suddenly inverted by falls in investment, in consumer demand, savings, loans, and by fast rising long term bond yields, and when presumably wholesale financial markets become 'one-way' (either buy-buy or sell-sell)!  The indicator values of sudden change are provided by the EU regulatory authorities, but in a narrative form and using an abstractionist, sometimes over-coy, language that is farcical.

This is not how the CRD and BIS Basel II regulations and guidelines specified the laws for this.

In the original regulations, stress-test calculations are expected to be conceived by the banks from their own macro-economic models including their own models of their banking sectors and of how and where their banks fit within those sectors. Instead, the regulators of today provide precise values by which asset class prices and some macro-economic variables are expected to adversely move according to a shock scenario, according to a narrative that cannot make sense to bankers or economists! Moreover, they do not provide models for how to assess the whole scenario. Consequently, banks must assess balance sheet impacts on a line by line basis and not worry about whether the overall picture or its timeframe is realistic or if there are allowances for effective risk mitigations, merely the summing of capital losses (nominal, as if crystallised instantly in full).

The reason for this is that banks proved to be extremely reluctant to invest time and resources in building macro-models of their business and of their banking sectors, largely because they also wished strenuously to avoid accepting responsibility for what happens in  underlying macro-economies, nationally and internationally. They are suspected of being likely to game any degrees of freedom given to them to try and intelligently work it out for themselves. This way no one will learn anything new and really useful.
Banks also strenuously resist, resent, and fear the political implications if governments or financial authorities are able to command the banks to alter or restructure their 'risk appetites', in effect to dictate their choice of lending priorities. If banks do not possess the tools to interpret instructions from above intelligently they think they can ignore any suchlike instructions from regulators, and often, in fact almost always, do ignore or reject instructions to change their risk profiles as a matter of political principle. 
For example, if governments or regulatory authorities such as central banks do ever order commercial banks to lend less to property and more to manufacturing or vice versa, or less to big firms and more to small firms, or more in cash-flow support of borrowers and less in fixed term loans, the best defense is to be able to claim ignorance of precisely how much is lent in these categories or to obfuscate the data, or to claim that disclosure of penalties income, trade finance, factoring, or over-the-counter transaction volumes, are commercially confidential and cannot be disclosed anyway. 
Greek, Spanish, UK, Belgian, Dutch, and Irish regulators have many times in the last 15 years precisely instructed their banking sectors to make major changes to their lending patterns only in every case to be individually and collectively rebuffed or simply ignored.

Few senior bankers ever read their regulators' quarterly stability reviews! So long as the law is not resorted to to force banks to do what the law states they must do, e.g. be more risk-diverse, or if regulators do not use the full powers the CRD law granted to them, and no precedents are set by court decisions, no fines, no penalties, then banks will continue to refuse to accept instructions from governments and regulators. 
This is a surprising situation after nearly a dozen major banks were nationalised, merged and broken up in the credit crunch and sovereign risk crisis and after one of the biggest in Europe (Fortis) had its license summarily cancelled. 
Banks will and do agree however to participate in stress-tests, something that both Bank of England Governor Mark Carney and ECG President Mario Draghi take very seriously they say as their first priority to progressively better ensure banking's future stability. It is extremely doubtful if the banks take these tests seriously, except for the idea that individual stress-test reports can be commercially sensitive with implications for their share price and bondholder confidence.

Personally, my own portfolios of bonds and deposits benefited (were saved) to the tune of tens of millions because of bank rescues. I therefore, and not only as an ex-banking risk officer and economist, take very seriously the failures of banks to learn from, and make honest, best efforts to comply with, both the spirit and the letter of the CRD law.

Today's stress-tests are not nearly as comprehensive as the CRD dictates. No stress tests, for example, include figuring out how much in secondary market transactions by value per bank could fail in a crisis, such as occurred in the wake of the closing of Lehman Brothers, reputed at the time to be over $4 trillions between USA and Europe. 
The stress tests do not address shock unwinding of derivatives positions or the concentration and timing risks of borrowing short from money market funds to lend long in mortgage loans, for example. Yet, such issues are in the law governing stress tests, in Basel II and the CRD.  
Stress tests are a legal and mandatory requirement in Pillar II of Basel II (also known as the Capital Requirements Directive, CRD, which is both national law and EU law).

The current simplified stress-tests require  banks to take latest macro-economic estimates of growth, unemployment, prices, and property values, provided to them by the regulators, to make projections for 2015-2016 and then calculate what it would mean if there was another sovereign risk crisis (expressed by fast-rising long term bond yields) with attendant foreign exchange shocks, commercial property falls and higher than expected unemployment etc. 
The stress-tests do not specify the equivalent of a credit crunch when money market funds dramatically raise their loan rates far above banks' interest rate margins. 
Generally, banks should be tasked to assess how their balance sheets perform over an economic cycle, and especially in severe downturns, also called 'adverse scenario' shocks, either a few shocks at a time, or in a well-established sequence, or all at the same time? The current stress-tests are characterised as mainly about asset-quality and capital reserve resilience, not about liquidity risk mismatches or quality of liabilities, and everything happens suddenly at once? 

The current stress-test designs of the ECB et al. are travesties of reductionism, falling far short of the original requirements as defined at law. The original law requires banks to build models using many years of historical data, at least over one, preferably two, economic and credit cycles. This requirement and its various implications, including the full context of Pillar II, are not present in either the recent or current US or EU stress-tests? 
These tests assume banks are risk-takers, not risk generators, merely passive recipients of economy shocks and not part, if only collectively, of what generates the shocks. Regulators, to specify such stress-test scenarios, should properly begin with the banking sector data they already possess in abundance and advise the banks of the implications of stress-tests across the whole of each national banking sector, and only then ask individual banks to assess themselves against those results. After all, whatever happens to a whole banking sector cannot be the responsibility of a single bank, except in criminal circumstances.

Instead, the individual banks are being asked to undertake intellectually challenging tasks such as to consider a cumulative 5-6% fall in national and Euro Area (or EU) real (inflation adjusted) GDP growth? 
Banks are insensitive to inflation-adjusted economic variables - only concerned with actual cash-flows, and just as they are concerned with nominal profits and unrealised losses more than actual crystallized profit & loss. 
Since banks are indifferent to inflation, except where this reflects exchange rate effects, to make sense of the 5-6% figure they have to make a sophisticated analysis of inflation and exchange rates, which is mostly beyond any of them to do so. A lot of the key data will be guesstimated in spread-sheets even in the biggest most sophisticated banks.

The banks do not have macro-economic models except those (of spreadsheet quality mainly) created merely to satisfy past stress-test exercises. These take top-down figures rather than empirically understanding how headline figures are accounted for in national income accounting, and their dependency links to national domestic versus international factors, and not integrated with financial sector, national accounting standard information. 
The Central Banks also entirely lack models (regulators similarly) to account for complex finance sectors and for cross-border dynamics including international financial markets. Such fuller, more comprehensive, and more complete, macro-economic, accounting standards based models require a skill-set that banks on their own, however big - and without one exception I know of - either lack, or resist, employing.

Part of the reason is failure of central banks and other financial sector regulators to develop comprehensive models beyond merely broad design concepts, and also because of the commercial confidentiality applying to in-house stress-testing whereby the work is confined strictly to as few people as possible on a strict need-to-know basis. 
Banks therefore only undertake the minimum to satisfy each item in the list of basis points and percentage rises and fall that EBA or others such as the BoE list in a scenario spreadsheet, and tick-box check-list? 
This is a defeat for the original purpose of Pillar II of Basel II, which is to force the banks to be sensitive to macro-economics and to imbed this sensitivity into all of their day to day operations, and as the basis for fully understanding the much bandied-about term, 'risk appetite'..

Stress-tests have become understood as onerous and artificial and not about anything  that could happen anytime soon, and merely about the solvency of banks. 
But that idea falls far short of the full scope and depth of Pillar II as understood, defined, and specified at great length by the BIS committees, which is also about forcing banks to make and maintain long run macro-economic models whereby banking sectors, and each bank's own circumstances within that sector, can be understood and monitored by each bank continuously.

Pillar II and stress-testing are about congregating the effects of all risk types (not just credit risk), to develop the ability to monitor risk diversity across the whole of the underlying economies in which any bank operates, as a governance principle, and, not least, to understand and measure the role and potential for banks to act counter-cyclically thereby taking the pressure of lenders of last resort and off the shoulders of governments when recovery out of recession is needed. 
It is part of the needed effort so that governments and central banks should not have to respond entirely unaided when tasked with digging economies out of recession holes or out of sub-optimally low growth. It is, of course, unsurprising that commercial banks resist sharing this burden for fear it will blunt their profitability or change their business model and bonus rewards system?

The adverse scenarios for stress-testing are currently merely abstracted, simplified, versions of what was officially acceptable as explaining why what happened in the 2008-onwards banking crisis and 2010-onwards sovereign risk crisis.

An example of EBA/ESRB narrative trying to express the sovereign risk crisis, if replayed in 2015-16:
"The proposed adverse scenario reflects the systemic risks that are currently assessed by the ESRB (European Systemic Risk Board) General Board as representing the most pertinent threats to banking sector stability: (i) an increase in global bond yields amplified by an abrupt reversal in risk assessment, especially towards emerging market economies (EMEs), and pockets of market liquidity; (ii) a further deterioration of credit quality in countries with feeble demand, with weak fundamentals and still vulnerable banking sectors; (iii) stalling policy reforms jeopardising confidence in the sustainability of public finances; and (iv) the lack of necessary bank balance sheet repair to maintain affordable market funding"

To both experienced macro-economists and bankers this guidance is not just over-abstract, but actually ludicrous and evidences the mindset of people with zero economics training, lawyers perhaps?, certainly not primary or secondary markets or macro-economics professionals. It is extremely primitive compared to the guidance provided in the original Basel II Pillar II papers. To turn this into good-sense requires genius-level interpretation. For example, who is capable of fully expanding on the expression "...and pockets of market liquidity" when we don't have any official data for markets liquidity? Do banks have good data at least of their internal "liquidity pockets" asset & liability class by class... er, no they absolutely do not!  If even they had that, with some standards created for this today, would they have it on a consistent basis over enough time from which to draw any realistic or reliable conclusions... er, again, very unlikely.

The ESRB guidance continues.
"In line with this ranking of risks, the scenario narrative takes as a starting point a rise in investor aversion to long-term fixed income securities, which results in a generalised re-pricing of assets and related sell-offs. In particular, this causes US long-term interest rates to rise, setting in motion a global increase in long-term bond yields, a steepening of yield curves and an additional market tantrum in emerging markets. This affects particularly the group of countries identified as the ‘Fragile Five’ and other BRICS."
But, there is no reference to ratings? We have over 20 major risk grades for fixed income - what effects are being suggested? - all fixed income falls in attractiveness world-wide regardless of ratings? Quite apart from a silo-mentality assessment of individual asset class effects and ignoring the many attendant factors that give rise to such changes, and that the world is various and does not move as one, and ignoring the fact that relative volatilities are commonplace. 
Above all, where is the consideration, beyond merely measuring for risk-weighted assets, that one essential aspect of good banking practise is risk-diversification across asset (& liability) classes and risk diversification also within asset classes and geographically. 
The narrative above does not make sense, not least the  cute use of an eccentric phrase, "...additional market tantrum..."  The phrase 'Fragile Five' clearly means PIIGS (Portugal, Ireland, Italy, Greece, Spain). 
Why should there be any anxiety about government bonds within the Euro Area any longer given all the standby funding power now set up to guarantee their credit rating? 
Surely, one implication for  a renewed sovereign risk crisis would have been failure of the European Stability Mechanism Treaty system, possible break-up of the Euro Area, and, before that, major banks collapsing needing to be refinanced on budget of national governments, i.e. not by the ECB i.e. ECB refusing to regrow its balance sheet by a couple of trillion Euros as it has done before, only very recently?  
Bond yields have converged and historically sovereign risk crises do not emerge until at the tail-end of a recession shock, not at its outset?

Where is 'investor aversion' to long-term fixed income securities possibly coming from in the minds of the ECB/ESRB. Is it from pension and life funds? Is it because of expectations of sharply rising inflation? The guidance can't explain that. 
Perhaps its authors daftly believe inflation follows long term bond yields (via secondary market pricing) as a uniform global phenomenon, and not vice versa, or simply cannot accept there may be no connection, either at some time or any time, or are we dealing in car-crash coincidences? 
Or, are we supposed to read into this 'long term government bonds' issued by certain countries only are in freefall, not others? Why would US long rates rise in response to others, if surely it is just as likely to do the opposite given assumptions about relative exchange rates!  
It is daft to speculate about a generalised loss of confidence and price fall in all fixed income or all sovereign bonds spreading worldwide? Apart from anything else that entirely ignores the value of risk diversification.

The imaginative narrative would  fail to win a good mark if even only a first year undergraduate essay. It continues. 
"These financial disturbances have  further important real economy  spillover effects, especially for emerging market economies (EMEs).The latter suffer from sizeable capital outflows, in a form which is similar to a ‘Sudden Stop’ episode, in which countries are excluded from international capital markets since they are perceived as too risky. Their internal demand then experiences a sudden fall. Overall, the negative effects, worldwide, of the financial turmoil on the real economy imply a marked deterioration of foreign demand for EU exports, putting significant downward pressure on GDP growth as a result. The global financial shock also acts as a trigger for all three other, EU domestic, vulnerabilities. This leads in particular to a further weakening of EU real economic activity, re-differentiation of EU sovereign bond yields according to associated perceptions of sovereign risk, with associated funding difficulties for respective banking sectors."  

Somehow the story gets from  a supposedly unexpected "aversion to long term fixed income securities" to "worldwide financial turmoil"!  Even the stupidest of bankers can only laugh at this? 
For example, foreign demand for EU exports variously employ between 4% and around 10% of jobs. Yet we are expected to believe a fall in demand among EMEs will significantly harm GDP growth? But trade patterns are far more resilient and risk diverse and complex than this supposed conjunction. 
One has only to consider Eurostat's latest external trade report. http://epp.eurostat.ec.europa.eu/cache/ITY_PUBLIC/6-18032014-AP/EN/6-18032014-AP-EN.PDF. 
This includes a paragraph on total trade for individual member states (including intra-EU trade) whose net effects are far greater than ex-EU trade. 
Scenarios would be far more realistic to take account of intra-EU capital flows and trade effects following from recessionary or market failure shocks than external ones. But, perhaps this is where certain realities become politically too hard to specify in EU stress test scenarios even if far more realistic?

While the so-called stress-test narrative is obviously very artificial and, in proportion to how the EU economies inter-work, actually nonsensical, the quantitative implications derived by the ESRB are surprisingly precise, country by country such as residential property values falling from 9% to 34%across all EU member states, and commercial property by between 4% and 27% (even if usually commercial property falls faster than residential in a recession), while unemployment rates rise moderately for some and almost double for others. 
These changes are only against rising or falling baseline trends, some of which are questionable when checked with the latest official data, but no matter that. 
In any case, there are exchange rate and other factors, and one wants to ask what of these explain why the 18-country Euro Area aggregates appear in every case better than the 28-country EU aggregates, implying that being outside the Euro Area means much higher % shock effects?

It is certainly therefore is not the case that banks' balance sheets for country by country domestic asset exposures are all being tested against the same adverse scenario percentages - and presumably because of nonsensical assumptions in the narrative above. 
This means the implications for regulatory capital and economic buffer reserves will not be the same across all countries if only because of very questionable, certainly simplistic, assumptions among which it is easy to note that German banks especially are faced by less testing shocks to their balance sheets that others such as the UK. 
This must be pleasing to the major German banks whose actual current exposures involve much narrower and more fragile (in profit terms) spreads and interest margins than for example UK domestic bank lending.

In conclusion, it must be hard for banks to take these stress tests seriously as it will also be hard for some countries not to suspect that there is a political or national bias operating in this system of stress-testing, however unrealistic or not.


Financial regulation, especially governance by law, is ultimately dictated by who puts the most money on the table. In Europe, this means The Bank of England and The European Central Bank. 

The ECB has the central role of integrating the landscape of European banking regulation within the Euro Area while the BoE takes care essentially of off-shore branch banking and wholesale financial markets primarily centred in London, together worth up to half of Europe's banking volumes. 
EU-wide direction is also given by the European Banking Authority (London) and The European Systemic Risk Board.  
BoE and ECB are banks, but both are themselves outside of the Basel II/CRD laws. The regulations they oversee do not apply to them. They do not apply to any wholly national-state owned banks. The same solvency tests enforced on non-state banks do not in any case make sense if applied to central banks, least of all to the ECB and BoE. 
This is worth noting also because they stand in certain respects in a contrary risk position to all other banks. The higher the reserve requirements imposed on all other banks the higher the liabilities that are forced onto the ECB and BoE balance sheets.

The ECB, whose primary function is banking stability within the Euro Area, is a central bank, but multi-state owned, a product of the Maastricht Treaty that set the criteria for membership of the Euro Area, the Stability & Growth Pact (articles 121 and 126 of The Treaty on the Functioning of the European Union) out of which also came the European Stability Mechaniism Treaty (amended article 136). 
This was originally intended to furnish a Euro500bn fund of intra-Euro Area state aid to secure the solvency of the Euro Area banking system, but only in part. The major financial support is provided via the ECB balance sheet that between 2006 and 2012 grew sixfold to three trillion Euros, half of this to compensate banks for loss of deposits due to private capital flight from southern EU states to 'the north'.  
As bank stocks have recovered and some capital flight reversed the ECB balance sheet has shrunk in the last year to two trillion Euros, or roughly 12% of all major banks' domestic assets (loans).

According to regulatory law, wholly state-owned banks do not have to comply with the regulatory laws of Basel II. The ECB, in this regard, is an anomaly because not nationally but multi-nationally state-owned, a category about which banking regulatory laws are unclear? As the Euro Area's supra-national Central Bank, the ECB is the ultimate stability guarantor (lender of last resort) and regulatory authority for stability issues (acting through the European Banking Authority, which covers all of the EU). 
And for the Euro Area it is the designer and evaluator of stress tests of 124 banks, or roughly half of each national banking sector. The BoE has deepened and changed some of the percentage shocks compared to the ECB/EBA/ESRB figures? Why it should dothis, given its own lack of any realistic model, I don't know, especially when it seems to me the percentages applied to the UK share of the scenario are already exaggerated? The BBA should object!

Short URL
Asymptotix on Twitter

Are the key legislative pillars such as Basel II & III, UCITS IV and Solvency II forcing you to re-examine how you identify, measure and manage risk and capital?

Asymptotix work closely with our partners to help clients develop a more proactive, systematic and integrated approach to governance and risk management to deliver proper value.

Asymptotix can offer the support you need to deliver on time. Read more...

Is the goal of your website to sell services or products, educate, or collect data?

A positive customer experience is vital to conversion, no matter what your conversion goals may be. Our designers and developers will create a positive experience to maximize your conversions and deliver the optimal return on your investment. We strive to find the perfect balance between the web site’s design and functionality.

Asymptotix implements interactive solutions for European companies. From corporate websites to social communities, our clients will tell you an investment in building a scalable online experience will deliver long-term tangible benefits.

Based in Luxembourg we can help you all over Europe. Our multi-lingual team can work with projects and speak your language! Read more...