Since 1494 and the contribution of Luca Paciolo, accounting is concerned with a higher level of reliability in financial record-keeping and reporting, starting
with the journal of transactions. Accounting, however, is only one of the pillars on which a dependable reporting system rests. There exists no universal method- ology which integrates the other pillars like management ethics and risk control.
But it is possible to learn from engineering, stretching the concept of reliability into the business domain. For instance,
● If we measure financing staying power at 0.9997 level of confidence as equal to AA credit rating by Standard & Poor’s,9
● Then we would require a fully dependable financial reporting system able of providing evidence which confirms that level, not any other level down the grading line.
This concept can be extended to the whole banking industry in a given market.
Reliability theory teaches that for a system with 100 parts connected in series, even if each has 99% mean reliability, the overall reliability will be no better than 40%. The notions to keep in mind are that there is a much bigger number of banks in any important market, and a 99% level of confidence is all that is required when reporting to regulators about their derivatives exposure.
A contrary view of what is written in the preceding paragraphs would be that financial markets do not work as ifthey were engineering systems, and therefore the rules underpinning reliability engineering do not apply. If such a statement were made, it will be very weak – and this is for two reasons:
● Rocket scientists working in the banking industry, have been extensively using the Weibull distribution in risk management.10
The tool they employ in financial studies is precisely the same Weibull distri- bution which has been used, with considerable success, for reliability engineer- ing studies since the early 1950s.11
● Colossal amounts of money in bilateral derivatives contracts, traded over the counter (OTC), interconnect major financial institutions among them- selves in a way closely resembling an engineering system with black boxes arranged in tandem.
Starting with the premise that proper study on this interdependence, and its impact on systemic risk, is not yet done (though it should be done without delay), one can in no way refute the argument about the use of reliability engi- neering in the study of systemic risk. Moreover, in science, we are much more
confident when we reject a hypothesis than when we accept it:
● When we reject a hypothesis, we do so because there is evidence it should be rejected (which is not the case of using the Weibull distribution in the financial industry).
● When we accept a hypothesis, what we practically say is that there is no evidence for rejecting it – until a ‘nasty’ new fact shows up to destroy a whole theory, not just a hypothesis.
This discussion fits nicely within the IFRS (and Basel II) perspective because, as explained in the Introduction, the key objective of the new accounting rules is to provide a realistic basis for valuing transactions and positions inventoried in the entity’s portfolio. In fact, professional associations are actively searching for an accounting base that will permit interactive frames of reference for risk manage- ment to be represented in crisp terms like the two interlinked cases shown in Figure 2.2.
The study of the long leg of a credit risk distribution, as the one shown in Figure 2.3, has a good deal to do with stress testing and risk management. In both cases, the background is provided by accounting data and statistics – and both must be reliable.
As we saw in Chapter 1, accounting, auditing, and risk control correlate. It is not just banks which must look most carefully into their capital adequacy. All enti- ties should be adequately capitalized, beyond the minimum regulators require, and they should provide capital adequacy information to:
● Their shareholders
● Supervisory authorities, and
● Financial market at large.12
Information on an entity’s capital adequacy as well as capital allocation requires a robust, unbiased, and forward-looking accounting system. This also helps in terms of market discipline – which is an issue underpinning arguments between stan- dards setters and banking supervisors. Many banks think they have a valid risk management function in place and a better perspective than the standards body about ‘what is needed’. This is, of course, the old argument about self-regulation, which has been shattered with:
● The 1929–32 Great Depression, and
● The September 1998 bankruptcy of LTCM, the Rolls-Royce of the hedge funds, which nearly tore the world’s financial fabric to pieces.
As a conclusion to this section, it is wise to borrow a page from the July 2004 Exposure Draft of IFRS. It states that the objective is to require entities to provide disclosures in their financial statements that enable users to evaluate the signif- icance of financial instruments for the firm’s financial position and performance;
nature and extent of risks arising from financial instruments to which the entity was exposed during the period and at reporting date; and adequacy of capital.
This statement kills two birds with one well-aimed stone:
● It complements, at accounting side, requirements already outlined by Basel II, and
BY INSTRUMENT
BY COUNTERPARTY
BY MARKET
MARKING TO MARKET
MARKING TO MODEL
DISCOUNTED CASH FLOW
Figure 2.2 Interactive frames of reference for total risk management which require forward- looking accounting standards
● It provides the company’s management with a brief description of what is required for an effective tool for first-class governance.
Figure 2.4 presents in a nutshell the sense of the second bullet point. Reliable financial reporting is at the core of many subsystems with vital management accounting information (see Chapter 8). As an example, Figure 2.4 shows a dozen of them which, taken together, provide corporate management with lead- ing edge technology – if, and only if, accounting information on which their interactive reports are based is absolutely reliable (see also in section 7 the dis- cussion on the U-curve).