Although it is possible to model the vast majority of financial contracts with a relatively small set of standardized CTs, there will always be outliers. This is not only true for the methodology described above, but holds for any analytical methodology. The very notion of standards implies the existence of outliers; in other words, where there are standards, there are also exceptions.
In this appendix we discuss the state of standardization in the financial sector and propose technical solutions to the problem.
Standardization in the Financial Sector
It is common to define standards in public-sector areas. Air, water, electricity, transport, and food are examples of important public sectors, which are more or less regulated. After all, these are public goods we all depend on. If I drink a glass of water I want to be sure that certain standards have been met.
Standards, if they are to be adhered to, require a language through which they can be communicated. In the food industry, for example, the language of chemistry, together with labelling standards for packaging, ensure that the declaration and the contents match.
Although the financial sector can be considered one of the most important public sectors and while it is much less difficult to describe food than the conditions of financial contracts, there is nothing comparable to a food declaration in the financial sector today. The imperative of food declaration—declaration and content must match—does not apply in finance. Worse still, the financial sector simply lacks a comparable language to even describe the content of its products. The declared intention of the OFR is to change this and to do so would require a concept of CTs.
The financial sector is, in terms of standardization, where the industrial sector was in the middle of the 19th century, when each screw had its own nut. Despite this desperate state, there is a tendency in the financial sector to undo or prevent any standardization effort. Much of this criticism seems to be driven from a protective mode. Whatever the reason behind this tendency,
we should remember that: the majority of the contracts do follow standards, though undeclared ones. Standards are a hidden fact. By obstinately focusing on the notorious difficulty to tame outliers of finance, we forego solving the vast majority of cases, where this is easily possible. By not being able to fix the one percent, we give up on the 99 % which would be easily fixable. The Commodities and Futures Trading Commission (CFTC) and Securities and Exchange Commission (SEC) prove that standardization is possible even in the “notorious corners” of finance. They do so by forcing Over-the-Counter (OTC) derivatives through an exchange. OTC derivatives such as Collateralized Debt Obligations (CDOs) and complex options have hitherto been considered the most buoyant and innovative part of finance and the most difficult to control. If it is possible to bridle this part, there is hardly an argument left against standardizing the rest. Even if it was impossible to find a solution for the outliers, a system capable of reining in the vast majority of contracts would still be very attractive for an OFR, if an approximation for the rest can be found. Assuming that we can model 99 % of the financial contracts to a high degree of precision with standard contracts and the remaining 1 % with an approximation of 70 % accuracy, we still would get highly precise overall results. It is even possible to standardize nonstandard products from an analytical perspective. This we discuss next. 15 Standardization of the Outliers
Although seemingly paradoxical, it is possible to standardize even outliers. From Fig.
, we can deduce that if it is possible to generate financial events, it is possible to derive the analysis elements. Generation of financial events in an unrestricted world, however, is difficult or impossible when we have to work with cash flows directly. By doing so, we skip the event line which contains the full information regarding liquidity and value/income/sensitivity. We can salvage the system by producing directly the expected cash flows in two forms: one for the liquidity side and one for value/income/sensitivity (Brammertz et al. 3.2 ). 2009
The system is like a mini-version of the whole system. Figure
is almost the same picture as the upper part of Fig.
. There are however four notable differences.
Standardizing nonstandard contracts
Nonstandard contract instead of financial contracts: This implies using a general purpose programming language to define the cash-flow rules as necessary.
Dashed arrows: The dashed arrows between the nonstandard contract and the other three input elements stand for the functional relationship between these elements. The output of the contract necessarily depends on the other input elements, including, in particular, standard CTs. If we think about an exotic contract any relationship is feasible such as dependence on the difference of the three-month interest rate and the mean temperature in Florida, or the difference between the market value of a certain bond and a stock. Ratings, prepayments (as part of behaviour), and many other factors can play roles, which are represented by the input factors. A nonstandard contract may depend on other standard (or even nonstandard) contracts, for example, in a case where the payoff is a function of the value difference between a stock and a bond. The standard contracts depend also on the same input elements, however with clearly defined relationships. In order not to overload the picture, the arrows indicating this dependence are left out.
Output is pure cash flow: While standard contracts produce about two dozen different event types, there is only one event type known to the nonstandard contract type: cash flow. In order to cover the full analytical spectrum it is necessary to write two pieces of code: one generating cash flows for liquidity related analysis and one for value/income/sensitivity related analysis.
The cash-flow generating code has to be delivered along with the product.
What is standard in this system is the structure of the process, which links risk factors, financial contracts, and cash flows.
Since the code has to be written for each product, which is an error-prone process, additional care must be taken from a regulatory perspective. Possible precautions include additional checks and capital charges.
Finally, nonstandard products can change into standard products with appropriate efforts, thus reducing the additional capital charges.
The Boundary Between Standard and Nonstandard CTS
The set of possible contracts is not conclusive or enumerative but must expand over time if new relevant instruments appear on the market. Where standardization ends and non-standardization sets in, is a question of choice. The OFR could set itself, for example, a target to model at least 99 % of all relevant contracts with a 99 % precision, where relevance could be measured by a value, sensitivity, or risk measure and precision by correctness of amount and timing of the expected cash flows. This would demand a study of all existing market instruments. We estimate that 99 % of the current existing variety of contracts can be covered with a mere two and a half dozen mentioned CTs on a 99 % precision level.
we discuss the additional effort nonstandard contracts take. We conclude that with a system such as the one proposed, the likelihood of appearance of new nonstandard contracts is reduced and while they still will appear, they will quickly mutate into standard products. 3.8.5