|
What types of data would the NIF collect? |
E-mail |
|
The NIF will collect four main types of data, all of which serve as inputs to models of systemic risk: - Instrument reference data — the legal and contractual structures that define financial instruments;
- Entity reference data — the participants in the issuance and transactions process and their relationships (e.g., parent-subsidiary hierarchies);
- Transactions and positions data — detail on the transactions and holdings of financial institutions;
- Market data — market prices, trading volumes, and other key measures (e.g., volatilities) that emerge from the trading process.
In addition, the NIF would utilize macroeconomic and other data (e.g., GDP, unemployment rates, tax rates, house price indexes, etc.) already available from other sources; there is no need to duplicate these other collection efforts. |
|
Last Updated on Tuesday, 20 October 2009 08:48 |
|
|
Will the NIF need comprehensive coverage across all industry participants? Why? |
E-mail |
|
Comprehensive collection across industry participants is required. Otherwise, problematic contracts and exposures will naturally migrate to avoid supervisory scrutiny. The build-up of exposures by AIG Financial Products Division provides a cautionary example of this phenomenon of "regulatory arbitrage." |
|
Last Updated on Tuesday, 06 October 2009 11:35 |
|
Will the NIF require counterparty information? Why? |
E-mail |
|
Counterparty information is required. Only by mapping the network of relationships among participants will analysts be able to understand and simulate the behavior of the system as a whole. It is these inter-relationships that make the financial system too complex to be represented by a simple sum of its individual institutions. This is the core of the macroprudential problem. |
|
Last Updated on Tuesday, 06 October 2009 11:35 |
|
Will the NIF require daily reporting? Why? |
E-mail |
|
The NIF will require daily reporting of transaction data. Getting this type of granular data is central to the NIF being able to do its job (as explained elsewhere in these FAQs). By requiring daily reporting, the NIF will be able to minimize the reporting burden on the reporting entities. Once data reporting is standardized, the least costly way to report transaction data is with the equivalent of an electronic “cc” automatically attached to each trade confirmation message.
|
|
Last Updated on Thursday, 19 November 2009 18:49 |
|
How will the NIF handle transactions involving non-U.S. institutions? |
E-mail |
Data will be required from U.S.-based financial institutions and their foreign affiliates. In addition, foreign institutions would be required to report on transactions conducted in the U.S.
|
|
Last Updated on Tuesday, 06 October 2009 12:13 |
|
Read more...
|
|
Why will the NIF need to collect so much data, of so many different kinds? |
E-mail |
|
The next crisis undoubtedly will surprise us as the previous ones have done. Policy makers will require answers to urgent questions that we cannot foresee today. Whatever answers are supplied will be acted upon in a crisis; getting such actions wrong can make any crisis much worse. Given that building a data collection system can take years, it will be too late to initiate such an effort when the need arises. The next crisis will have to be addressed with the data held in the NIF at the time it occurs. This makes it essential that the NIF hold at least a complete, up-to-date, high-quality, standardized, ready-to-process set of foundational data (entity data, instrument data, transaction data) and the necessary analytical tools and systems to evaluate their implications. |
|
Last Updated on Friday, 20 November 2009 07:36 |
|
Won’t the NIF simply be swimming in data when crucial answers are needed? |
E-mail |
|
By providing common reference databases and a common reporting language, the NIF will be able to integrate the data it collects in an automated fashion. Because the NIF will require entities to send electronic cc’s for all of their transactions, on a regular basis, the NIF will always have a current view of the financial system. Because these data will be properly organized and tied together in a consistent manner and because the NIF will have dedicated high performance computing resources to process the data, the NIF will be able rapidly to provide decision-makers with needed summaries and aggregations – especially during a time of crisis. |
|
Last Updated on Friday, 27 November 2009 13:28 |
|
What is meant by “granular” data? |
E-mail |
|
Data on contractual terms and conditions -- such as the interest rate, principal, and maturity of a loan -- are key ingredients in risk calculations. We recommend that such data be captured at the most granular level: as attributes of specific legal contracts. The alternative is to capture the information at some higher level of aggregation, such as the average interest rate and maturity across all loans in a portfolio. However, aggregated data are necessarily and unavoidably less informative than fully granular data. Detailed data will provide the flexibility required to feed a diverse range of risk models. The resulting insights will be invaluable in identifying future systemic fragilities, whose source cannot be predicted today. Because granular, contract-level data will typically constitute business confidential information, data security will be a core feature of the NIF. |
|
Last Updated on Friday, 23 October 2009 08:39 |
|
How can granular or contract-level data tell us anything about system-level risk? |
E-mail |
|
The solvency of a financial entity depends, in part, on assets that are guaranteed by other financial entities (counterparties). Understanding how financial entities are connected contractually is critical to understanding how failures can cascade through the financial network. Aggregation and netting are based on assumptions that vary from market participant to market participant, assumptions that do not always hold in practice and that are especially vulnerable to failure during a crisis. |
|
Last Updated on Tuesday, 10 November 2009 13:12 |
|
How will granular or contract-level data help regulators understand an emerging crisis? |
E-mail |
|
The NIF must be able to understand the capacity of the financial system to manage and distribute risks; to monitor how risk flows through the financial network, and to identify potential points of failure. This understanding can only be built up based on a detailed view of market liquidity (how frequently and deeply instruments trade), and a clear analysis of the obligations and rights of liquidity providers. For example, financial entities that are under stress are forced to sell assets, which can lead to a fire sale and a lack of liquidity. As liquidity disappears in one market they are forced to sell other assets, which can quickly lead to a lack of liquidity in other markets. The failure of multiple markets can cause a widespread lack of confidence, resulting in runs on the entire market. |
|
Last Updated on Tuesday, 10 November 2009 13:14 |
|
How will granular or contract-level data help regulators deal with the growing complexity of financial contracts? |
E-mail |
|
Some of the risks that are hardest to measure are those associated with highly complex and innovative financial instruments whose payoffs depend on other financial instruments, assets, or external factors, often in intricate ways. Without detailed, granular data, it can be difficult or impossible to understand the risks involved with such complex instruments. For example, shocks to the value of underlying collateral at the bottom of a complex structured product (e.g., mortgage-backed or asset-backed structured products) can result in “toxic assets” that cannot be priced. In order to understand the risk to solvency of financial entities that hold these complex instruments, and the risk in turn to the counterparties of these entities, the NIF must be able to see how all of these complex financial instruments connect to their underlying collateral. This capability depends on access to granular, contract-level data.
|
|
Last Updated on Thursday, 19 November 2009 20:10 |
|
Why would the NIF need granular data to monitor systemic risks? Why wouldn’t aggregated or netted data be sufficient? |
E-mail |
|
Data aggregation could potentially create serious blind spots. The process of aggregation results in the irretrievable loss of information that could be critical during a crisis. It is not possible to know in advance the most appropriate way to aggregate data for the purposes of systemic risk analysis. The NIF needs to collect and maintain granular data exactly in order to determine the most appropriate ways to aggregate data for the purposes of monitoring systemic risk and responding to crises whose sources cannot be predicted in advance. |
|
Last Updated on Tuesday, 09 March 2010 10:38 |
|
Read more...
|
|
Are there situations where the NIF would collect or use aggregated data? When and why? |
E-mail |
|
Wherever practicable, the NIF would collect data at the most granular level. Aggregation is a one-way street: granular data can be aggregated, but the original detail ordinarily cannot be recovered from aggregated data. Collection of granular data is recognized as a best practice, and it governs data collection strategies at most financial, commercial and telecommunication companies today. It is facilitated by advances in data storage and data access software. Once collected, however, the same data might be used in aggregated form for any number of purposes, such as calculating regional summary statistics, or peer group benchmarks for industry subgroups. “Tick” (or trade-level) data is a good example: data on each trade enables analysis at the most granular level (trade detail), as well as at aggregated levels (end-of-day, quarter by quarter, year over year, etc.). |
|
Last Updated on Tuesday, 24 November 2009 12:44 |
|
Could costs be reduced by collecting smaller volumes of aggregated data, rather than larger volumes of granular data? |
E-mail |
|
To the contrary, it is actually easier, both for regulators and financial firms, if data are collected in the same forms in which they are generated originally. Once data reporting processes are standardized, the least costly and least burdensome way to report transaction data is with the equivalent of an electronic “cc” automatically attached to each trade confirmation message. A requirement that firms aggregate data prior to transmission would actually increase the burdens on firms. |
|
Last Updated on Thursday, 26 November 2009 20:01 |
|
|
