Published October 25, 2012 | October 2012 issue
Pre-Crisis Supervisory Use of Market Data
Because of the attractive features of market data we discussed, supervisors made use of market data to supplement other inputs prior to the crisis.25 Pre-crisis reviews of supervisory use of market data also describe an upward trend in the use of this external data. By the late 1990s and early 2000s, supervisory use of market data was relatively common, although “use” often reflected a fairly informal monitoring role not fully integrated into other supervisory processes.26 Later reviews in the mid-2000s found increasing resources devoted to supervisory use of market data—albeit from a modest level. More generally, such reviews found that supervisory use of market data was “roughly consistent with the researching findings.”27 In practice, this meant that large-bank supervisors were monitoring market prices and quantities, often on a high-frequency basis.
The use of market data, even though it has limitations, has been promoted by Federal Reserve policymakers. The then chair and vice chair of the Board of Governors spoke directly to this end as use of market data grew: Vice Chair Roger Ferguson captured the spirit, noting: “Our examiners are extremely good at what they do, but any good examiner recognizes that data should come from a variety of different sources, including the signals that come from the market.”28
Post-crisis improvements to Federal Reserve supervision reinforced the use of market data as one of several inputs for supervisory assessment. Governor Daniel Tarullo noted, for example, “The Federal Reserve is also putting in place a permanent quantitative surveillance mechanism for the large, complex financial organizations we supervise. This mechanism will incorporate supervisory information, firm-specific data analysis, and market-based indicators to identify developing strains and imbalances that may affect multiple institutions, as well as emerging risks to specific firms.”29
Tarullo provided additional details on the use of market data in the supervision of systemically important firms in the new regime:
There are other ways to incorporate non-governmental views into the regulatory system. We have already taken steps in this direction in conjunction with the Federal Reserve’s overhaul of its approach to supervising the largest financial holding companies. As part of this effort—and with the aim of advancing both our microprudential and macroprudential goals—we have created a quantitative surveillance mechanism (QSM) to regularize the collection and analysis of relevant data. Among other things, the QSM will use market-based indicators such as stock prices, option prices, credit default swap spreads, and short-term funding costs to provide an external perspective on the condition of these institutions—one that will be formally presented to regular meetings of senior supervisory and other Federal Reserve staff. Market-based indicators of macroeconomic and financial market risks that could pose threats to the largest institutions also will be used to assess their condition.30
As part of the enhanced formality and process noted by Tarullo, the Federal Reserve requires appropriate discussion and review of the current supervisory posture to firms with outlier market signals. Supervisors will also review, as relevant, implications of the market signals for firm operations and performance in areas such as funding. This enhanced process does not necessarily make use of new market data or even increase use of market data by some supervisors. Rather, it encourages consistent use of market data in the context of a well-functioning, cross-firm supervisory review process.
25 The Federal Reserve Bank of Minneapolis has advocated for use of market data in the supervisory process and tracked that use for some time. See, for example, Feldman and Rolnick (1998), Feldman and Levonian (2001) and Stern (2001).
26 See Schmidt (2004) and the papers at clevelandfed.org/research/conferences/2004/march/index.cfm for discussion of supervisory use of market data.
27 See Furlong and Williams (2006).
28 See the interviews with Ferguson (2000) and Greenspan (2001).
29 See Tarullo (2009).
30 See Tarullo (2010).