FR2900.20060425.omb

Supporting Statement for the Reports of Deposits

(FR 2900; OMB No. 7100-0087)

(FR 2910a; OMB No. 7100-0175)

(FR 2915; OMB No. 7100-0237)

(FR 2930; OMB No. 7100-0088)

Summary

The Board of Governors of the Federal Reserve, under delegated authority from the Office of Management and Budget (OMB), proposes to extend for three years, with revision, the reports of deposits. This group of reports consists the:

• Report of Transaction Accounts, Other Deposits and Vault Cash (FR 2900; OMB No. 7100-0087)

• Annual Report of Total Deposits and Reservable Liabilities (FR 2910a; OMB No. 7100‑0175)

• Report of Foreign (Non-U.S.) Currency Deposits (FR 2915; OMB No. 7100-0237)

• Allocation of Low Reserve Tranche and Reservable Liabilities Exemption
  (FR 2930; OMB No. 7100‑0088)

Depository institutions submit deposit data either weekly, quarterly, or annually. Larger depository institutions generally must submit deposit data more frequently than smaller ones. These mandatory data are used by the Federal Reserve for administering Regulation D (Reserve Requirements of Depository Institutions) and for constructing, analyzing, and monitoring the monetary and reserve aggregates.

The Federal Reserve proposes several changes to the structure of deposit reporting to be implemented in 2006 and 2007. In light of the decision to eliminate the M3 monetary aggregate, these recommended changes include using a new measure—total transaction accounts, savings deposits, and small time deposits—for determining the frequency at which depository institutions file deposit reports, rather than total deposits, which includes large time deposits, a component of M3. Implementation of this measure would be delayed until 2007 to allow time for depository institutions to adjust to the changes.

The proposed changes for 2006 are designed to reduce reporting burden. Under the current indexation formula, the Federal Reserve announced, in October 2005, the values of the nonexempt deposit cutoff and the reduced reporting limit—two parameters used to determine the frequency at which a depository institution files a report of deposits. The Federal Reserve has traditionally used the triennial report renewal process as an opportunity to provide additional burden reduction for depository institutions, while remaining mindful of possible harmful effects on the accuracy of the monetary aggregates, by raising one or more of these parameters above their indexed values. The staff is proposing to increase the nonexempt deposit cutoff again in 2006.

This proposal focuses on a nonexempt deposit cutoff of $229.1 million in total deposits, which is equivalent to $200 million based on the new measure. Such a cutoff would provide substantial burden reduction by shifting 615 FR 2900 reporters from the weekly to the quarterly panel while keeping likely benchmark revisions to M2 within acceptable bounds. The Federal Reserve is not proposing any revision to the reduced reporting limit. It would remain at $1.206 billion in total deposits or $1.120 billion based on the new measure.

The recommended revisions summarized below are grouped to show those that would go into effect this year and those that would be delayed until next year.

Revisions Proposed for Implementation in 2006

• Raise the nonexempt deposit cutoff to $229.1 million (compared with an indexed level of $181.1 million) and set the reduced reporting limit at its indexed value of $1.206 billion beginning in September 2006. These parameters are expressed in terms of total deposits and would determine the frequency at which depository institutions submit deposit data for the period from September 2006 to September 2007.

• Combine the FR 2930 and the FR 2930a into one reporting form (the FR 2930) that would be used by all respondents, effective September 30, 2006.

• Revise the instructions to the FR 2930 to reflect the merging of two reports.

• Revise the instructions for the FR 2900 to enhance clarity.

Revisions Proposed for Implementation in 2007

• Calculate the nonexempt deposit cutoff and the reduced reporting limit using the sum of total transaction accounts, savings deposits, and small time deposits, rather than total deposits, beginning with the September 2007 panel shift.¹

• Index the nonexempt deposit cutoff and the reduced reporting limit annually to 80 percent of the June-to-June growth in total transaction accounts, savings deposits, and small time deposits at all depository institutions.² The actual values of the nonexempt deposit cutoff and the reduced reporting limit to be used in September 2007 will be announced under the usual schedule, in October 2006.

• Revise the FR 2910a reporting form (which is also used in determining reporting frequency), as follows and revise the instructions accordingly, effective for the June 30, 2007, report date:

• Replace data item 1, “Total Deposits,” with “Total Transaction Accounts, Savings Deposits, and Small Time Deposits;”

• Delete the parenthetical text from data item 1, “(If the amount reported for this item is less than or equal to $7.0 million, Items 2 and 2.a need not be completed);” and

• Change the reporting form title from, “Annual Report of Total Deposits and Reservable Liabilities,” to “Annual Report of Deposits and Reservable Liabilities; and

• Require depository institutions to submit either a positive or negative value in data item 2.a, “Net Transaction Accounts,” rather than reporting negative values as zero, as is currently required.³

The proposed revisions to the family of deposit reports would decrease the total estimated annual reporting burden from 692,543 to 591,737 hours, a reduction of 100,806 hours (15 percent), representing a savings to the public of $2 million annually. Copies of the FR 2930 and FR 2910a reporting forms, marked to show the revisions, are attached. Although data on large time deposits are no longer needed for constructing M3, no revisions to the FR 2900 or FR 2915 reporting forms are recommended since data on large time deposits are still needed to construct an aggregate balance sheet for commercial banks, and the staff believes it is less burdensome to banks to continue reporting large time deposits on the FR 2900 than to add the data to another information collection.

Background and Current Reporting Structure

The current system of reporting is designed to meet the requirements of the Federal Reserve Act as amended by both the Monetary Control Act of 1980 (MCA) and the Garn‑St Germain Depository Institutions Act of 1982 (Garn‑St Germain Act). The MCA imposes reserve requirements on all depository institutions that have transaction accounts or nonpersonal time deposits.⁴ In implementing MCA, the Federal Reserve elected to limit the reporting burden on smaller institutions by reducing their frequency of reporting. As a result, institutions with total deposits below a deposit cutoff submit FR 2900 data at a quarterly rather than weekly frequency. The Garn‑St Germain Act imposes a zero‑percent reserve requirement on a specific amount of a depository institution’s reservable liabilities (the “exemption amount”), in effect exempting from reserve requirements all depository institutions whose total reservable liabilities are less than or equal to the exemption amount. The Garn-St Germain Act also requires that depository institutions with a zero-percent reserve requirement be subject to less overall reporting requirements than other depository institutions consistent with the Federal Reserve’s responsibility to monitor and control the monetary and credit aggregates.

The current reporting framework for the main deposit reports was implemented in April 1983 to fulfill the reduced reporting requirements of the Garn‑St Germain Act. This framework originally comprised five reporting categories, but in 2000, the number of reporting categories was reduced to four (defined in detail in the next section, entitled “Deposit Reports”).⁵ Since September 2003, the boundaries of the four reporting categories have been defined by three measures: the exemption amount, the nonexempt deposit cutoff, and the reduced reporting limit.⁶ Membership in the four reporting categories is reviewed annually, and the assignment of institutions to reporting panels (known as panel shifts) occurs each September. These shifts have reflected movements in depository institutions’ total deposits and net transaction accounts across the prevailing boundaries that separate the reporting categories.

The exemption amount, which is governed by statute, and the nonexempt deposit cutoff and the reduced reporting limit, which are both determined by the Federal Reserve, are each indexed annually. The exemption amount, which was initially set at $2 million in 1983, is indexed annually by 80 percent of the annual growth rate of reservable liabilities at all depository institutions.⁷ The deposit cutoff and reduced reporting limit are each indexed annually by 80 percent of the annual growth rate of total deposits at all depository institutions. The initial deposit cutoff was set at $5 million in 1985, while the initial reduced reporting limit was set at $1 billion in 2003. Appendix C provides a history of the settings for each of these three measures since their initial implementation. Without the proposed Federal Reserve intervention, these indexing procedures would call for the nonexempt deposit cutoff to be set to $181.1 million and the reduced reporting limit to be set at $1.206 billion (in terms of total deposits), for the September 2006 panel assignments.

Deposit Reports (FR 2900 and FR 2910a)

With the exceptions noted below, an institution’s reporting status is currently determined by the levels of its net transaction accounts and total deposits.

Nonexempt institutions—defined as those with net transaction accounts greater than the exemption amount or with total deposits equal to or greater than the reduced reporting limit—file the fifteen‑item FR 2900 weekly if their total deposits are equal to or greater than the nonexempt deposit cutoff and quarterly if their total deposits are less than the nonexempt deposit cutoff. U.S. branches and agencies of foreign banks and banking Edge and agreement corporations are required to submit FR 2900 data weekly regardless of their deposit size.

Exempt institutions—defined as those with net transaction accounts less than or equal to the exemption amount and with total deposits less than the reduced reporting limit—file the three‑item FR 2910a covering one day in June if their total deposits are greater than the exemption amount. Exempt institutions with total deposits less than or equal to the exemption

amount are not required to submit deposit data to the Federal Reserve if other data sources, such as Call Reports, are available.⁸ Any institution that is assigned to a particular reporting category may, using established procedures, elect instead to submit deposit data (and, if applicable, maintain reserves) in accordance with a higher frequency reporting category.

Tables 1.a and 1.b show the number of institutions and the amount of total deposits, respectively, by data source—FR 2900 weekly, FR 2900 quarterly, FR 2910a, or Call Report—and by entity type, as of June 2005.

Table 1.a

Number of Institutions by Entity Type and Data Source

(as of June 2005)

Entity type	Data source				All DIs
	FR 2900 weekly	FR 2900 quarterly	FR 2910a	Call Report
Commercial Banks	2,298	3,965	1,151	176	7,590
S&Ls, Savings Banks	595	361	334	34	1,324
Credit Unions	642	706	4,120	3,569	9,037
Corporate Central Credit Unions	30	1	0	0	31
Banking Edge and Agreement Corporations	18	Not eligible for quarterly or annual reporting			18
U.S. Branches & Agencies of Foreign Banks	229				229
All DIs	3,812	5,033	5,605	3,779	18,229

Table 1.b

Volume of Total Deposits by Entity Type and Data Source

(as of June 2005, billions of dollars)

Entity type	Data source				Total deposits
	FR 2900 weekly	FR 2900 quarterly	FR 2910a	Call Report
Commercial Banks	4,798	313	50	1	5,162
S&Ls, Savings Banks	985	35	19	2	1,041
Credit Unions	374	58	140	9	581
Corporate Central Credit Unions	83	0	0	0	83
Banking Edge and Agreement Corporations	5	Not eligible for quarterly or annual reporting			5
U.S. Branches & Agencies of Foreign Banks	709				709
All DIs	6,954	406	209	12	7,581

Foreign Currency Deposit Report (FR 2915)

The FR 2915 collects seven‑day averages of the amounts outstanding for foreign (non‑U.S.) currency-denominated deposits held at U.S. offices of depository institutions, converted to U.S. dollars and included in the institution’s FR 2900 data. Foreign currency deposits are subject to reserve requirements and, therefore, are included in the FR 2900 data submission. However, because foreign currency deposits are not included in the monetary aggregates, the FR 2915 data are used to remove foreign currency deposits from aggregated FR 2900 data in constructing the monetary aggregates. All weekly and quarterly FR 2900 respondents offering foreign currency deposits file the six-item FR 2915 quarterly, on the same reporting schedule as quarterly
FR 2900 respondents.

Allocation Reports (FR 2930 and FR 2930a)

Institutions with two or more offices (or groups of offices) that file separate FR 2900 reports at either a weekly or quarterly reporting frequency are required to file at least annually the FR 2930 or FR 2930a. An institution’s net transaction accounts up to the exemption amount ($7.8 million in 2006) are reserved at 0 percent.⁹ In addition, net transaction accounts up to the low reserve tranche ($48.3 million in 2006) are reserved at 3 percent, while amounts in excess of this amount are reserved at 10 percent. Only a single exemption amount and a single low reserve tranche are allowed per depository institution (including subsidiaries). Therefore, an institution that submits separate FR 2900 data covering different groups of offices is required to file the FR 2930 or FR 2930a at least annually to allocate its single exemption amount and low reserve tranche among its offices.

Description of Information Collection

Report of Transaction Accounts, Other Deposits and Vault Cash (FR 2900)

The FR 2900 is the primary source of data used for the calculation of required reserves and applied vault cash and for the construction and interpretation of the monetary aggregates. Data are also used for indexing the exemption amount and low reserve tranche amount each year, as required by statute, and for indexing the nonexempt deposit cutoff and reduced reporting limit each year as determined by the Federal Reserve. The Federal Reserve proposes that the nonexempt deposit cutoff be raised to $229.1 million in total deposits, a significant increase from its 2006 indexed value of $181.1 million in total deposits. The Federal Reserve proposes that the reduced reporting limit remain at its 2006 indexed value of $1.206 billion in total deposits. Federal Reserve proposes that, owing to the elimination of the M3 monetary aggregate, the nonexempt deposit cutoff and reduced reporting limit be recalibrated in terms of total transaction accounts, savings deposits, and small time deposits. The Federal Reserve also proposes that the nonexempt deposit cutoff and reduced reporting limit be indexed annually to 80 percent of the June-to-June growth in the new measure. Each of these proposed changes is discussed below. No changes to data coverage or reporting frequency are recommended.

Data Coverage. The Federal Reserve does not recommend any change to the FR 2900 reporting form. The FR 2900 reporting form currently consists of fifteen items. Twelve items are reported at a daily frequency for the Tuesday‑through‑Monday reporting period. These twelve items are necessary for the calculation of reserve requirements, applied vault cash, and for the construction of the monetary aggregates. Three items—those that supply data on the nontransaction components of total reservable liabilities—are submitted at an annual frequency for use in the annual indexation of the exemption amount and low reserve tranche amount. See Appendix B for further detail on the current uses of each of the fifteen items.

Although the Federal Reserve has discontinued the publication of M3, The Federal Reserve is not recommending that large-denomination time deposits be removed from the FR 2900. Data on large time deposits are needed for constructing the aggregate balance sheet of commercial banks published each week in the Federal Reserve’s H.8 statistical release, Assets and Liabilities of Commercial Banks in the United States. Given that daily data on small-denomination time deposits (total time deposits minus large time deposits) are necessary for constructing the M2 monetary aggregate, the Federal Reserve believes that it would be least burdensome on depository institutions to continue collecting total and large time deposit data on the FR 2900, rather than collecting small time deposit data directly on the FR 2900 and large time deposit data separately on the bank credit family of reports.¹⁰

Reporting Frequency. The Federal Reserve does not recommend any change to the FR 2900 reporting frequency. Weekly reporting of the current twelve daily FR 2900 items by larger nonexempt institutions facilitates calculation of reserve requirements, construction of the monetary aggregates, and the implementation of monetary policy. Quarterly reporting of those same FR 2900 items by smaller nonexempt institutions reduces the reporting burden on these institutions.¹¹ All FR 2900 reporters, regardless of how frequently they submit the daily data items, submit the three annual FR 2900 data items one day each year.

U.S. branches and agencies of foreign banks and banking Edge and agreement corporations submit the FR 2900 data on a weekly basis, regardless of their size. The relationship between these institutions and their parent organizations makes possible short‑term transfers of liabilities and assets between reporting dates in order to avoid reserve requirements. To eliminate the possibility of reserve avoidance, continued submission of weekly data for these institutions is deemed necessary.

For each of the twelve daily items, the Federal Reserve also recommends the continued collection of seven days of data each report week, rather than data for a single day or weekly average data. Single‑day data would increase the ability of institutions—in particular the larger ones—to avoid reserve requirements by managing their balance sheets to reduce net transaction accounts on the report date. In addition, single‑day data—for example, levels submitted only for Monday—are more volatile than the weekly average of daily data currently used to construct the money stock.

For several reasons, submitting data for each of the seven days separately is preferable to submitting seven‑day averages only. First, the ability to make day‑to‑day (as well as week‑to‑week) comparisons through editing procedures is important in maintaining the quality of the weekly average data. This ability is particularly important with the advent of retail sweeps, which result in very large fluctuations in the daily deposit data. In addition, respondents would still need to compile the daily data to compute the weekly average. Finally, daily data, not weekly averages, are used in constructing final month‑average levels of the monetary aggregates. Monthly figures could be constructed from prorating weekly-average figures, but such figures would only provide approximations. For those components of the monetary aggregates that have pronounced intra‑weekly movements, such as vault cash and demand deposits, monthly averages of daily figures normally differ substantially from monthly figures constructed by prorating weekly averages.

In the past, daily data have been useful for measurement and analysis of the monetary aggregates. The availability of daily data has facilitated the interpretation of sharp movements in the money stock when the timing of these movements within a week coincided with major disturbances, such as quarter‑end balance sheet adjustments by commercial banks, the terrorist attacks that occurred on September 11, 2001, the shutdown of a financial center due to a power failure, and weather-related disruptions to depository institutions. In addition, the availability of daily data has made the detection of new retail sweep programs much easier than would have been the case based only on weekly-average data.

Reporting Panels. The nonexempt deposit cutoff and the reduced reporting limit are used to determine the frequency at which depository institutions submit deposit data. These values have been indexed each year to the growth in total deposits. Based on that indexation procedure and absent further Federal Reserve action, the nonexempt deposit cutoff and reduced reporting limit would be $181.1 million and $1.206 billion, respectively, as approved and announced by the Federal Reserve in October 2005.

The Federal Reserve proposes that the nonexempt deposit cutoff be raised to $229.1 million in total deposits and that the reduced reporting limit be set at $1.206 billion in total deposits, effective in September 2006. The proposed increase in the nonexempt deposit cutoff reduces reporting burden on depository institutions, while keeping adverse consequences for the measurement of money and reserves to what the staff believes are an acceptable level.

Owing to the discontinuance of M3, the Federal Reserve proposes that, beginning in September 2007, the nonexempt deposit cutoff and the reduced reporting limit be expressed in terms of total transaction accounts, savings deposits, and small time deposits, rather than total deposits. This change is proposed so that the frequency at which an institution submits FR 2900 deposit data is tied to the institution’s issuance of deposits included in the M2 monetary aggregate, rather than its issuance of all types of deposits, some of which were included only in M3. Likewise, the Federal Reserve proposes that, going forward, the nonexempt deposit cutoff and the reduced reporting limit be indexed annually to grow at 80 percent of the June 30- to-

June 30 growth rate of total transaction accounts, savings deposits, and small time deposits at all depository institutions.¹² The results of this new indexation procedure will be announced in October 2006 in keeping with normal practice.

Under the proposal to raise the nonexempt deposit cutoff to $229.1 million in total deposits, the staff estimates that 615 nonexempt institutions (18 percent of the weekly panel) would become eligible to shift to quarterly FR 2900 reporting (Table 2).¹³

Table 2

Effect of Proposed Nonexempt Deposit Cutoff on FR 2900 Panel

Panel	Cutoff ($ millions)	Number of FR 2900 weekly reporters	Number of FR 2900 quarterly reporters
Indexed Value of Nonexempt Deposit Cutoff	181.11	3,367	5,478
Proposed Nonexempt Deposit Cutoff	229.11	2,752	6,093
Difference		(615)	615

1. A nonexempt deposit cutoff of $181.1 million based on total deposits yields the same number of weekly FR 2900 reporters as a cutoff of $153.4 million based on the new measure: Total transaction accounts, savings deposits, and small time deposits. Similarly, $229.1 million in total deposits yields the same number of weekly FR 2900 reporters as $200.0 million does using the new measure.

Loss of weekly reporters to the quarterly reporting category has several adverse effects on the measurement of the monetary and reserve aggregates. Each of these effects is discussed below.

(1) Loss of accuracy in the initial weekly estimates of the monetary aggregates. As institutions move from weekly to quarterly reporting, initial weekly estimates of the monetary aggregates become less accurate because fewer institutions are submitting weekly data. The Federal Reserve must estimate the weekly deposits of all non-weekly reporters, which it does by using actual deposit growth for a sample of the smaller weekly reporting institutions (known as the EDDS surrogate panel). In addition, the higher nonexempt cutoff removes some of the smaller weekly-reporting institutions that would otherwise be available for the surrogate panel, so that the remaining weekly reporting institutions may become less representative of the non-weekly reporters. These weekly and quarterly projection errors show up when new quarterly data are received, at which time the monetary aggregates are revised.

Appendix A analyzes the effects of alternative nonexempt deposit cutoff levels on quarterly revisions to M2 and on reporting burden. The tables and the discussion in the appendix show how the errors involved in estimating the quarter‑end levels of M2 deposits at quarterly reporting institutions rise with higher levels of the nonexempt deposit cutoff as more institutions shift from weekly to quarterly reporting status. The increase in the nonexempt deposit cutoff from $153.4 million in total transaction accounts, savings deposits, and small time deposits (the equivalent of its 2006 indexed value, $181.1 million, in total deposits) to $200.0 million (the equivalent of $229.1 million in total deposits) would increase the size of the expected error in estimating quarter‑end M2 deposits at quarterly reporters. (See column 3 of Table A.1 of Appendix A.) This increase is deemed acceptable given the burden reduction that would result.

(2) Permanent loss of information between quarters. As institutions move from weekly to quarterly reporting status, the staff loses information about the actual daily, weekly, and monthly movements of deposits at these institutions between quarterly report dates. The estimated historical pattern is likely to differ more from the true weekly deposit holdings when less actual weekly data are collected, but there is no way to directly measure or correct for such error. Thus, the loss of higher frequency information that accompanies a higher nonexempt cutoff is permanent, and cannot be corrected by quarterly benchmarking.

(3) Potential for reserve avoidance. For quarterly FR 2900 reporters, vault cash held during the reporting week is applied against required reserves during the subsequent quarter. Thus, these depository institutions could acquire extra vault cash during the reporting week and reduce the balances—called required reserve balances—that they are required to maintain to satisfy reserve requirements throughout the following quarter. The number of institutions that could benefit from the temporary acquisition of vault cash—namely, quarterly FR 2900 reporters whose normal vault cash holdings are less than their reserve requirements (bound institutions)—increases when the level of the nonexempt cutoff is raised.

Although raising the nonexempt deposit cutoff would make it easier for some institutions to reduce their required reserve balances, this effect is dwarfed by the main method used to reduce required reserve balances—namely, retail sweep programs. By implementing such programs, large weekly reporters have substantially lowered their required reserve balances, and yet there has been no adverse effect on the Federal Reserve’s ability to implement monetary policy. Likewise, it is unlikely that allowing smaller institutions to shift to quarterly reporting status would reduce required reserve balances enough to interfere with the implementation of monetary policy.

Table 3 shows required reserve balances of potential quarterly reporters at selected nonexempt deposit cutoff levels, as of the maintenance period ending June 22, 2005.

(4) Loss of high frequency information to track issues facing the banking and thrift industries. At times staff uses the weekly deposit data to evaluate the impact of natural or manmade disasters on specific regions or even specific institutions and to assess industry trends by specific entity types or geographic regions. These efforts are most successful when the weekly reporting panel is sufficiently broad to provide good coverage when data are disaggregated by entity type or region. Experience has shown that large weekly reporters are able to provide sufficient data on current trends in the financial services industry and on product innovation, as well as sufficient information during temporary dislocations in the financial markets. Indeed, reflecting ongoing consolidation in the financial services industry, financial activity is mainly concentrated at large institutions. In addition, consolidation has meant that there are now weekly reporters in all geographic markets. As a result, there is less need to have many individual reporters to ensure adequate geographic coverage.

Table 3

Effect of Proposed Nonexempt Deposit Cutoff on Required Reserve Balances

Nonexempt deposit cutoff 1	Total Required Reserve Balances ($9,271 million)
	Percent at quarterly reporters	Percent at weekly reporters	Quarterly reporters ($millions)	Weekly reporters ($millions)
$153.4 (current)2	6	94	522	8,749
$200 2	9	91	851	8,420
$250	14	86	1,296	7,975
$300	17	83	1,621	7,650
$350	20	80	1,872	7,399
$400	23	77	2,096	7,175
$450	25	75	2,274	6,997
$500	26	74	2,400	6,871

1. Based on the new measure: Total transaction accounts, savings deposits, and small time deposits.

2. Recall that equivalent total deposit amounts in millions of dollars for 153.4 and 200.0 are 181.1 and 229.1, respectively.

In summary, the proposed nonexempt deposit cutoff of $200.0 million in terms of the new measure would allow the Federal Reserve to decrease the reporting burden of weekly nonexempt reporters without significantly jeopardizing the accuracy of the monetary and reserve aggregates.

Annual Report of Total Deposits and Reservable Liabilities (FR 2910a)

Currently, the three‑item FR 2910a is generally filed by exempt institutions whose total deposits (as shown on their December Call Report) are greater than the exemption amount but less than the reduced reporting limit. Respondents submit single‑day data as of June 30. The Federal Reserve proposes one change in FR 2910a data coverage.

Data Coverage. The FR 2910a currently consists of three items: total deposits, reservable liabilities, and net transaction accounts. The first data item submitted on the FR 2910a, total deposits, comprises the sum of total transaction accounts, savings deposits, total time deposits, and other reservable liabilities. Data on total deposits submitted on the FR 2910a are currently used to determine whether an institution will continue to be eligible for reduced reporting and, if not, the frequency at which the institution must submit FR 2900 data (weekly or quarterly). In addition, these data are used in the annual indexation of the nonexempt deposit cutoff and the reduced reporting limit.

The second item, reservable liabilities, consists of the sum of net transaction accounts, nonpersonal savings deposits, and nonpersonal time deposits, regardless of maturity. Data on reservable liabilities are used for the annual indexation of the exemption amount (as required by statute).

Net transaction accounts, the third item, consist of total transaction accounts plus ineligible acceptances and obligations issued by affiliates maturing in less than seven days; less demand balances due from depository institutions and cash items in process of collection. Data on net transaction accounts are used in the annual indexation of the low reserve tranche (as required by statute) and to determine whether an institution will continue to be eligible for reduced reporting.

Reporting Panels and Frequency. Any depository institution that does not file an FR 2900 and that has total deposits¹⁴ greater than the exemption amount must submit an FR 2910a each year on June 30.¹⁵

Proposed Revisions to the FR 2910a

In light of the elimination of M3 and the staff’s proposal to use total transaction accounts, savings deposits, and small time deposits, rather than total deposits, for FR 2900 panel assignments in 2007, the Federal Reserve proposes to replace data item 1 “Total Deposits”—which includes large time deposits—with “Total Transaction Accounts, Savings Deposits, and Small Time Deposits.” In addition, the Federal Reserve proposes to change the reporting form title from “Annual Report of Total Deposits and Reservable Liabilities” to “Annual Report of Deposits and Reservable Liabilities.” These revisions would be effective June 30, 2007. ¹⁶ (See attached FR 2910a reporting form marked to reflect these proposed changes.)

The Federal Reserve also proposes to revise the FR 2910a instructions to require that depository institutions submit negative data values of net transaction accounts (data item 2.a). Currently, respondents are instructed to enter a zero if their calculation of net transaction accounts yields a negative value. Under the proposed revision, any such negative value would be submitted on the reporting form. Each year the Federal Reserve indexes the low reserve tranche to 80 percent of the June 30-to-June 30 growth in net transaction accounts. These calculations are performed at an aggregate level, by first summing net transaction accounts across all depository institutions, by year, and then calculating 80 percent of the year-to-year growth. By requiring FR 2910a reporting institutions to submit negative amounts of net transaction accounts rather than replacing negative amounts with zero, measurement of aggregate net transaction accounts would be improved. This revision would be effective June 30, 2007. In order to avoid artificially depressing the growth of net transaction accounts for indexation calculations, the negative values would not be used in indexation until the fall of 2008.

Report of Foreign (Non-U.S.) Currency Deposits (FR 2915)

All FR 2900 respondents, both weekly and quarterly, that offer deposits denominated in foreign currencies at their U.S. offices file the FR 2915. The FR 2915 was implemented in January 1990 following the decision by the Federal Reserve in late 1988 not to object to issuance of foreign currency deposits at depository institutions in the United States after December 31, 1989. At that time a procedure for converting the value of such deposits into dollars for submission purposes was established. Data collected on the FR 2915 are mainly used in the construction of the monetary aggregates. (These data are included in deposit data submitted on the FR 2900 for reserve requirement purposes, but they are not included in the monetary aggregates.) The FR 2915 is the only source of data on such deposits. As of June 27, 2005, 107 depository institutions submitted about $15 billion of total foreign currency deposits on the FR 2915. No changes to data coverage or reporting frequency are proposed.

Data Coverage. The Federal Reserve does not recommend any change to the FR 2915. The amounts of foreign currency deposits held at U.S. offices of a depository institution are converted to U.S. dollars and included in the appropriate item on the institution’s FR 2900 data, which collects outstanding balances as of the close of business each day of the seven-day reporting week. The six items on the FR 2915 break out the amounts of these foreign currency-denominated deposits that are included in selected FR 2900 data items. Specific FR 2900 data items are referenced on the face of the FR 2915 reporting form.

The staff has looked for alternative sources of quarterly data on foreign currency deposits. However, the item on foreign currency deposits was removed from the Call Report in March 1996; therefore, the FR 2915 data are now the sole source of data on foreign currency deposits from depository institutions. The staff has also examined whether the number of items collected on the reporting form could generally be reduced. In order to construct the monetary aggregates, foreign currency deposits must be broken out by major deposit type and must eliminate interbank transactions.

Reporting Frequency. The Federal Reserve does not recommend any change to the FR 2915 reporting frequency. The FR 2915 collects seven‑day averages at a quarterly frequency. All FR 2915 respondents file data according to the FR 2900 quarterly reporting schedule; data are filed for the seven-day report week beginning on the third Tuesday of March, June, September, and December. The Federal Reserve determined in 1994 that quarterly measures would suffice both for backing these data out of the monetary aggregates as well as for monitoring the overall volume of the deposits, and a quarterly collection frequency is still sufficient.

Although foreign currency data on large time deposits will no longer be needed to construct M3 after March 2006, the Federal Reserve proposes that the FR 2915 continue to collect these data in order to maintain the parallel reporting system that currently exists between the FR 2900 and the FR 2915. In addition, foreign currency data on small denomination time deposits are necessary for the construction of M2, and these data are obtained from the items currently reported on the FR 2915. The continued reporting of total time and large time deposits on the FR 2915, instead of the direct reporting of small time deposits on the FR 2915, would not measurably affect respondent reporting burden.

Allocation of Low Reserve Tranche and Reservable Liabilities Exemption (FR 2930/2930a)

The FR 2930 and FR 2930a collect data on the allocation of the low reserve tranche and reservable liabilities exemption amount for depository institutions having offices (or groups of offices) that file separate FR 2900 deposit reports.¹⁷ The FR 2930 is filed by U.S. branches and agencies of foreign banks and banking Edge and agreement corporations; the FR 2930a is filed by other types of depository institutions. Both reporting forms collect the same data. However, the instructions and explanatory information differ.

Proposed Revisions to the FR 2930 and FR 2930a

Staff proposes the two reporting forms be combined into one reporting form (FR 2930) that would be used by any entity type (both foreign-related and domestic institutions). The instructions for the FR 2930 reporting form would be modified to reflect this change. The effective date of this revision would be September 30, 2006. (See attached FR 2930 reporting form marked to reflect this proposed change.)

Staff recommends no revisions to the data content of the proposed reporting form. As noted earlier, an institution’s net transaction accounts up to the exemption amount are reserved at 0 percent. In addition, net transaction accounts up to the low reserve tranche are reserved at a lower ratio than amounts in excess of this level. Only a single exemption and a single low reserve tranche, however, are allowed for all U.S. offices of the same parent institution. Therefore, in order to calculate the reserve requirement of an institution that submits separate FR 2900 data for two or more offices, that institution is required to allocate, using the FR 2930, the low reserve tranche and the exemption among those offices. Data must be submitted to each Reserve District in which a reporting office is located.

The revised FR 2930 data would continue to be collected from all respondents toward the end of each calendar year (or early in the next calendar year, as appropriate) to take into account the annual indexing of the low reserve tranche and the exemption amount, as well as any reallocation of these amounts that institutions may make among their various offices. The data would still be required at least one week before the beginning of the reserve computation period (a 14‑day period beginning on Tuesday and ending on Monday of the second week following) in which the revised allocations are to be effective. Relating to the 2005 indexation of the tranche and exemption amount, 160 depository institutions filed the FR 2930. An allocation reporting form is also required at any time during the year when an institution changes the number of FR 2900 reporting forms it submits.¹⁸

Reporting Instructions

The instructions for the FR 2910a and FR 2930 would be revised to reflect the revisions to the reporting forms, as discussed above. In addition, the instructions for the FR 2900 would be updated and clarified. These clarifications would include removing obsolete text and making the text parallel the Call Report instructions, where appropriate.

Justification for Proposed Revisions

The Federal Reserve discontinued compiling the M3 monetary aggregate in March 2006. The construction and publication of M3—which consists of M2, large-denomination time deposits, repurchase agreements, Eurodollars, and institutional money market mutual funds—consumed resources beyond that required to construct and publish M2. Depository institutions prepared and submitted additional data used only for M3. The Federal Reserve had to validate, construct, and publish these data. These activities involved costs in the form of staff time and data processing resources. M3, however, had not been closely tracked by policymakers for some time, nor was it routinely analyzed by Federal Reserve staff. Staff work suggested that M3 did not convey any additional information about economic activity that was not already embedded in M2. Although academic papers had occasionally used M3 as a variable in empirical work, the staff’s literature review did not surface any studies that concluded that M3 was an important financial indicator.

The nonexempt deposit cutoff and the reduced reporting limit are used to determine the frequency with which depository institutions must submit deposit data. These measures are currently defined in terms of total deposits, and they are currently indexed each year to growth in aggregate total deposits. With the discontinuance of M3, the Federal Reserve proposes that these two measures be recalibrated in terms of total transaction accounts, savings deposits, and small time deposits, to be effective with the annual deposit panel assignments in September 2007. This recalibration would improve the coverage of M2 deposits for a given number of weekly FR 2900 panel members and it would not change aggregate reporting burden for the family of deposit reports. Going forward, the Federal Reserve proposes that these panel-determination parameters be indexed annually to the growth in total transaction accounts, savings deposits, and small time deposits at all depository institutions instead of total deposits. The Federal Reserve’s proposal to raise the nonexempt deposit cutoff to $229.1 million, in terms of total deposits, would reduce reporting burden without significantly impairing the measurement of M2 and reserves aggregates.

In order to implement the changes discussed in the previous paragraph, the Federal Reserve proposes, effective June 30, 2007, to revise the FR 2910a by replacing data item 1, “Total Deposits,” with “Total Transaction Accounts, Savings Deposits, and Small Time Deposits”. To improve the measurement of net transaction accounts each June 30 for use in annual indexation of the low reserve tranche and exemption amount, the Federal Reserve also proposes removing certain text from the FR 2910a instructions that raised negative values of net transaction accounts to zero. This change would not affect reporting burden. It would be implemented in June 2007 but not used in indexation calculations until the fall of 2008 in order to avoid artificially depressing growth in net transaction accounts. Finally, the combination the FR 2930 and FR 2930a into one reporting form to be used by all entity types is proposed to reduce the number of forms. This change would require some revision to the reporting instructions but would not affect the reporting burden.

Table 4 summarizes the proposed deposit reporting requirements that would become effective September 2006 and provides an estimate of the number of institutions that would be assigned to each reporting category.

Table 4

Proposed Deposit Reporting Categories

(effective September 2006)

Exempt institutions (Eligible for reduced reporting) Net transaction accounts < $7.8 million AND Total deposits < $1.206 billion		Nonexempt institutions (Not eligible for reduced reporting) Net transaction accounts > $7.8 million OR Total deposits  $1.206 billion
3,779 Nonreporters	5,605 Annual reporters	6,093 Quarterly reporters	2,752 Weekly reporters
 Have total deposits < $7.8 million	 Have total deposits  $7.8 million	 Have total deposits < $229.1 million	 Have total deposits > $229.1 million
 Do not file an FR 2900 family report	 File the 3-item FR 2910a report each June 30	 File the FR 2900 report: 12 daily items one week each quarter, and 3 annual items on Monday of the June report week	 File the FR 2900 report: 12 daily items each week, and 3 annual items each June 30

Time Schedule for Information Collection and Publication

Reserve Banks collect the deposit data from respondents on a schedule that allows the Banks to meet the deadline for reporting those data to the Federal Reserve Board. The Reserve Banks review the data for accuracy and then make the data available to the Federal Reserve Board.

FR 2900 quarterly respondents submit daily data for the week beginning with the third Tuesday and continuing through the following Monday in March, June, September, and December. FR 2910a respondents submit data annually, as of close of business June 30 each year. FR 2915 respondents submit data for the week beginning on the third Tuesday and ending the following Monday in each of March, June, September, and December. FR 2930 respondents submit data annually, as of December 31, or upon the establishment of an office outside the home state or Federal Reserve District.

Publication. The data are used in the production of public statistical releases and internal reports. Aggregate data for deposits, reserves, or both are published in numerous publicly available statistical releases: Aggregate Reserves of Depository Institutions and Monetary Base (H.3 statistical release; Statistical Supplement Table 1.20); Money Stock Measures
(H.6 statistical release; Statistical Supplement Table 1.21); Reserves and Money Stock Measures (Statistical Supplement Table 1.10); Reserves and Borrowings (Statistical Supplement
Table 1.12); and, Assets and Liabilities of Commercial Banks in the United States (H.8 statistical release; Statistical Supplement Table 1.26).

Legal Status

The Federal Reserve Board’s Legal Division has determined that the deposit reports are required by law, as follows:

FR 2900 and FR 2930: 12 U.S.C. 248(a), 461, 603, and 615

FR 2910a and FR 2930a: 12 U.S.C. 248(a) and 461

FR 2915: 12 U.S.C. 248(a)(2), and 347(d)

The data are given confidential treatment under Section b(4) of the Freedom of Information Act [5 U.S.C. 552(b)(4)].

Consultation Outside the Agency

There has been no consultation outside the Federal Reserve System.

Sensitive Questions

These reports contain no sensitive questions as defined by OMB guidelines.

Estimated Cost to the Federal Reserve System

The proposed cost to the Federal Reserve System for collecting and processing the FR 2910a and FR 2930 is estimated to be $173,867 per year, an increase of 9.0 percent from the current cost of $159,367. The one-time cost to implement the revised reports is estimated to be $37,784.

Respondent Burden and Costs

As shown in Table 5, the current annual reporting burden for the reports of deposits is estimated to be 692,543 hours. The proposed changes represent a net burden reduction of 100,806 hours, or 15 percent for this family of reports. The proposed total burden represents 11.9 percent of all Federal Reserve System reports, down from 13.9 percent.

The FR 2900 panel shift would represent a net burden reduction of 103,320 hours or 15 percent. The decline in burden is due to some FR 2900 reporting institutions becoming eligible to submit data less frequently; that is, the proposed increase in the nonexempt deposit cutoff above its 2006 indexed value would result in a shift of 615 FR 2900 respondents from the weekly to the quarterly reporting panel.

One of the proposed revisions replaces FR 2910a, data item 1, “Total Deposits” with “Total Transaction Accounts, Savings Deposits, and Small Time Deposits” effective in June 2007. This change could make it more difficult for credit unions, savings and loan institutions, and federally chartered savings banks to submit data on the FR 2910a because they are not required to submit small and large time deposits data as separate data items on their respective Call Reports. As a result, the estimated time per response for the 4,454 credit unions, savings and loan institutions, and federally chartered savings banks that submit FR 2910a data would increase from 30 minutes to 1 hour. In addition, the proposed revision to data item 1 would increase the estimated time per response for the 1,151 commercial banks, state chartered cooperative banks, and state chartered savings banks that submit FR 2910a data, from 30 minutes to 45 minutes. This increase in burden is due to the additional time it would take depository institutions to calculate the proposed data item 1. This proposed revision would increase the estimated FR 2910a annual burden, starting in 2007, from 2,803 to 5,317 hours, an increase of 2,514 hours, or less than 0.5 percent of the overall respondent burden for the deposit reports.

Based on an hourly cost of $20, the estimated annual cost to the public would decline from $13,850,860 to $11,834,740 a year, a reduction of $2,016,120.

Table 5

Estimated Respondent Burden for Deposit Reports ¹

	Number of respondents	Annual frequency	Estimated average hours per response	Estimated annual burden hours
Current
FR 2900 weekly	3,367	52	3.50	612,794
FR 2900 quarterly	5,478	4	3.50	76,692
FR 2910a	5,605	1	0.50	2,803
FR 2915	107	4	0.50	214
FR 2930/2930a	160	1	0.25	40
Total				692,543
Proposed
FR 2900 weekly	2,752	52	3.50	500,864
FR 2900 quarterly	6,093	4	3.50	85,302
FR 2910a 2 (Commercial banks and state savings banks only)	1,151	1	0.75	863
FR 2910a 2 (Federal savings banks, S&Ls, and credit unions only)	4,454	1	1.00	4,454
FR 2915	107	4	0.50	214
FR 2930	160	1	0.25	40
Total				591,737
Change				(100,806)

1. The number of respondents is calculated as of June 27, 2005. The split between FR 2900 weekly and quarterly reporters is relative to the estimated panel sizes that would prevail if the 2006 indexed value of the nonexempt deposit cutoff ($181.1 million in total deposits or $153.4 million in total transaction accounts, savings deposits, and small time deposits) were implemented. The FR 2900 weekly data exclude the Federal Home Loan Banks.

2. Proposed estimates for FR 2910a reporters would be effective in June 2007.

APPENDIX A¹⁹

The Cutoff between Quarterly and Weekly Reporters

Nonexempt depository institutions submit FR 2900 data either quarterly or weekly depending on their level of total deposits. One result of the 1985 deposit reports clearance process was that the cutoff between quarterly and weekly reporters was raised from $15 million to $25 million and was then indexed annually to grow at 80 percent of the growth rate of total deposits at all depository institutions.²⁰ The cutoff for exempt institutions to submit data annually on the FR 2910a or quarterly on the FR 2910q was set to equal the FR 2900 quarterly versus weekly cutoff. Under the next regular three‑year review of the deposit reports in 1988, the single cutoff was further increased from its indexed value of $30 million to $40 million. In each of the next five years it was indexed, reaching $44.8 million in 1993.

In the 1994 reports review, the single deposit cutoff structure was replaced with a dual cutoff scheme—one deposit cutoff was applied to nonexempt institutions to determine whether they reported the FR 2900 on a weekly or quarterly basis (the nonexempt deposit cutoff), and a second deposit cutoff was applied to exempt institutions (the exempt deposit cutoff) to determine whether they reported quarterly (on the FR 2910q) or annually (on the FR 2910a). When first implemented during the 1994 panel shifts, the nonexempt deposit cutoff was set to $55.0 million and the exempt deposit cutoff was set to $44.8 million, each to be indexed annually thereafter according to the usual procedure. In 1997, the cutoff for nonexempt reporters was increased to $75.0 million from its indexed value of $59.3 million, while the cutoff for exempt reporters was left at its indexed value of $48.2 million. In 2000, the cutoff for nonexempt reporters was increased to $95 million from its indexed value of $84.5 million, and the exempt cutoff was eliminated entirely along with the associated FR 2910q report.

In 2003, the cutoff was increased to $150 million from its indexed value of $112.3 million, and a reduced reporting limit (RRL) was established so that institutions whose net transaction accounts were less than or equal to the exemption amount would be required to submit the
FR 2900 weekly if their total deposits were greater than or equal to the RRL. The RRL was initially set at $1 billion (causing 12 exempt institutions to be added to the FR 2900 weekly panel) and was to be indexed in the same way as the deposit cutoff. The indexing procedure calls for the nonexempt cutoff to rise to $181.1 million and the indexed RRL to rise to $1.206 billion, effective September 2006.

This appendix considers the implications of raising the nonexempt deposit cutoff above its indexed value of $181.1 million in total deposits (or the equivalent value of $153.4 million in total transaction accounts, savings deposits, and small time deposits). The benefit of increasing the nonexempt cutoff, of course, is that reporting burden would be reduced. The principal cost,

which will be examined here, is the adverse effect such a move would have on the measurement of the monetary aggregates. In particular, we examine how much, on average, quarterly revisions to the aggregates would increase if the nonexempt cutoff were raised. In addition, we examine how quarterly revisions would have varied, given alternative values for the reduced reporting limit.

With the discontinuance of M3, the Federal Reserve has proposed that, in 2007, instead of total deposits the cutoff should be based on a concept more tailored to the monetary and reserve aggregates of interest. The proposed measure for determining the frequency of reporting is total transaction accounts, savings deposits, and small time deposits and is the sum of total transaction accounts (regardless of owner), savings deposits, small time deposits, and ineligible acceptances and obligations issued by affiliates maturing in less than 7 days.

If the indexed cutoff based on total deposits, $181.1 million, were applied to June 2005 data, 3,367 depository institutions would be required to submit weekly data. The same number of institutions (3,367) would be required to submit weekly data if the cutoff were $153.4 million of total transaction accounts, savings deposits, and small time deposits. Therefore, we consider $153.4 million to be the equivalent value in terms of total transaction accounts, savings deposits, and small time deposits, and this serves as the base against which we compare alternative cutoffs.

Effect of the Cutoff on the Measurement of the Aggregates

FR 2900 quarterly reporters submit their deposit data to the Federal Reserve for just one week each quarter. For these institutions, also known as QEDS reporters, deposit levels during the intervening weeks are estimated using a surrogate panel drawn from FR 2900 weekly reporters (called EDDS reporters). When the next quarter’s QEDS data become available, the QEDS time series are revised to agree with the reported single-week data, and appropriate revisions are made to the values for the weeks falling between the quarterly report weeks.

Deposits at depository institutions that do not file either EDDS or QEDS reports are estimated in much the same way as are the deposits of QEDS institutions. Quarterly single‑day deposit data for virtually all depository institutions not on the EDDS or QEDS panels are available from their Call Reports, and the surrogate panel is used to estimate deposits of these institutions between the quarterly Call Report dates. In this appendix, for the purpose of summarizing revisions, QEDS and Call Report (only) quarterly reporters are grouped together and labeled “quarterly reporters.”²¹

These end‑of‑quarter revisions for the quarterly reporters are the summary statistics to be used in evaluating the nonexempt cutoff’s effect on the measurement of the monetary aggregates. Time‑series of estimates of deposits and revisions for quarterly reporters were calculated under simulated, hypothetical conditions—that is, for several potential nonexempt cutoff values, what would revisions have been if the institutions that currently submit weekly data but are now below the hypothetical nonexempt cutoff had been assigned to quarterly reporting and reported the same data that they had reported on EDDS for those quarterly (QEDS) reporting weeks, and the surrogate panel had to be selected from those institutions now falling above the hypothetical nonexempt cutoff. For the purposes of this appendix, the surrogate panel included the smallest few hundred institutions still reporting weekly and that are not known to be retail sweepers. The study period consists of the twenty quarters from March 2000 through March 2005.

The results of the simulations are summarized in Table A.1. Column 2 shows how the amount of deposits at quarterly reporters increases with increases in the nonexempt cutoff. Looking down column 2, as the amount of deposits being estimated increases, it is natural to expect that errors in estimating that amount will increase as well, as has generally, but not uniformly, turned out to be the case. Error statistics are shown in the remaining three columns. The mean absolute quarterly revisions are shown in Column 3, and root mean square (RMS) or standard deviations of the revisions are given in Column 4. For example, at a nonexempt cutoff of $153.4 million, the sample RMS revision for M2 deposits (and therefore the estimate of the long‑run standard deviation) is $1,809 million. (In Table A.1, it is assumed that the RRL is left unchanged. Effects of increasing the RRL will be addressed below.)

The last column in Table A.1 (Column 5) translates the RMS revision of M2 deposits into the implied RMS revision of quarter-to-quarter growth rates of M2. (To construct this column, a revision to a quarterly average was taken to be approximately half the revision for the quarterly reporting week, and the revisions were compared to the level of M2 as of March 2005—$6,472.3 billion.) As noted, the RMS revisions are estimates of the long-run standard deviations of the revisions. As a rule of thumb, two-thirds of revisions are expected to be less than the RMS revision and 95 percent are expected to be less than twice the RMS revision.

As expected, the simulated errors generally increase with increases in the cutoff level. The increases become much sharper as the cutoff reaches around $300 million. Note, however, that the hypothetical QEDS data for the DIs currently reporting weekly are the values reported on EDDS for the QEDS reporting weeks. Many DIs with total transaction accounts, savings deposits, and small time deposits exceeding $300 million offer retail sweep accounts, and their deposits are difficult to estimate on the basis of deposits at other banks. If these DIs were to discontinue their sweep activity after moving from EDDS to QEDS, the revision statistics in Table A.1 would tend to overstate the quarterly revisions to be expected in the future. More likely, the DIs moving to QEDS would simply change their sweep algorithms to minimize reservable deposits during the QEDS reporting week, rather than minimizing the average over the month. In that case, the statistics in Table A.1 may understate the revisions to be expected in

the future. (Including sweepers in the surrogate panel probably wouldn’t help much, since their intra-monthly time series patterns would be different from those of the QEDS sweepers.)

Finally, note that the statistics shown in Table A.1 just reflect experience over the past five years, and the RMSs, taken as estimates of the long-run standard deviations, are themselves subject to error. This caveat suggests that one should be conservative in selecting the maximum acceptable cutoff.

Burden Reduction

The second criterion for evaluating the cutoff is the savings to the Federal Reserve System and to the banking system that would result from the reduced processing associated with an increased nonexempt cutoff. Panel sizes as functions of the nonexempt cutoff are shown in Table A.2. These panels are based on 2005 second-quarter data. The first row shows the sizes of the existing FR 2900 weekly and quarterly panels. The second row shows the panel sizes as they would be if the 2006 indexed total deposits cutoff of $181.1 million were in effect today. The third row shows the results for the equivalent value in terms of the new measure—total transaction accounts, savings deposits, and small time deposits. As stated earlier, the equivalent value produces the same number of weekly reporters as would the indexed total deposits cutoff (but not exactly the same DIs; 87 DIs qualify for weekly reporting using the total deposits criterion but not using the new measure, and 87 DIs qualify using the new measure but not using the indexed total deposits cutoff). These indexed panel sizes are the base from which one can measure burden reduction from further increases in the cutoff. With a nonexempt cutoff of, say, $200.0 million based on the new measure, an estimated 615 institutions would leave the weekly FR 2900 panel, a reduction of 18 percent. The quarterly FR 2900 panel would increase accordingly.

The Reduced Reporting Limit and Further Burden Reduction

As noted earlier, twelve DIs were added to the FR 2900 weekly panel in 2003 because their total deposits exceeded the RRL. Currently, 24 DIs whose net transaction accounts are less than or equal to the exemption amount have total deposits that are equal to or greater than the indexed RRL ($1.206 billion), and would be expected to submit weekly data. The RRL equivalent in terms of total transaction accounts, savings deposits, and small time deposits would be $1.120 billion, that is, 24 DIs would be required to submit weekly data using this criterion. Table A.3 shows the results of simulations assuming that the RRL is increased. To keep this table of manageable size, we just focus on the root mean square revisions. However, the caveats regarding the summary statistics in Table A.1 also apply to this table. Also, the incremental effect of increasing the RRL is shown for just two representative levels of the cutoff. The simulated RRLs range from the current level of $1.120 billion up to $32 billion, at which point no additions to the weekly FR 2900 would be made, as no DIs with net transaction accounts less than or equal to the exemption amount are that large. The simulations indicate that the RRL could be set anywhere between $1.120 billion and $2 billion (in terms of total transaction accounts, savings deposits, and small time deposits) without materially affecting the quarterly revisions. If the RRL were increased to, say, $2 billion, only 8 DIs, rather than the current 24, would be required to submit the FR 2900 weekly, and the weekly panel sizes shown in Table A.2 could be reduced by 16. (The FR 2900 quarterly panels are unaffected—RRL only determines

whether the DI submits the FR 2900 weekly; if not, its Call Report data are used as the data source for the monetary aggregates.)

Table A.1

Quarter‑End Revisions of M2 at Various Nonexempt Deposit Cutoffs

March 2000 – March 2005

(1)	(2)	(3)	(4)	(5)
FR 2900 cutoff levels ($mil) 1	M2 deposits at quarterly reporters (percent) 2	Sample mean absolute quarterly revision ($mil)	Sample root mean square of quarterly revisions ($mil)	Root mean square effect on annualized quarterly growth rate (percent)
153.4	9.4	1,490	1,809	0.06
200	11.4	1,771	2,128	0.07
250	13.5	2,227	2,706	0.08
300	15.1	3,544	4,821	0.15
350	16.8	4,731	6,180	0.19
400	18.2	7,158	8,799	0.27
450	19.4	8,063	10,115	0.31
500	20.3	11,778	14,293	0.44

1. In terms of total transaction accounts, savings deposits, and small time deposits.

2. As of March 2005. For the purposes of this table, M1 deposits = demand deposits adjusted + other checkable deposits, and M2 deposits = M1 deposits + savings deposits + small time deposits, all aggregated over EDDS weekly and QEDS and Call quarterly reporters.

Table A.2

Number of Respondents at Various Nonexempt Deposit Cutoffs

(as of June 27, 2005)

Alternative cutoffs ($mil)	FR 2900 weekly reporters 1	FR 2900 quarterly reporters
Existing panel 2	3,812	5,033
Indexed 3	3,367	5,478
Total transaction accounts, savings deposits, and small time deposits:
153.4	3,367	5,478
200	2,752	6,093
250	2,274	6,571
300	1,921	6,924
350	1,653	7,192
400	1,468	7,377
450	1,319	7,526
500	1,209	7,636

1. This column includes 18 banking Edge and agreement corporations, 229 U.S. branches and agencies of foreign banks, 12 bankers’ banks and 30 corporate credit unions. These institutions report the FR 2900 weekly, regardless of their size. Excluded are 12 Federal Home Loan Banks, which appear on the FR 2900 data file although they do not file an FR 2900 report (Federal Reserve Banks record some information on reserves-related items for these banks, so they are included in the weekly data flow to the Federal Reserve Board).

2. In 2005, before the September panel shift, the nonexempt deposit cutoff was $161.2 million.

3. In September 2006, the indexed cutoff is $181.1 million in terms of total deposits.

Table A.3

Quarter‑End Root Mean Square Revisions of M2

at Various Reduced Reporting Limits and Selected Nonexempt Deposit Cutoffs

March 2000 – March 2005

(1)	(2)	(3)	(4)
Reduced reporting limit ($mil) 1	Number of depository institutions affected	RMS revision at nonexempt cutoff of $153.4 million ($mil)	RMS revision at nonexempt cutoff of $400 million ($mil)
1,120	24	1,809	8,799
2,000	8	1,912	7,888
10,000	3	3,560	10,283
32,000	0	3,784	10,957

1. In terms of total transaction accounts, savings deposits, and small time deposits.

Appendix B

Uses of FR 2900 Items

Table B.1 summarizes the current uses of each item on the FR 2900. As shown in the table, the separate reporting of various deposit categories is needed because of the different treatment of particular items in the definitions of reservable liabilities or monetary aggregates. For example, all demand deposits are classified as transaction accounts for calculating required reserves, but two items—demand deposits due to depository institutions and U.S. government demand deposits—are not included in the monetary aggregates and, therefore, must be submitted separately. Similarly, time and savings deposits are treated the same way for purposes of reserve requirements, but separate data are needed for construction and publication of the monetary aggregates.

In 1991 the Federal Reserve reduced the number of items collected on the FR 2900 from twenty-one to fourteen by consolidating some items that were previously submitted separately. In addition, the definition of one item was expanded.²² In 2003, the reporting frequency of two items was reduced from daily to only one day a year, and a new annual item—net Eurocurrency liabilities—was added.²³

Table B.1

Current Uses of each FR 2900 Item

Item	Calculation of required reserves1	Construction of monetary aggregates2	Comment
A. Transaction accounts:
1. Demand Deposits: a. Due to depository institutions	 Reserved as transaction accounts (3%/10%) 3	n.a.	 Not included in the monetary aggregates.
 b. Of U.S. government	Reserved as transaction accounts (3%/10%) 3	n.a.	Not included in the monetary aggregates, but published as a memorandum item on the H.6.
 c. Other demand	Reserved as transaction accounts (3%/10%) 3	M1	Monetary aggregates. For banks, other demand (item A.l.c) is used to calculate the demand deposits adjusted component of M1, which is published on the H.6. For thrifts, this item is a component of other checkable deposits.
2. ATS accounts and NOW accounts/share drafts, and telephone and preauthorized transfers	Reserved as transaction accounts (3%/10%) 3	M1/M2	Monetary aggregates. ATS and NOW accounts (item A.2) are included in the other checkable deposits component of M1, while telephone and preauthorized transfer accounts are introduced at the M2 level. With all three types of accounts submitted as a single total on the FR 2900, the Federal Reserve estimate the amount of telephone and preauthorized transfer accounts to be subtracted from that total and included in M2.
B. Deductions from transaction accounts:
1. Demand balances due from depository institutions in the U.S. 2. Cash items in process of collection	Deducted from transaction accounts before application of reserve requirement ratio. Deducted from transaction accounts before application of reserve requirement ratio.	n.a. M1	Reserve calculations. The sum of demand balances due from depository institutions in the U.S. and cash items in process of collection (items B1 and B2) is deducted from gross transaction account (the sum of items A.1.a, A.1.b, A.1.c, A.2, and A.A.1) in order to produce “net transaction accounts,” which is subject to reserve requirements. Monetary aggregates. Cash items in process of collection (item B.2) are deducted from other demand deposits in calculating the demand deposits adjusted component of M1 for banks and the other checkable deposits component of M1 for thrifts.

Table B.1, Continued

C. Total savings deposits (including MMDAs)	Personal accounts are not reservable; nonpersonal savings deposits are reservable, but with a reserve ratio of zero. See Schedule BB.1 below.	M2	Monetary aggregates. Savings deposits are a major component of M2.
D. Total time deposits	Personal accounts are not reservable; nonpersonal or negotiable time deposits are reservable, but with a reserve ratio of zero. See Schedule BB.1 below.	M2/M3 4	Monetary aggregates. Small‑denomination time deposits (those in amounts less than $100,000) are introduced at the M2 level, while large‑denomination time deposits (those in amounts of $100,000 or more) are introduced at the M3 level. The small‑denomination portion is calculated by subtracting memorandum item F.1, large time deposits, from total time deposits (Item D).
E. Vault cash	Total required reserves less vault cash equals the amount of reserves to be maintained at the Federal Reserve Bank	M1	Monetary aggregates. Vault cash is deducted from currency in circulation to arrive at the currency component of M1. Reserve aggregates. Vault cash is used to meet reserve requirements. The amount used is applied vault cash.
F. Memorandum Item:
1. All time deposits with balances of $100,000 or more (included in Item D above)	n.a.	M2/M3 4	Monetary aggregates. See comments for item D above. Also used in the construction of the Federal Reserve’s weekly H.8 statistical release, Assets and Liabilities of Commercial Banks in the United States.

Table B.1, Continued

Schedule AA:
1. Ineligible acceptances and obligations issued by affiliates maturing in less than 7 days	Reserved as transaction accounts (3%/10%) 3	n.a.	Reserve calculations. A depository institution is required to maintain reserves against ineligible acceptances and certain obligations issued by a nondepository affiliate if the proceeds of such obligations are channeled to the depository institution. These obligations are not direct obligations of the depository institution but are reservable obligations under Regulation D.
Schedule BB: Nonpersonal items
1. Total nonpersonal time and savings deposits (included in Items C and D above)	Reserved as nonpersonal time deposits (0%)	n.a.	A component of total reservable liabilities, used in the annual indexation of the exemption amount. (Annual indexation of the exemption amount is required by statute.)
2. Ineligible acceptances and obligations issued by affiliates maturing in 7 days or more (nonpersonal only)	Reserved as nonpersonal time deposits (0%)	n.a.	A component of total reservable liabilities that is used in the annual indexation of the exemption amount. (Annual indexation of the exemption amount is required by statute.)
Schedule CC:
Net Eurocurrency liabilities	Reserved as non-transaction accounts (0%) 5	n.a.	A component of total reservable liabilities that is used in the annual indexation of the exemption amount. (Annual indexation of the exemption amount is required by statute.)

Notes to Table:

1. Detailed procedures for the calculation of required reserves are presented in the Reserve Maintenance Manual (http://www.frbservices.org/).

2. For additional information on the monetary aggregates, please see the H.6 statistical release or Tables 1.10 and 1.21 in the Statistical Supplement to the Federal Reserve Bulletin.

3. The reserve ratios shown are those in effect as of January 2005. The amount of total net transaction accounts equal to or below the low reserve tranche is reserved at 3 percent, while the amount in excess of the tranche is reserved at 10 percent. Total net transaction accounts are gross transaction accounts (the sum of Section A, Items 1 and 2, plus Schedule AA, Item 1) less deductions as submitted in Section B of the reporting form (Items B.1 and B.2). In addition to the ratios shown in the table, the first $7.0 million of an institution’s reservable liabilities are subject to a reserve requirement of 0 percent in 2005 (this amount is referred to as the exemption amount).

4. As of March 23, 2006, the Federal Reserve ceased construction and publication of M3. Given that daily data on small denomination time deposits (the difference between total time deposits and large time deposits) are necessary for use in constructing the M2 monetary aggregate, the Federal Reserve has determined that it is least burdensome for depository institutions to continue to collect total and large time deposit data on the FR 2900 form.

5. Prior to 1991, the FR 2950 and FR 2951 were also used in the calculation of required reserves. These reports were discontinued in May 2004

Appendix C

Historical Levels of the Exemption Amount, Deposit Cutoff(s), and

Reduced Reporting Limit

(Used to Define Reporting Category Boundaries for the Reports of Deposits)

(millions of dollars)

Calendar year	Exemption amount	Deposit cutoff level(s)		Reduced reporting limit
		Nonexempt	Exempt
1983	2.11	15.02	15.02	B
1984	2.2	15.0	15.0	B
1985	2.4	25.03	25.03	B
1986	2.6	26.8	26.8	B
1987	2.9	28.6	28.6	B
1988	3.2	40.04	40.04	B
1989	3.4	42.1	42.1	B
1990	3.4	43.4	43.4	B
1991	3.4	44.0	44.0	B
1992	3.6	44.8	44.8	B
1993	3.8	44.8	44.8	B
1994	4.0	55.05	44.85	B
1995	4.2	55.4	45.1	B
1996	4.3	57.0	46.4	B
1997	4.4	75.06	48.26	B
1998	4.7	78.9	50.7	B
1999	4.9	81.9	52.6	B
2000	5.0	95.07	B	B
2001	5.5	101.0	B	B
2002	5.7	106.9	B	B
2003	6.0	150.08	B	1,0008
2004	6.6	161.2	B	1,074
2005	7.0	169.8	B	1,131
2006: Indexed 9 2006: Proposed 10 2006: Proposed Equivalent Value 11	7.8	181.1 229.1 200.0	B B B _	1,206 1,206 1,120

Notes to table:

1 . The exemption amount was implemented for the reserve computation period beginning December 9, 1982. The

1983 adjustment, from $2.0 million to $2.1 million, was made effective with its initial implementation.

2. The deposit cutoff was initially set at $5 million. When the reduced reporting system was implemented in 1983,

it was raised to $15 million. This cutoff originally determined whether nonexempt institutions submitted the FR 2900 weekly or quarterly and whether exempt institutions submitted the FR 2910q or FR 2910a.

3. In conjunction with the triennial review of the deposit reports in 1985, the Federal Reserve raised the cutoff level

to $25.0 million, to be indexed each year thereafter at 80 percent of the June 30-to-June 30 growth rate of total

deposits at all depository institutions.

Appendix C, Continued

4. In conjunction with the triennial review of the deposit reports in July 1988, the Federal Reserve raised the cutoff

level to $40.0 million from its indexed value of $30.0 million.

5. In conjunction with the August 1994 triennial review of the deposit reports, the Federal Reserve elected to

reduce reporting burden by replacing the single deposit cutoff level with two separate cutoffs—one for

nonexempt institutions and the other for exempt institutions. The nonexempt deposit cutoff was raised from the

indexed amount of $44.8 million to $55.0 million, while the exempt deposit cutoff was retained at the indexed

level of $44.8 million. Thereafter, each of the two deposit cutoffs was indexed annually to total deposits.

6. In conjunction with the August 1997 triennial review of the deposit reports, the Federal Reserve lifted the

nonexempt deposit cutoff from the indexed amount of $59.3 million to $75.0 million, while the exempt deposit

cutoff was kept at its indexed level of $48.2 million.

7. In conjunction with the July 2000 triennial review of the deposit reports, the Federal Reserve discontinued the

FR 2910q report, thereby eliminating the exempt deposit cutoff. In addition, the Federal Reserve raised the

nonexempt deposit cutoff from the indexed amount of $84.5 million to $95.0 million.

8. In conjunction with the June 2003 triennial review of the deposit reports, the Federal Reserve raised the

nonexempt deposit cutoff from the indexed amount of $112.3 million to $150.0 million, and implemented the

reduced reporting limit with an initial value of $1.0 billion, to be indexed annually to total deposits.

9. These values were calculated according to the usual procedures and were approved by the Federal Reserve and

announced in October 2005.

10. Values recommended by the Federal Reserve in this proposal and expressed in terms of total deposits.

11. Equivalent value is expressed in terms of total transaction accounts, savings deposits, and small time deposits instead of total deposits.

1. “Total transaction accounts, savings deposits, and small time deposits” for panel-selection purposes is the sum of total transaction accounts, total savings deposits (including MMDAs), gross small time deposits, and ineligible acceptances and obligations issued by affiliates maturing in less than seven days. “Total deposits” for panel-selection purposes is defined as the sum of total transaction accounts, total savings deposits (including MMDAs), total time deposits, total ineligible acceptances and obligations issued by affiliates, and net Eurocurrency liabilities.

2. As with the indexation of the reserve requirement exemption amount, if total transaction accounts, savings deposits, and small time deposits decline in that period, the Federal Reserve would not adjust these parameters downward.

3. In order to avoid artificially depressing the growth in net transaction accounts, the change in reporting will be implemented in June 2007 but not used in indexation calculations for the low reserve tranche and exemption amount until the fall of 2008, when both the base period (June 30, 2007) and the end period (June 30, 2008) will be reported on the same basis.

4. The Federal Reserve imposes reserve requirements on U.S. branches and agencies of foreign banks under the authority of the International Banking Act of 1978.

5. The FR 2910q was discontinued on September 25, 2000, thereby eliminating one of the reporting categories. Improved timeliness and processing procedures made it possible to use data from Call Reports, rather than data from the FR 2910q, in the construction of the monetary aggregates. (See footnote 9 for the definition of Call Report.)

6. The reduced reporting limit was implemented by the Federal Reserve in September 2003 to improve the coverage of weekly-reported deposit data.

7. No adjustment is made to the exemption amount if total reservable liabilities at all depository institutions should decline. The annual growth rate is measured from June 30 one year to June 30 the next year, and then used in calculating the exemption amount for the subsequent year. (It is implemented beginning with the first maintenance period in the subsequent year that includes January 1.)

8. In this document the term Call Report will refer to the commercial bank Consolidated Reports of Condition: (FFIEC 031and 041; OMB No. 7100-0036); the Thrift Financial Report (OTS 1313; OMB No. 1550-0023); and the Statement of Financial Condition (NCUA 5300/5300SF; OMB No. 3133-0004).

9. The reserve ratios that apply to the nontransaction account components of reservable liabilities are zero.

10 The bank credit family of reports comprise the Weekly Report of Selected Assets (FR 2644; OMB No. 7100-0075), the Weekly Report of Assets and Liabilities for Large U.S. Branches and Agencies of Foreign Banks (FR 2069; OMB No. 7100-0030), and the Weekly Report of Assets and Liabilities for Large Banks (FR 2416; OMB No. 7100-0075).

11. The reporting weeks for the quarterly respondents begin on the third Tuesday of March, June, September, and December. For the purposes of constructing the monetary aggregates, weekly deposits and vault cash for quarterly respondents are estimated between quarterly report dates from reported movements at a class of small weekly FR 2900 reporters. When actual data from the quarterly FR 2900 reporters become available, these weekly estimates are adjusted. Reserve requirements for quarterly respondents are satisfied during weekly maintenance periods, and are set quarter-by-quarter based on the data reported for a single week each quarter.

12. If total transaction accounts, savings deposits, and small time deposits at all depository institutions decline during the period, no downward adjustments would be made, i.e. the nonexempt deposit cutoff and the reduced reporting limit would remain at their levels from the previous year.

13. See Appendix A, Table A.2. Note that while some FR 2900 weekly reporters that are eligible to shift to
FR 2900 quarterly reporting may choose not to do so, for purposes of this document, burden reduction calculations are based upon the eligible number of shifters.

14. Institutions will continue to be added to the FR 2910a reporting panel based on total deposits (not total transaction accounts, savings deposits, and small time deposits) because nonpersonal savings and time deposits (a component of FR 2910a data item 2, Reservable Liabilities) typically include some large time deposits.

15. Any non-FR 2900 institution that does not file a December Call Report, or whose Call Report is not readily available, must submit an FR 2910a. Any institution that adjusts its FR 2910a reported values in order to qualify for reduced reporting will be shifted to an FR 2900 reporting panel.

16. Note that under this proposal, when calculating the indexed levels of the nonexempt deposit cutoff and reduced reporting limit to be announced in the fall of 2006 (to become effective in September 2007), the levels proposed in this memo for 2007 would be indexed based on 80 percent of the growth rate of total transaction accounts, savings deposits, and small time deposits between June 30, 2005, and June 30, 2006. In performing this calculation, total transaction accounts, savings deposits, and small time deposits as of both June 30, 2005 and June 30, 2006, would have to be estimated for some FR 2910a reporters because certain Call Reports—OTS 1313 and NCUA 5300—do not include data on the split between small and large time deposits. The estimation procedure would be needed again for the indexation of the nonexempt deposit cutoff and reduced reporting limit in the fall of 2007 (to become effective in September 2008) but only for the June 30, 2006 base period. After the sum of total transaction accounts, savings deposits, and small time deposits would be available on the FR 2910a, beginning June 30, 2007, the indexation performed in the fall of 2008 (to become effective in September 2009) would no longer require such estimation since the new measure would be available in both the base period (June 30, 2007) and the end period (June 30, 2008) used for indexation.

17. Federal Reserve Regulation D requires that offices of an Edge or agreement corporation with the same parent bank, and branches and agencies of the same foreign bank located within the same state and the same Federal Reserve District, file an aggregated FR 2900. However, separate FR 2900 must be filed if such institutions have offices in more than one state or District. In addition, for a brief period following a merger, some domestic institutions are allowed to submit separate FR 2900 data to facilitate consolidation of reporting and other operational systems.

18. Further changes in the allocations may be made as follows: (1) When a new office of an institution already filing an allocation report is established, the low reserve tranche and/or reservable liabilities exemption allocation for any or all of the offices may be changed as of the first reserve computation period beginning in any calendar month; or (2) if, under the existing allocation, the low reserve tranche is not being fully utilized during each reserve computation period by a depository institution that filed an allocation report, or if the existing allocation is having an adverse effect on the operations of the institution, the allocation may be changed as of the first reserve computation period beginning in any calendar month.

19. Darrel Parke and Hank Leddon in the Micro Statistics Section, Division of Research and Statistics, prepared this study.

20. The growth rate in total deposits is measured from June one year to June the next, and the new indexed value of the deposits cutoff goes into effect at the time of the panel shift in the subsequent year (unless the Federal Reserve chooses to override the indexed value and implement a higher cutoff).

21. The estimation procedure for the quarterly reporters works this way: The surrogate panel is formed each week from the smallest EDDS reporters for that week. (Institutions involved in a merger are excluded from membership in the surrogate panel for the week of the merger.) Weekly growth rates of the deposits at the EDDS surrogate panel are applied to the deposits of quarterly reporters by growing the last reported quarter-end levels one week at a time, for each successive week in the quarter, according to the weekly growth in the EDDS surrogate panel, to reach estimates of the next quarter-end levels. When data from the new QEDS and Call Reports become available, they replace the estimated (grown) quarter-end levels. The quarter-end revisions are then smoothed in over the previously estimated weekly data for the quarter, so that weekly changes are adjusted by a constant equal to the quarter-end revision divided by the number of weeks in the quarter. The pattern of weekly data for QEDS and Call reporters thus still reflects the relative weekly growth rates from the surrogate panel.

22. Reservable time and savings deposits were combined into a new memorandum item nonpersonal time and savings deposits; personal time deposits and the breakdown of nonpersonal time deposits by maturity were combined into one item time deposits; data on MMDAs and other savings deposits were combined into one item savings deposits; telephone transfer accounts were combined with the item ATS accounts and NOW accounts/share drafts; and the definition of item 2 in schedule AA was broadened to include all such transactions with original maturities of seven days or more.

23 . The two items with reduced reporting frequency were nonpersonal savings and time deposits and ineligible acceptances and obligations issued by affiliates maturing in seven days or more.

25