J.P. Morgan Chase & Co.
270 Park Avenue
Floor 28
New York, NY 10017
Telephone: 212-270-7559
Facsimile: 212-270-9589
August 2, 2002
Mr. Jonathan G. Katz, Secretary
U.S. Securities and Exchange Commission
450 Fifth Street, NW
Washington, D.C. 20549-0609
Subject: File No. S7-16-02
Dear Mr. Katz:
J.P. Morgan Chase & Co. appreciates the opportunity to comment on the Securities and Exchange Commission's ("the Commission") Proposed Rule: Disclosure in Management's Discussion and Analysis About the Application of Critical Accounting Policies ("Proposed Rule") Release Nos. 33-8098 and 34-45907. The stated goal of the Proposed Rule is to enhance investors' understanding of the application of companies' critical accounting policies to yield greater transparency about how past results may be used as a gauge of possible future performance. We support initiatives to enhance investors' understanding of financial results. We are concerned, however; that the Proposed Rule as we interpret it will not achieve improved understanding since it emphasizes the isolated impact of a change in discrete material assumptions rather than a broader perspective of a firm's results. We think that there are better approaches available than those contained in the Proposed Rule to achieve the stated goals, such as the Commission's concept to disclose what management considers the principal factors in determining the financial results.
Our key concerns with the Proposed Rule are as follows:
Following are our recommendations for changes to the Proposed Rule to reduce scope and focus while meeting the Commission's stated objectives.
Specific Recommendations Regarding the Critical Accounting Estimates Proposal
Scope of the Rule
We believe the definition, as described in the Proposed Rule, of a critical accounting estimate is too broad. The shift from the principles established in Financial Reporting Release (FR) No. 60, Cautionary Advice Regarding Disclosure About Critical Accounting Policies, to the set of rules established under the Proposed Rule will require a significant number of additional disclosures. We support the scope of FR-60 as summarized by Robert K. Herdman, Chief Accountant, on January 24, 2002: "A critical accounting policy is one that is both very important to the portrayal of the company's financial condition and results, and requires management's most difficult, subjective or complex judgments." Applying this principle to our year-end filings, we identified two critical accounting policies. A comparative review against selected competitors' disclosures identified that this was consistent in number and selection. While the content of registrants' disclosures under FR-60 did not satisfy the Commission, we strongly believe the scope was appropriate.
In contrast, under the Proposed Rule, the Commission outlines the following detailed rules to determine what is a critical accounting estimate:
1) It requires the company to make assumptions that are highly uncertain (as defined) at the time of the estimate; and
2) Another estimate could have reasonably been used in the current period, or changes in the estimate are reasonably likely to occur from period to period and would have a material impact.
Based on these criteria, we preliminarily identified the following items as potential critical accounting estimates. Each item listed has at least one assumption that meets the Commission's definition of highly uncertain and changes in that assumption would have a material impact.
Additionally, depending on materiality, the following estimates may also be in the scope of the Proposed Rule:
A significant amount of disclosure is already provided for several of these items under existing requirements either in the footnotes or in the MD&A. We support the existing substantial qualitative and quantitative disclosure requirements. However, the Proposed Rule appears to require that these current disclosures be repeated in a section of MD&A as critical estimates. For example, six different sections of our Annual Report would contain a discussion of the Allowance for Loan Losses. We urge the Commission to refine the scope based on the principles of FR-60.
Furthermore, we urge the Commission to eliminate the requirement to disclose the potential financial statement impact of possible alternative accounting methods. When management adopts a policy, they adopt the most appropriate policy. Management evaluates results based on the consistent application of the one policy; they do not view a business based on various accounting methods and then select one to disclose. As such, we do not think investors would find it useful to know what the financial results would have been under a less appropriate policy. The Proposed Rule would require maintenance of an alternate set of records that is not done today and would be inconsistent with how management views the results of the business. We support a qualitative disclosure of the adopted policy containing an explanation of the methodology, assumptions, and consistent application.
Nature of the Rule
The Proposed Rule requires a sensitivity analysis of estimates based on changes that result from "(i) making reasonably possible, near-term changes in the most material assumption(s); or (ii) using in place of the recorded estimate the ends of the range of reasonably possible amounts which the company likely determined when formulating its recorded estimate." In developing our current critical estimates, we do not calculate ranges. Thus, we would be required to apply the first alternative. The difficulty we have in applying this to our current critical accounting policies and other accounting estimates is that there are numerous interrelated assumptions in the estimates. The breadth of disclosure from separately stress-testing the individual material assumptions and discussing the results of the analysis would be massive. For example, we would be required to disclose an analysis that shows the impact to the Allowance for Loan Losses of changing, in isolation, the significant assumptions for each component of the allowance. We do not believe that this type of disaggregated data could be presented to an investor in a format that is meaningful, particularly because management does not analyze the allowance in this manner. An investor attempting to interpret all of the data surely will be bewildered.
The Commission has already identified the principal concept for a more meaningful disclosure model for financial results. Specifically, the Commission states in the Proposed Rule that it is considering a separate disclosure proposal for "what management considers the most important factors in determining its financial results and condition, including the principal factors driving them, the principal trends on which management focuses and the principal risks to the business." We anticipate that such a disclosure model would include some of the disclosures in the Proposed Rule but in a much more meaningful fashion. We would support the Commission further developing this "management perspective" macro-level type of disclosure rather than the detailed estimate approach.
We endorse the development of a more comprehensive model because it should focus on the broader economic factors, trends, and concentrations that affect the overall results of the business rather than the potential variability of single assumptions in various line items. In Attachment I, we provide an example of how we envision this type of disclosure could apply for our trading portfolio. We are not proposing this in addition to what we already disclose, but as a replacement or rationalization of the relevant disclosures. If the Commission adopted this approach, we would expand the disclosure to all items carried at fair value by business segment. Other registrants would make similar comprehensive disclosures for their most critical estimates as selected by them. If the Commission does adopt such a comprehensive disclosure approach, it should consider postponing the implementation of an accelerated timeframe for 10-Q and 10-K filings in order for registrants to adopt this rule appropriately.
We identified the following benefits of adopting a more comprehensive model versus the Proposed Rule:
We appreciate the opportunity to submit comments on your proposal and hope that you find our recommendations for disclosure improvements useful. If you have any questions or if we can be of further assistance, please contact me at 212-270-7559.
Very truly yours,
/s/
Joseph L. Sclafani
Executive Vice President & Controller
This attachment is intended to illustrate a potential disclosure method. We anticipate that companies would apply judgment in determining the content and format of their disclosures based on the factors that management believes are most pertinent. We do not believe that a specific format for disclosure should be mandated.
Critical accounting policies used by the Firm
JPMorgan Chase's (the "Firm") accounting policies are integral to understanding the results reported. The Firm's most complex accounting policies require management's judgment to ascertain the valuation of assets and liabilities. The Firm has established detailed policies and control procedures that are intended to ensure valuation methods are well controlled and applied consistently from period to period. In addition, the policies and procedures are intended to ensure that the process for changing methodologies occurs in an appropriate manner. The following is a description of the Firm's current policies involving significant management valuation judgments as it relates to trading assets and trading liabilities.
Fair Value Methodology
Fair value is defined as the value at which positions could be closed out or sold in a transaction with a willing and knowledgeable counterparty over a period of time consistent with the Firm's trading strategy. The fair value of a majority of our trading positions is determined based on quoted market prices. If listed prices or quotes are not available, then fair value is based on internally developed models. These models use as their basis independently sourced market parameters including, for example, interest rate yield curves, option volatilities, and currency rates. The valuation process then takes into consideration factors such as counterparty credit quality, liquidity, and concentration concerns. Management applies judgment in the determination of these factors. For example, there is often limited market data to rely on when estimating the impact of holding a large or aged position. Similarly, judgment must be applied in estimating prices where no market parameters exist. Finally, other factors can affect estimates of fair value including market dislocations, model assumptions, and relationships between various market parameters, i.e., correlations. Imprecision in estimating these factors can impact the amount of revenue or loss recorded for a particular position. The largest valuation adjustment relates to credit risk associated with the derivatives in the Firm's portfolio. The credit valuation adjustment, or CVA, represents the estimated credit component of the Firm's derivative counterparty exposures. The CVA is recalculated on a daily basis, taking into account market and counterparty-related activity, such as changes in credit ratings and market prices, master netting agreements, and collateral. CVA also includes additional adjustments for name-specific credit risk not reflected in credit spreads.
For some positions, the fair value is determined using internally developed models, where certain inputs to the model are not verifiable in the market. Generally, these types of contracts are complex derivatives, e.g. long dated currency options, and represent approximately x% of the trading portfolio at December 31, 2001. To ensure the mark-to-model valuations of such contracts is appropriate, the Firm has implemented various compensating controls. These controls include an independent review and approval of valuation models, detailed review and explanation for profit and loss analyzed daily and through time, decomposing mark-to-model valuations into components and benchmarking valuations where possible to similar products, and the validation of valuation estimates through settlement of actual cash flows. Importantly, we ensure a consistent approach is employed through time. (Note: Any material changes in methodology or assumptions and the impact of those changes would be disclosed).
Trading Strategy and Instruments
The Firm utilizes financial instruments to meet the financial needs of its customers, to generate revenues through its trading activities, to manage its exposure to fluctuations in interest and currency rates, and to manage the Firm's own credit risk. These market making, trading, and investing activities encompass a broad range of cash and derivative financial instruments in interest rate, foreign exchange, equity, and credit markets. The net risk position the Firm takes in these instruments creates earnings volatility.
Trading assets include debt and equity securities held for trading purposes that the Firm owns ("long" positions). Trading liabilities include debt and equity securities that the Firm has sold to other parties but does not own. These securities are "short" positions, and the Firm is obligated to purchase these securities at a future date. Trading assets and liabilities also include the fair value of trading derivatives. All of the Firm's trading assets and liabilities are carried at fair value on the Consolidated Balance Sheet, with changes in the fair value recorded in Trading Revenue. Trading positions are recorded on a trade date basis. The reported receivables (unrealized gains) and payables (unrealized losses) related to derivatives include the effect of master netting arrangements as permitted under FASB Interpretation No. 39.
Management of Trading Instruments
Two of the major risks to which the Firm's trading portfolios are exposed are Market Risk and Credit Risk.
Market Risk - Management
The Firm employs a comprehensive approach to market-risk management for its trading portfolios. JPMorgan Chase controls market risk primarily through a series of limits. Approved by the Firm's Risk Management Committee, these limits align specific risk-taking activities with the overall risk appetite of the Firm and of the individual business divisions.
Limits are set based on analyses of both the market environment and business strategy. The analyses encompass criteria such as market volatility, liquidity of the products involved, business track record, and management experience and depth. JPMorgan Chase reviews risk limits regularly and the Firm's Market Risk Management Group updates the risk limits at least twice a year. The Market Risk Management Group further limits the Firm's exposure by specifically designating which financial instruments each business division may trade.
Market Risk - Measurement
Because no single measure can reflect all aspects of market risk, the Firm uses several risk measures, both statistical and nonstatistical. This combined approach gives JPMorgan Chase a more comprehensive view of its exposure, and enhances the stability of the Firm's revenues from market-risk-taking activities.
JPMorgan Chase's statistical risk measure Value-At-Risk, or VAR, gauges the dollar amount of potential loss from adverse market moves in an ordinary market environment. Each business day, the Firm undertakes a comprehensive VAR calculation that includes the trading portfolios. JPMorgan Chase's VAR calculation is highly granular, comprising more than 500,000 positions and 220,000 market prices, e.g., securities prices, interest rates, and foreign-exchange rates.
To calculate VAR, the Firm uses historical simulation, which measures risk across instruments in a consistent, comparable way. This approach assumes that actual historical changes in market prices are representative of possible future changes. The simulation is based on data for the previous 12-month period. All statistical models involve a degree of uncertainty, depending on the assumptions they employ. The Firm prefers historical simulation because it involves fewer assumptions about the distribution of portfolio losses than parameter-based methodologies.
To evaluate the soundness of the Firm's VAR model, JPMorgan Chase conducts daily back testing of VAR against actual financial results. In addition, the Firm rigorously assesses the quality of the market data, since their accuracy is critical to computing VAR.
JPMorgan Chase calculated the VAR numbers in the table below using a one-day time horizon and a 99% confidence level. This means the Firm would expect to incur losses greater than predicted VAR estimates only once in every 100 trading days or about 2.5 times a year. For 2001, JPMorgan Chase posted positive daily market-risk related revenue for x out of x days, with x days exceeding positive $x million. No daily trading losses were incurred in excess of $ x million. Of the x days on which JPMorgan Chase posted trading losses, the actual trading losses exceeded the VAR on x days, a result that does not differ significantly from the 99% confidence level.
| Year Ended December 31, 2001 | ||||
| Average VAR | Minimum VAR | Maximum VAR | At December 31, 2001 | |
Trading portfolio: |
||||
Interest rate |
xxx | xxx | xxx | xxx |
| Foreign exchange | xxx | xxx | xxx | xxx |
| Equities | xxx | xxx | xxx | xxx |
| Commodities | xxx | xxx | xxx | xxx |
| Hedge fund investments | xxx | xxx | xxx | xxx |
| Less: Portfolio diversification | xxx | xxx | xxx | xxx |
| Total trading VAR | xxx | xxx | xxx | xxx |
Market Risk - Stress Testing
While VAR reflects the risk of loss due to unlikely events in normal markets, stress testing captures the Firm's exposure to unlikely but plausible events in abnormal markets. Stress testing is equally important as VAR to the Firm in measuring and controlling its risk. The Firm stress tests its portfolios at least once a month, at both the corporate and line-of-business levels, using multiple scenarios. Scenarios are continually reviewed and updated to reflect changes in the Firm's risk profile and economic events.
The Firm's stress testing methodology assumes that, during an actual stress event, no action would be taken to change the risk profile of the Firm's portfolios. This lets the Firm capture the decreased liquidity that often occurs with abnormal markets, and results, in the Firm's view, in a conservative stress test result.
The following table represents the potential economic value stress test loss (pre-tax) in JPMorgan Chase's trading portfolio predicted by JPMorgan Chase's stress test scenarios:
| Year ended December 31, 2001 | ||||
| (in millions) | Average | Minimum | Maximum | At December 31, 2001 |
| Stress test loss - pre tax | $(xxx) | $( xxx) | $(xxx) | $( xxx) |
Note: Companies would provide the appropriate type of market risk disclosures based on management's judgment of factors that are pertinent to their business.
Credit Risk Management
Credit risk management begins with an assessment of the risk of loss resulting from the default of a borrower or counterparty. The Firm uses the same credit risk management procedures for derivative and foreign exchange transactions as those used for traditional lending products. In addition, the Firm actively manages the credit risk of the derivative counterparties with credit, interest rate, foreign exchange, equities, and commodity derivatives.
In terms of credit risk outstanding exposure, the true measure of risk from derivative and foreign exchange contracts is the mark-to-market value of the contracts at a point in time (i.e., the cost to replace the contract at the current market rates should the counterparty default prior to the settlement date). For most derivative transactions, the notional principal amount does not change hands; it is simply an amount that is used as a reference upon which to calculate payments. While notional principal is the most commonly used measure in the derivative and foreign exchange markets, it is not a measure of credit risk. The $x trillion of notional principal of the Firm's derivative and foreign exchange contracts outstanding at December 31, 2001, does not represent a proxy for, and significantly exceeds, the possible credit losses that could arise from such transactions. At December 31, 2001, the associated mark-to-market value of the contracts, or the amount owed to the Firm, was $x billion after taking into consideration the benefit of legally enforceable master netting agreements. Further, after taking into account $x billion of collateral held by the Firm at year-end, the net credit exposure owed to the Firm relating to derivative and foreign exchange contracts was $x billion.
The Firm is operating in a recessionary environment and has seen an increase in credit-related costs. Nonperforming trading derivative contracts increased/decreased in 2001 from $x million to $x million. The Firm's focus is on distribution of credit and on continuing the reduction of its commercial credit exposure, both on- and off-balance sheet. Note: We would provide the fair value of the derivative portfolio by the relevant rating category.
Fair Value of Trading Instruments by Product, Geography and Industry
The following tables present information on the fair value of trading assets and trading liabilities in a manner consistent with how management monitors the position, by product. The majority of these positions are in the Investment Bank.
| Fair Value at December 31, 2001 | |
| Trading assets | |
| Debt & equity instruments: | xxx |
| US Governments | xxx |
| Certificates of Deposit | xxx |
| Foreign government debt securities | xxx |
| Corporate securities & others | xxx |
| Total trading assets - debt & equity instruments | xxx |
| Derivative receivables: | |
| Interest rate contracts | xxx |
| Foreign exchange contracts | xxx |
| Debt, equity, commodity & other contracts | xxx |
| Total trading assets - derivative receivables | xxx |
| Trading liabilities | |
| Total trading liabilities - debt & equity instruments | xxx |
| Derivative payables: | |
| Interest rate contracts | xxx |
| Foreign exchange contracts | xxx |
| Debt, equity, commodity & other contracts | xxx |
| Total trading liabilities - derivative payables | xxx |
Trading-related revenues, which include trading-related net interest income, increased/decreased by x% due to the following factors {described}. The following table sets forth the components of total trading-related revenue.
| 2001 | 2000 | 1999 | |
| Trading revenues: | |||
| Equities | xxx | xxx | xxx |
| Fixed income & other | xxx | xxx | xxx |
| Trading revenue | xxx | xxx | xxx |
| Net interest income impact | xxx | xxx | xxx |
| Total trading-related revenue | xxx | xxx | xxx |
Selected Geographic Exposures
The Firm has a process for measuring and managing its country exposures and risk. The following table presents JPMorgan Chase's exposure of trading activities to selected countries at December 31, 2001. This disclosure is based on management's view of country exposure.
| December 31, 2001 | |
| Brazil | xxx |
| Mexico | xxx |
| Argentina | xxx |
| Venezuela | xxx |
| South Africa | xxx |
| Japan | xxx |
| Indonesia | xxx |
| Turkey | xxx |
| United States | xxx |
Industry Diversification Profile
JPMorgan Chase focuses on diversifying its commercial credit-related trading assets. The table below illustrates the Firm's 10 largest credit-related industry groups.
| Debt & equity instruments | Derivatives | Total | |||
| Investment Grade | Noninvestment Grade | Investment Grade | Noninvestment Grade | ||
| Commercial Banking | xxx | xxx | xxx | xxx | xxx |
| Holding & Investment Companies | xxx | xxx | xxx | xxx | xxx |
| Securities Brokers, Dealers, Exchanges | xxx | xxx | xxx | xxx | xxx |
| Investment Management/Private Banking | xxx | xxx | xxx | xxx | xxx |
| Telecommunications Services | xxx | xxx | xxx | xxx | xxx |
| Central Government | xxx | xxx | xxx | xxx | xxx |
| Real Estate | xxx | xxx | xxx | xxx | xxx |
| Investment & Pension Funds | xxx | xxx | xxx | xxx | xxx |
| Oil & Gas Exploration/Production | xxx | xxx | xxx | xxx | xxx |
| Business Services | xxx | xxx | xxx | xxx | xxx |