Patent Application
20100114745
A portion of the disclosure of this patent document may contain material which is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure as it appears in the Patent and Trademark Office patent files or records, but otherwise reserves all copyrights whatsoever.
Business entities, e.g., banks, enter into a large number of transactions in the ordinary course of their operations. Some of these transactions carry financial risks such as currency or foreign exchange (FX) risks, commodity price risks, interest rate risks, stock price risks, and counterparty risks, to name a few. For example, individual loans carry the risk of debtor default, currency exchange rate fluctuations, or changing interest rates for variable rate loans or imminently mature loans (whose principal likely will be reinvested at a new interest rate). Financial risks are often affected by market data changes such as changing costs or currency values. For example, an increased or decreased value of the U.S. Dollar relative to the Euro affects FX risks in transactions involving those currencies. As another example, an increased or decreased cost of a commodity like aluminum, copper, gold, or oil affects the commodity price risks for that commodity.
Typically, the business entities' internal policies or banking regulations of governing regulatory bodies, e.g., the International Accounting Standards Board (IASB), which has promulgated International Financial Reporting Standard (IFRS) 39, Financial Instruments: Recognition and Measurement, or the Financial Accounting Standards Board (FASB), which has promulgated Financial Accounting Statement (FAS) 133, Accounting for Derivative Instruments and Hedging Activities, require, at least in some instances, that the business entities own instruments, typically derivatives such as options, whose behavior counterbalances risks presented by the transactions. This is called “hedging.”
Risk exposures presented by a first, typically numerically large, set of instruments are counterbalanced by performance of a second, typically much smaller, set of instruments (called “hedging instruments” herein), such that when risk rises with respect to the instruments that present the risk exposures, risk falls in the hedging instruments. For example, a set of instruments are grouped and treated as a single exposure that is to be hedged. One or more hedging instruments counterbalance the exposure group. The exposures or exposure groups and their corresponding hedging instruments are grouped into corresponding hedging relationships. A hedging relationship associates one or more particular hedging instruments with a particular exposure or exposure group. Accordingly, use of hedging relationships aids in management of risk exposures and corresponding hedging instruments and facilitates compliance with hedging policies or regulations.
Additionally, the business entities' internal policies or banking regulations of governing regulatory bodies such as the IASB or FASB typically require, at least in some instances, that the business entities prospectively test their risk exposure management strategies (for example, the effectiveness of their hedging instruments) by simulating future market data. Both short-term (e.g., 5-30 days) and long-term (e.g., 2-10 years) market data are relevant to risk exposure management; thus, market data forecast models should accurately simulate both short- and long-term trends to improve hedging effectiveness.
Previous strategies for simulating future market data, however, do not accurately preserve correlation values—for example, the relationship between the price of oil and the relative strength of the U.S. Dollar versus the Euro—while accurately accounting for short- and long-term trends. For example, extrapolating historical data into the future maintains correlation values but ignores short-term trends. On the other hand, the well-known Monte Carlo modeling method preserves correlation values in the short term but is too noise-sensitive (that is, too easily influenced by random irrelevant or meaningless data) to accurately model long-term trends. A hybrid approach generates a Monte Carlo model while assuming a steady market data change rate, but this model sacrifices correlation value accuracy.
Available computer applications employ improved strategies for simulating future market data. For example, existing applications may employ scenarios, which forecast future market data based on hypothetical values at specific times input by the user. For example, a user may assume that the Dow Jones Industrial Average closes at 11500 in 5 days. Based on that assumption (i.e., “scenario”) and current market data, the application might calculate a market data change rate and apply it to predict, for example, the value of the U.S. Dollar relative to the Euro at the Dow's close in 5 days. The user, however, must manually enter his or her assumptions. Moreover, since scenarios are calculated from absolute values, an output value may not be “reused”; that is, although a user may create a model based on multiple assumptions, he may not use that model to create another model based on new or different assumptions. For example, a user may design a scenario that predicts the price of oil in 5, 10, and 15 days based on certain values; that scenario, however, is useless to predict the price of oil in 20 days. The user must instead enter new hypothetical values and rerun the scenario application. Existing applications may also employ “shifts,” which apply mathematical formulae to current market data to predict market data values for a particular time in the future. Like scenarios, shifts require manual entry of data, assumptions, and/or formulae.
Even these improved strategies, however, are ill-suited to hedging or risk management. For example, these strategies may employ random values to simulate the market's reaction to, for example, strategy changes within a company. In this case, the same input set may yield different output sets upon multiple executions. This lack of reproducibility may cause auditing difficulties, especially if a business entity's hedging percentage (i.e., the percent of financial risk that is hedged) is low (for example, 85% or less). In addition, any data for documentation or auditing may be stored in the application, not the system, resulting in more storage space consumed.
More complex strategies maintain correlation values while applying them to current market data; however, these strategies require continually rebuilding the model to reflect the latest market data. Thus, applications which employ these complex strategies consume massive resources and are appropriate only for the most sophisticated users, such as, for example, brokerage firms.
The common business user thus requires a simpler, more efficient solution which simulates both short- and long-term development of market data while preserving correlation values.
Embodiments of the present invention enable users to maintain market data change rate (“MDCR”) information for different risk factors using a unified interface. In example embodiments, a single interface may facilitate maintaining, storing, and/or retrieving market data. Embodiments of the present invention may further provide access to historical market data and/or simulate future market data.
Embodiments of the present invention relate to a market data solution which may handle at least one of: extracting and storing historical information, calculating market data change rate sets, and/or applying market data change rate sets to forecast future market data.
Embodiments of the present invention enable a customer to manage historical market data and simulate long-term development of future market data. In example embodiments, these simulations may be used to predict, for example, the fair market value of a currency or commodity at a given time in the future. In example embodiments, these simulations may also model, for example, how market data changes affect risk exposure positions. Because they implicitly preserve correlation values and produce consistent results when repeated with identical input, example embodiments of the present invention are well-suited for testing hedging and risk management strategies while providing the reproducibility desired for auditing.
Embodiments of the present invention further provide a central storage place for market data, eliminating the need for data replication in each application. The tool is flexible and may be easily enhanced or adapted to existing or future customer systems, such as customer planning and production systems. In example embodiments, the tool provides seamless integration into existing hedge accounting solutions in compliance with, e.g., IFRS 39 or FAS 133. In an embodiment, the tool provides seamless integration into the analytical components of Treasury and Risk Management, which makes both internal operational Risk Management and external Risk Reporting (e.g., according to IFRS 7, Financial Instruments Disclosures) more efficient and accessible.
For purposes of illustration, the below example embodiments of the present invention largely concern stock price and foreign exchange (FX) risks. However, the embodiments may be used for other purposes as would be evident to one of skill in the art. For example, embodiments of the present invention may manage or simulate all kinds of risks, such as financial risks including foreign exchange risk, interest rate risk, commodity price risk, stock price risk and counterparty risk.
In step 101, example embodiments of the present invention extract and store historical market data. The data may be extracted from one or more sources of financial data according to any method known in the art. Alternatively, some or all historical market data may be manually entered into the system.
In step 102, example embodiments of the present invention generate market data change rate (“MDCR”) sets, which model market data values as change rates over given periods. As contemplated by embodiments of the present invention, an MDCR set may include market data change rates (“MDCR data points”) generated from the same historical market data. As an example, one MDCR data point may indicate that the price of oil is forecast to increase by 3% annually over the next 180 days. Another MDCR data point may indicate that the Euro is forecast to gain 2% relative to the U.S. Dollar in the next 180 days. Still another MDCR data point may indicate that the price of oil is forecast to decrease by 4% over the next 365 days, and so on. By thus reflecting relative changes, the model implicitly preserves correlation values. For example, the model could preserve the correlation between the price of oil and the relative strength of the U.S. Dollar versus the Euro while accurately modeling long-term trends.
In example embodiments, the system extrapolates historical market data (for example, the data stored in step 101) using, for example, the well-known Elliott wave principle. In this way, the model simulates development curves and cycles while ignoring short-term tremors caused, for example, by day-trading. Similarly, example embodiments incorporate any random values employed by the formulae into the forecast change rate, resulting in a condensed and, therefore, repeatable value that may be archived for auditing purposes. Moreover, by assigning each MDCR set a unique identifier, embodiments of the present invention may enable automatic versioning of forecast data.
In step 103, example embodiments of the present invention apply one or more of the MDCR sets calculated in step 102 to known or hypothetical market values, such as to current market data. In this way, embodiments of the present invention may forecast, for example, the fair market value of a hedging instrument over a period of time. Since example embodiments apply forecast values independently, the system need not generate new MDCR sets for each data point to be forecast. Rather, the system may calculate MDCR sets periodically (e.g., monthly) but apply them as demanded by daily use. Thus, decoupling the calculation (step 102) and application (step 103) of MDCR sets not only facilitates versioning and archiving of forecast market data but also results in faster execution since the MDCR calculation is performed less frequently. To that end, embodiments of the present invention initially calculate a constant rate of return over a long period, for example from 1920 through 1988. As new market data is extracted into the system, embodiments of the present invention periodically generate and store MDCR sets from the new data into the system. As the new MDCR sets are generated and stored, embodiments of the present invention annualize or otherwise combine the stored change rates, creating condensed MDCR sets which may be applied to subsequent market data.
For illustration purposes, the example in
In step 301, the system checks to ensure that all required input parameters are present. As an example, a procedure to simulate FX scenario values may require as input: a pointer to one or more MDCR sets (as, for example, calculated and stored in step 102 of
In step 305, the system calculates the FX scenario value for each Horizon Date. First, the system determines a change rate c for each Term in the Horizon Table, as shown in step 305a. In an example embodiment, the change rate c for a Horizon Table Term of Δt days is the change rate of the shortest MDCR Value Table Term that is greater than or equal to Δt days; if no MDCR Value Table Term is greater than or equal to Δt days, then the change rate c for that Horizon Table Term is the change rate of the longest MDCR Value Table Term. In an alternative embodiment, the change rate c for a Horizon Table Term of Δt days is the change rate of the longest MDCR Value Table Term that is shorter than or equal to Δt days; if no MDCR Value Table Term is shorter than or equal to Δt days, then the change rate c for that Horizon Table Term is the change rate of the shortest MDCR Value Table Term. Returning to the above example, if the MDCR Value Table contained change rates of +5%, +1%, and −3% for Terms of 50, 40, and 20 days, respectively, as shown in
In step 305b, the system uses the change rate c determined in step 305a to calculate an FX scenario value for each Horizon Date. Example embodiments calculate the FX scenario value using a compound change rate formula, wherein the change rate compounds daily, i.e., for a Term of Δt days:
Scenario_Value(Δt)=Start_Value*(1+c)Δt/365
Returning to the example in
In step 305c, the system appends the FX scenario value calculated in step 305b to an output table, for example as shown in
As another example,
In step 501, example embodiments of the present invention determine whether a dividend schedule exists for the given stock. If so, the system converts the dividend schedule into the given stock's currency, as shown in step 502.
In step 601, the system checks the input parameters to ensure that all required data to convert the dividend schedule is present. For example, in addition to the schedule itself (as shown, for example, in
If all input parameters are valid, then the system creates a local Horizon Table wherein each Horizon Date corresponds to a dividend payment due date in the dividend schedule table, as shown in step 603. Example embodiments first verify that the dividend schedule table contains only one currency from which to convert, as shown in step 603a. If a second currency is found in the dividend schedule table, then the system raises an error and/or throws an exception as shown in step 602, and the process ends. Otherwise, the system traverses through each record in the dividend schedule table, creating a corresponding local Horizon Table entry, as shown in step 603b.
In step 604, the system calculates the FX scenario value for each record in the local Horizon Table, for example using the process illustrated in
Returning to . If any input parameters are missing or invalid, then the system raises an error and/or throws an exception as shown in step 504, and the process ends. Otherwise, the system retrieves the stored MDCR values which correspond to the input into an MDCR Value Table, as shown in step 505. In the above example and given the data set shown in
In step 507, the system calculates the stock scenario value for each Horizon Date. First, the system determines a change rate c for each Term in the Horizon Table, as shown in step 507a. In an example embodiment, the change rate c is calculated in the same manner as described in step 305a above. Returning to the above example, if the MDCR Value Table contained change rates of +5%, +1%, and −3% for Terms of 50, 40, and 20 days, respectively, as shown in
In step 507b, the system uses the change rate c determined in step 507a to calculate an initial stock scenario value for each Horizon Date. Example embodiments calculate the initial stock scenario value in the same manner as described in step 305b above, i.e., for a Term of Δt days:
Init_Scenario_Value(Δt)=Start_Value*(1+c)Δt/365
Returning to the example in
If a dividend schedule exists for the given stock, then the system next adjusts the stock scenario value based on the dividend schedule, as shown in step 507c. Example embodiments use the converted dividend schedule generated in step 502 to calculate an adjusted scenario value by subtracting each dividend payment Di due at Δti during the Term, i.e., where Start_Date<Δti<Δt. An example embodiment also subtracts the compounding effect of each Di. That is, Scenario_Value(Δt)=
Returning to the example in
If the system encounters an error while calculating the adjusted stock scenario value (for example, if the converted dividend currency differs from the reference currency assigned to the risk factor), then it returns an error and/or throws an exception as shown in step 504, and the process ends. If step 507c completes without error, or if no dividend schedule exists for the given stock, then the system appends the stock scenario value calculated in step 507b or 507c to an output table, as shown in step 507d and illustrated, for example, in
Returning to
Likewise, those skilled in the art can appreciate from the foregoing description that the present invention can be implemented in a variety of forms. For example, the above embodiments may be used in various combinations with and without each other. Therefore, while the embodiments of the present invention have been described in connection with particular examples thereof, the above embodiments are for illustration purposes only and are not meant to limit the scope of the present invention. Other modifications will become apparent to the skilled practitioner upon a study of the present application.
This application is related to U.S. patent application Ser. No. 12199775, filed Aug. 27, 2008, entitled “System and Method for Exposure Management.”
