Our experience in risk management, applied mathematics, and software development has provided solutions to institutions that include the New York Clearing House, Deutsche Bank, HSBC, KPMG, and various hedge funds and boutique securities issuers. In addition to performing purely mathematical analysis, we offer hands on development and project management for implementing software in such languages as C++ and JAVA ORACLE PL/SQL on UNIX and NT platforms. Ed’s computing experience includes devising, prototyping and patenting data compression algorithm United States Patent #5537551.
MULTI-MILLION DOLLAR RETURNS FROM “GAMMA RENT” INTEREST RATE CAP TRADING METHODOLOGY
A methodology for identifying “cheap” interest rate caps and quantifying their departure from fair value was developed by formalizing the concept of “gamma rent.” A position in the cap, delta hedged by a FRA, was found to change its value overnight due to
- time decay of the option, including the effect of riding down the forward rate curve and the forward volatility curve;
- the hedging FRA’s change in value due to the effect of riding down the forward rate curve; and
- the interest earned due to investment of the option premium.
Applying, among other things, Black’s Model and Ito’s Lemma, formulas were derived to take advantage of these characteristics by calculating the “gamma rent” of the option. A low gamma rent implied that rates do not require high volatility for the long option position to have positive value. The formulas, which were then coded in C, produced a report showing market opportunities based on gamma rents of the more liquid caps.
DESIGN OF WEB BASED RAROC CALCULATOR THAT RE-ALLOCATED $85 MILLION AWAY FROM UNFAVORABLE LOANS IN FIRST 18 MONTHS.
A multi-trillion dollar institution had software, widely viewed as obsolete and an undue burden on network resources, for pre-deal determination of the appropriateness of risk-return characteristics for large commercial loans. Major savings were realized after a proposal for rewriting the software using more modern technology was accepted and implemented under direct supervision, with personal participation in the coding of critical analytics.
STRESS TEST AND VALIDATION OF FUNDAMENTAL CHANGE TO INTERBANK PAYMENT SYSTEM FOR DAILY CLEARING OF $2 TRILLION.
A consortium of money center banks sought to implement multilateral netting in which a payment from Bank A to B and B to C would be replaced by a single net payment from A to C. Independent review was complicated by the code having been written in uncommented ALGOL. Nevertheless, a functional description of the algorithms was developed, a number of flaws in the detailed design were identified, and basic parameters of the project were confirmed. By eliminating unnecessary payments, the system dramatically decreased the likelihood of system failure.
SYSTEM DEVELOPED FOR MONITORING $3 BILLION PER DAY IN REPURCHASE AGREEMENTS (“REPOS”)
Existing methodology for estimating Repo credit risk at a multi-trillion dollar institution was both incomplete and arbitrary. A comprehensive methodology for measuring credit risk was drafted and implementing software written, with previously undiscovered risks identified.
SALE OF MULTI-TRILLION DOLLAR INSTITUTION?S FIRST INDEX PRINCIPAL SWAP
Sale of the institution’s first product in a new line was slated to fail because of the customer?s refusal to do the deal at the institution’s quoted note yields.
Viewing the problem less as “quant” than as a trader, and observing that forward interest rates are slightly higher over New Year’s weekend, a solution was found, and the deal closed, with a forward starting structure that began and ended two weeks later than originally envisioned.