9
discuss which parts to buy and which to build for better performance, let alone compete
on the NASCAR circuit.
To the lay investor, building a trading/investment system might be as simple as the
NASCAR example—buy a few off-the-shelf parts and away you go. But, unless you truly
understand the workings of a trading/investment system inside and out, you cannot even
begin to discuss how to fine tune performance in the real markets, against real compe-
tition. Real professionals understand that a fully functioning trading/investment system
blends together off-the-shelf and proprietary components and algorithms. Benchmarks
additionally enable the running of real-world and what-if scenarios.1.8. A Quality Approach to Development
Over the course of this book we apply quality techniques to systematic trading and invest-
ment. A systematic, methodical approach can control development, evaluation, and ongoing
management of financial trading/investment systems. Furthermore, a process approach will
provide a consistent framework for stepped allocation of seed capital for development and
knowledge and team management. Applying a quality development methodology will sat-
isfy the most important quality principle—customer satisfaction—as well as the interests of
seed capital providers, employees, and regulators.
Our four-stage DTIM (Design, Test, Implement, Manage), 16-step trading sys-
tem development methodology forms the structure of this book. The four DTIM stages
spiral in a Plan-Benchmark-Do-Check (PBDC) framework, revolving around quan-
titative methods, data, technology, and portfolio monitoring, respectively. (Our meth-
odology is in sync with the investment process proposed by Grinold and Kahn—
“ researching ideas (quantitative or not), forecasting exceptional returns, constructing and
implementing portfolios, and observing and refining their performance ” as set forth in
their seminal book Active Portfolio Management. We have implemented some of Grinold
and Kahn ’ s methods at investment firms, and for the underlying financial concepts we
have chosen not to address, we strongly recommend their text. This is also true of the
books published by Jacobs and Levy. This text focuses on the engineering aspects of
implementing the concepts presented in these and the many other books on quantitative
finance.)
In each stage and at each step of a trading/investment system development project
exist both risk and opportunity. For example, research into quantitative methods gives
rise to model risk. Prototyping in Excel/Resolver gives rise to spreadsheet risk. Back-
testing gives rise to data cleaning risks and optimization risk. Deming also presented
seven deadly diseases to articulate certain risks. We have developed our own list
of seven deadly diseases for trading/investment system design and development. They are
a lack of:- A skilled, balanced, multifunctional product team.
- Understanding of investment cycles and lack of a marketing plan that incorporate
investment cycles. - Complete description of the trading or investment strategy.
- Adequate backtesting and execution system testing for shadow trading.
- Data and misunderstanding of data.
1.8. A QUALITY APPROACH TO DEVELOPMENT