92 CHAPTER ◆ 8 Describe Trading/Investment Idea
the logic implications. Every time our trading system receives a price update from the
market, it must recalculate the moving averages and understand the new state. From lowest
value to highest, these states are- Price, short MA, long MA
- Price, long MA, short MA
- Long MA, price, short MA
- Long MA, short MA, price
- Short MA, price, long MA
- Short MA, long MA, price.
Now consider also that such a system, fully automated, may or may not have a position
on long or short. So, multiply these six states by 3—long, short, flat—for 18 possible total
states. Also, our system must be able to identify crossovers. If the previous state was that
the short MA was above the long MA, but upon recalculation, the short MA is now below
the long MA we have to raise a flag indicating a bearish (or alternatively, a bullish) signal.
Remember, MAs are one thing, but our system trades on prices, and when it takes a posi-
tion, it may bracket the position with a target price and a stop price. But not just any stop
price, a trailing stop price. And one last thing, what happens if a partial fill is received and
the state changes? Should the system cancel the remaining quantity?
This is a lot of logic, a lot of “if ” statements. The number of possible combinations
grows exponentially as the number of logic statements grows. Trading logic like this can
quickly turn into a spaghetti bowl of logical comparisons; if any potential state is unac-
counted for, the trading system may go out of control. Logic leaks are dangerous. Putting
pen to paper forces a product team to think through all logical outcomes. A clear logic
map or logic flow diagram will enable them to plan for and prove all logical outcomes.
We recommend that you map all of your logic before you start prototyping. Any
change to the logic must first be changed in your logic map, because changes will have
unthought of implications in other places. You must be sure that all states are accounted
for, that all logical outcomes are handled. Once you have started coding, it is too late; you
will not be able to figure it out on the fly.8.3. STEP 1, LOOP 1: Review Money Document
Description
The K|V methodology begins with a review of the input, the trading/investment strategy
description from the Money Document, and a plan for research.We know of cases where the research drives the development of new trading systems. In K|V, the Money
Document, a sales tool, provides the impetus for research and development.In the step following this one, K|V 1.2, team members will benchmark the quantita-
tive methods and business rules that comprise the trading/investment system ’ s strategy.
Disciplined research will ensure that teams produce proper specifications, establishing a
common understanding of what the different parts of the trading system will do and, fur-
thermore, facilitating testing of prototypes since testers can check whether the prototype