Nature of the Linear Progression Across the Liquidity Spectrum
Are single instrument rules valid or should they be confirmed by results in other markets? If so, what basket of stocks, bonds, futures, etc. is best used to test against?
What I believe you are asking is whether the "physics" that describe the movements of prices in one market are the same as in other markets. When stated this way, the answer is clearly no. "Rules" that work in one market, e.g. stocks, do not work in others, e.g. gold or bonds. For example, stock prices tend to make rounded tops and spike bottoms, like the path of a bouncing tennis ball. Gold, on the other hand, tends to make rounded bottoms and spike tops.
My belief is that this is related to nature of the linear progression across the liquidity spectrum. Oooph! Big words! Let me explain.
I contend that it is possible to array financial instruments on a one-dimensional axis (like a number line) related to their relative liquidity, or better yet, to their "money-likeness". In this sense, actual stocks are more "money-like" than options or futures which are more ethereal. And similarly, cash is more "money-like" than shares of stock. So when investors panic, they panic out of options and into stocks. They panic out of stocks, and into cash. Progressing further in the direction of money-likeness on this linear scale, it is presumable that investors in a cash currency might panic out of that cash, and into something more money-like, e.g. gold (or canned food, or shotgun shells).
So this brings us to the paradox of why stock prices make spike lows, but gold makes spike highs. The answer lies in the units we use. Both stocks and gold are expressed in $/unit, i.e. $/share or $/ounce. But that does not reflect the correct relationship on the liquidity spectrum. To clear up the conflict, the units should be arranged with the more liquid instrument in the numerator and the less liquid instrument in the denominator. We already have that relationship established for stock prices in $/share, but if we use this liquidity-based convention then gold should be expressed in ounces/$. If it were, then the chart patterns for gold would be inverted, and the spike tops would then be spike bottoms, clearing up the conflict. Panics would result in a sharp decline in the ounces/$ "price", just as stock market panics result in sharp declines in the $/share "price".
For "markets" or "cross rates" like board feet of lumber per T-Bond, or bushels of apples per pound of copper, which don't share the same panic linearity, the physics should be expected to be different. And so, therefore, the rules that one uncovers to try to exploit those physics should therefore also be different from "market" to market, and must therefore be adjusted/tuned for the market being studied.