The Multiplication Principle of Multiple Independent Programs - Mathematical Foundations for Reducing Signal Failure Rates

Zen Theory Mar 24, 2026 01:55

Any single stock selection program inevitably faces signal failure, and reducing the failure rate of a single program below a meaningful threshold proves exceedingly difficult. This essay introduces the multiplication principle from probability theory for independent events, demonstrating that combining multiple mutually independent programs can compress the composite failure rate to remarkably low levels. It further discusses the construction logic for three categories of independent programs — technical indicators, relative valuation and capital flow, and fundamental analysis — along with the criteria for verifying genuine independence among them.

The Multiplication Principle of Multiple Independent Programs - Mathematical Foundations for Reducing Signal Failure Rates

I. The Inevitability of Signal Failure and Its Distinctive Difficulty

The preceding essay established a binary classification framework for investment targets: all stocks at any given moment are to be divided into actionable and non-actionable categories, with operations strictly confined to the former. Regardless of which classification program is employed, one fact remains inescapable: no program can guarantee that every stock admitted to the actionable category will ultimately produce a sustained upward trend. In other words, every operational program must inevitably confront the problem of signal failure — the target briefly advances after satisfying entry conditions, then rapidly retraces, invalidating the breakout and converting the entry into a trapped position.

Signal failure constitutes the most intractable challenge in investment operations for a fundamental reason: it cannot be confirmed prior to entry. Unlike the market activity screening discussed in the previous essay — where a chronically dormant, wholly inactive target can be identified and excluded without committing any actual capital — signal failure reveals itself only after the investor has established a real position. Every instance of signal failure detection therefore carries genuine capital risk. Compounding the difficulty is the lack of stable predictability: the same stock exhibiting the same technical pattern may produce a valid breakout on one occasion and an immediate reversal on another. Past success under identical conditions provides no guarantee of future outcomes.

Many investors and market commentators claim that when certain specific technical formations or market conditions appear, a stock "will definitely" rise. In practice, however, every condition combination carries a substantial probability of signal failure. Conflating "high probability of validity" with "certainty of validity" ranks among the most dangerous confusions in investment cognition. A great many losses in the market arise not from choosing the wrong analytical framework, but from overestimating the certainty of a correct framework, thereby neglecting position sizing and risk control.

II. The Primary Response Principle Following Signal Failure

Before discussing how to reduce the signal failure rate, a prerequisite principle must be established: once signal failure occurs, the position must be exited immediately. This principle admits no exceptions. Even if the target subsequently resumes its advance and develops a powerful uptrend, no regret or second-guessing should attach to the exit decision.

The logical foundation of this principle is as follows: the occurrence of signal failure means the conditions upon which the original entry program relied have been violated. Continuing to hold the position at that point is equivalent to engaging in unstructured speculation without any programmatic support. Even if the target does subsequently recover, that recovery constitutes a new, independent market event unrelated to the original entry logic. Inferring the correctness of a process from its outcome is the most classic fallacy in probabilistic reasoning.

Furthermore, if a target is to re-enter an effective uptrend after signal failure, it typically must pass through an extended adjustment period. During this interval, capital remains locked in a target of extremely high uncertainty, forfeiting the opportunity to engage other targets currently exhibiting valid breakouts. The universe of available investment targets is vast; consuming capital and time on a target that has already emitted a failure signal is unacceptable from both capital efficiency and opportunity cost perspectives.

These principles represent only the most basic response framework. If the investor has established a rigorous system for scaled entry and scaled exit, combined with a comprehensive position management regime, the losses caused by signal failure can be further compressed to manageable levels. Position management is a broad subject that will receive dedicated treatment in subsequent essays. The remainder of this essay addresses a different dimension: how to reduce the probability of signal failure at the level of entry program design itself.

III. The Upper Bound of Single-Program Failure Rates and the Multiplication Principle

The fundamental cause of signal failure is that any entry program is, at best, an approximate description of market dynamics rather than a precise replication. The actual operation of markets inevitably produces anomalous situations that fall outside any program's descriptive scope, causing the program's output signals to diverge from the market's actual trajectory. For any specific program, its signal failure rate can be estimated through long-term backtesting against historical data.

Consider the simplest possible random program: deciding whether to buy based on a coin toss. This program's signal failure rate is at least 50%. Clearly, any thoughtfully designed program should perform significantly better than random, but even so, reducing a single program's failure rate below 10% is extraordinarily difficult. A program with a stable failure rate below 10% would mean that out of every ten operations, no more than one results in failure — a performance level that is virtually impossible to maintain consistently over the long term in practice.

However, the situation is far less hopeless than it appears. A fundamental principle from probability theory can radically transform this landscape: the multiplication rule for independent events. This rule states that if multiple events are mutually independent, the probability of their simultaneous occurrence equals the product of their individual probabilities.

Applied to investment program design: suppose three mutually independent entry programs have signal failure rates of 30%, 40%, and 30% respectively — all quite ordinary, unremarkable programs. If the rule is that a buy is executed only when all three programs simultaneously emit entry signals, then the composite signal failure rate of the program group is 30%×40%×30%=3.6%30\% \times 40\% \times 30\% = 3.6\%30%×40%×30%=3.6%. This means that out of one hundred operations following this program group, signal failure would occur fewer than four times. The leap from individual failure rates of 30–40% to a composite failure rate of 3.6% is achieved entirely through mathematical structure, requiring no extraordinary predictive power from any single component program.

The practical implications of this conclusion are profound: the investor does not need to pursue a single "perfect" indicator or method. It suffices to identify several individually adequate but mutually independent programs and combine them. Pursuing the ultimate refinement of a single indicator is a dead end; constructing combinations of multiple independent programs is the correct direction.

IV. Constructing Three Categories of Independent Programs

The critical prerequisite of the multiplication principle is genuine mutual independence among the component programs. If three programs are highly correlated with one another, combining them produces no reduction in failure rate whatsoever — three programs that essentially measure the same thing, however combined, remain functionally a single program. Therefore, the core challenge in program group design lies not in increasing the number of programs but in ensuring independence among them.

The following outlines one viable approach to constructing three categories of independent programs.

The first category is based on the technical indicator system. The input variables of all technical indicators are ultimately derived from price and volume data, which means all technical indicators share varying degrees of correlation. MACD, RSI, KDJ, Bollinger Bands, and so forth all draw from homogeneous underlying data sources and do not possess genuine mutual independence. For this reason, within the technical indicator dimension, the investor need only select any single indicator to construct a trading program; stacking multiple technical indicators provides no gain in terms of independence. For more technically proficient investors, a candlestick chart incorporating moving average systems and volume information provides richer information than any single technical indicator's output and can serve as the complete program for this dimension.

The second category is based on relative valuation and capital flow dynamics. No individual stock exists in isolation; it occupies a position within the broader market's web of relative price relationships. The relative pricing of individual stocks versus sector indices, sectors versus the broad market index, and the domestic market versus international markets — changes in these relative prices reflect the directional flow of capital among different targets. When capital persistently flows from one sector to another, relative price relationships undergo systematic shifts that can be quantified and structured into independent trading signals. It should be noted that all indicators directly related to market capital flow, including sentiment indicators, turnover rate dynamics, and institutional fund monitoring data, are highly correlated with relative valuation programs and cannot be treated as independent.

The third category is based on fundamental analysis, though the term "fundamental" here extends well beyond conventional financial statement analysis. Traditional fundamental metrics such as price-to-earnings ratios, return on equity, and revenue growth rates are certainly part of the picture, but in practical investment operations, the most discriminating fundamental factors are often those concerning the behavioral logic and interest structures of market participants. For example, the management of large state-owned enterprises, driven by reputational considerations, typically will not tolerate their company's stock price remaining below its IPO price for extended periods. Issuers of stocks with associated put warrants, motivated by the desire to avoid exercise and settlement obligations, possess an intrinsic incentive to support the stock price. The buying and selling behavior of industrial capital reflects the value judgments of insiders who understand the company's operating conditions most intimately. This class of analysis, grounded in participant interest structures and human behavioral logic, possesses genuine independence from both purely price-volume-based technical analysis and capital-flow-based relative valuation analysis, because its information sources and analytical logic operate in an entirely different dimension.

V. Independence Verification as the Core of Program Group Design

The three program categories outlined above are illustrative. Any investor may design their own independent program group according to their own knowledge base and circle of competence. But regardless of design, the absolute prerequisite for the multiplication principle to function is genuine independence among the component programs. If three substantially correlated programs are combined, the composite failure rate will show no meaningful decline, and the mathematical advantage of the multiplication principle will be entirely nullified.

The basic method for determining whether two programs are independent is to trace their underlying information sources and logical chains. If the input data of two programs can ultimately be traced to the same information source (for example, both derived from price and volume data), or if the logical derivation processes of two programs share common intermediate links (for example, both depending on judgments about market capital flow direction), then they lack independence. Only when two programs draw from information sources in completely different dimensions, with no shared nodes in their logical chains, can they be considered independent programs whose failure rates may be validly compressed through the multiplication principle.

This verification work may seem abstract, but in practice it is of paramount importance. A large number of investors believe they are employing a "multi-confirmation" approach by simultaneously consulting multiple technical indicators or observing multiple capital-flow-related data points. Because these indicators and data are fundamentally highly correlated, the supposed "multi-confirmation" amounts to nothing more than repeated affirmation of the same signal and possesses no mathematical efficacy in reducing the failure rate. Genuinely effective multi-confirmation must be founded on true independence across information source dimensions.

VI. Systematic Methods and Market Rhythm

The discussion of the multiplication principle ultimately aims to transform investment operations from a mode dependent on individual judgments to one dependent on systematized program groups. Any single judgment, whether its source is technical analysis, fundamental analysis, or market intuition, inevitably carries a relatively high failure rate. But when judgments from multiple independent dimensions are systematically integrated, the composite failure rate decays at multiplicative speed, establishing for the investor a genuinely reliable decision-making foundation in the probabilistic sense.

The application of this methodology is not confined to individual stock selection. At the more macroscopic level of sector rotation, the same logic applies with equal force. Rotational movement among different sectors in the market follows discernible patterns: when a leading sector enters a phase of consolidation at elevated levels, market capital tends to flow toward previously lagging sectors, initiating a new rotation cycle. The rhythm of such sector rotation can likewise be incorporated into the multi-independent-program analytical framework as an auxiliary dimension for entry timing decisions.

What the investor needs is not to predict whether a specific stock will rise or fall tomorrow, nor to chase a specific thematic narrative, but to build a complete, systematized analytical and operational framework based on multiple independent dimensions, letting probability and mathematical structure work on their behalf. Specific investment targets and entry timing are the natural outputs of this framework in operation, not products of subjective conjecture. Once the methodology is mastered, the recognition of opportunity becomes a matter of natural consequence.

Image source: Shutterstock