Do Five Agreeing Technical Indicators Provide Five Pieces of Evidence?
Identify overlapping inputs in moving averages, MACD, and RSI. Distinguish composite buy ratings from independent evidence and assess information value by removing indicators one at a time.
Even when several indicators point in the same direction, the number of independent pieces of evidence has not increased if they merely transform the same price data. Rather than counting indicators, check whether their inputs and the questions they answer differ. A composite rating summarizes predefined calculation rules; it does not automatically represent a profit probability or a business valuation.
Classify inputs and calculation structures rather than names
Moving averages and MACD, which uses the difference between two averages, share a price-smoothing structure. RSI and stochastic use different calculations but are both influenced by the recent price path. This explains why one large advance can generate positive readings on several screens at once. Record each indicator’s raw data, period, transformation, and intended purpose in a checklist. Start by avoiding treating several lines based on the same closes and similar periods as opinions from separate market participants.
Votes in a composite rating are not probabilities
Some platforms apply individual criteria to multiple moving averages and oscillators to display a composite rating. Having many items labeled “buy” means many of that platform’s conditions have been met. A score that sums related conditions is not a proportion of independent successful cases. Reading the formula helps reveal whether the same characteristic—that current price is above averages of several lengths—is counted repeatedly. Alongside the rating label, it is useful to state which price conditions generated the score.
Check whether the added indicator changes the decision
Before adding another indicator, identify the unresolved question in the records produced by the existing rules. For example, if direction is understood but the size of price swings is not, a volatility measure can add a different question. Even so, do not assume the inputs are completely independent. Over the same historical window, remove indicators one at a time and compare changes in candidate count, decision timing, number of trades, and the path of losses. Little difference may mean you added only complexity; even if there is a difference, check whether it recurs in other periods.
Worked example
Hypothetical example: Today’s price is 110 won, and the 5-day, 10-day, and 20-day averages are 108 won, 106 won, and 104 won. Counting price above each average as a vote produces “agreement among three indicators.” But these are views of the recent upward price path through overlapping windows. Adding a holding of 100 shares and an assumed loss of 8 won per share produces an exposure-check figure of 800 won. That information answers a different question from the number of directional votes.
What to check
- Record each indicator’s raw data and observation period.
- Check the individual decision formulas behind a composite rating.
- Do not turn the number of agreeing items into a success probability.
- Compare decision changes and costs before and after removing indicators.
Sources and primary materials
This is educational material about research methods. For an actual company decision, verify the filings and primary sources for the relevant period.