What are RSI and moving averages, and why limitations matter?
RSI (Relative Strength Index) and moving averages are technical indicators that translate past price data into numeric values. RSI typically measures the magnitude of recent price gains versus losses over a chosen lookback window. A moving average smooths price by averaging it over a chosen number of periods, producing a line that can be used to judge trend direction and momentum.
Their limitations matter because the “story” they tell depends on assumptions: which price you feed in (close vs. other), how many periods you use, and what you consider a relevant market regime. When those assumptions don’t match your situation—or when the market stops behaving like the past—RSI and moving averages can become less informative.
How RSI and moving averages work (stable mechanics first)
RSI is commonly computed from average gains and average losses over a fixed lookback period. The result is scaled into a bounded range (often described as 0 to 100). A higher RSI value indicates that recent gains have dominated recent losses; a lower RSI indicates the opposite.
A moving average is computed by taking the mean of a price series over a rolling window. If you use more periods, the line usually smooths more but reacts more slowly to change. If you use fewer periods, it reacts faster but tends to be more sensitive to noise.
In both cases, the mechanics are stable: given the same price series, timeframe, and parameter choices, the formulas produce the same numerical outputs. What changes in real use is the market data context and how you interpret the outputs.
Evidence and example of why they can fail
A common failure mode is “whipsaw” in range-bound or choppy conditions. Imagine a market that repeatedly moves up and down within a relatively narrow range. A moving average can switch from “below” to “above” price multiple times as the smoothed line lags behind short moves. Meanwhile, RSI can oscillate within its own bounded range, repeatedly approaching levels that people often associate with strength or weakness.
Even if the indicator values are computed correctly, the interpretation can be unstable because the market’s direction is not sustained long enough for trends to develop. In that situation, both RSI and moving averages may generate frequent changes in bias that do not correspond to meaningful follow-through.
Another limitation is sensitivity to parameter selection. If you change the RSI lookback window or the moving average length, the timing of turning points and the frequency of crossings can change substantially. Two analysts can look at the same chart but use different settings and reach different conclusions, even without making a mathematical error.
Material limitations and risks to verify independently
-
Lag and regime dependence (trend vs. no-trend): Moving averages generally react to change after it happens, because they are based on averages of past periods. RSI, while not a moving average, still depends on recent gains and losses. If the market shifts from one regime (trend) to another (range or sudden reversal), the same readings may no longer mean what they did before.
-
False confidence from historical relationships: Indicators are derived from historical price patterns. Historical relationships do not establish that the same relationship will persist in future conditions. This uncertainty applies whether you use RSI alone, a moving average alone, or both together.
-
Hidden assumptions about data and execution: Even when indicators are “parameter-based,” the inputs come from market data. Differences in timeframe, the exact price used, and how data is sampled can alter indicator values. Additionally, real-world outcomes can be affected by costs and execution, so indicator-based expectations may not map cleanly to realized results.
-
Combination does not eliminate uncertainty: Using RSI with a moving average can sometimes make interpretations more selective (for example, filtering trades to align with a perceived direction). However, combining indicators mainly changes how you decide; it does not remove the core limitations: noise, regime shifts, parameter sensitivity, and the non-guarantee nature of pattern-based reasoning.
How to verify claims about RSI and moving average limitations
You can verify limitations without relying on predictions by checking how the indicators behave under different conditions in your own historical data.