What is a worked example of RSI?
A worked example of RSI is a fully numeric walk-through that starts with an explicit list of prices, states the RSI settings (such as the lookback length), computes gains and losses, averages them, and then calculates the final RSI value.
Below is one self-contained example using only fixed assumptions and no live market data.
RSI mechanism and definition
RSI stands for Relative Strength Index. It is a momentum oscillator that maps recent price changes into a bounded value between 0 and 100.
The standard RSI workflow is:
- Choose a lookback length N (commonly 14).
- For each step in the lookback, compute the change from the previous price.
- Split changes into:
- Gain: the positive change amount (otherwise 0)
- Loss: the absolute value of the negative change amount (otherwise 0)
- Compute average gain and average loss over the lookback.
- Compute the relative strength RS = (average gain) / (average loss).
- Convert to RSI: RSI = 100 − (100 / (1 + RS)).
Assumptions matter because different RSI implementations may average gains and losses differently (for example, using a smoothed method). The worked example below uses the simple averages so every step is directly verifiable.
Worked numerical example (all assumptions stated)
Assumptions
- Lookback length N = 5 (shortened from 14 so the arithmetic fits in one example).
- Prices are a sequence of closes measured at equal intervals.
- RSI uses simple averages of gains and losses across the N periods (not a smoothed average).
- No fees, spreads, or trading outcomes are involved; RSI is computed purely from the price sequence.
Input prices
Assume we observe these 6 closing prices:
- Day 0: 100
- Day 1: 102
- Day 2: 101
- Day 3: 103
- Day 4: 103
- Day 5: 104
That creates 5 consecutive changes (the lookback N periods):
- Change 1 (Day 1 − Day 0): +2
- Change 2 (Day 2 − Day 1): −1
- Change 3 (Day 3 − Day 2): +2
- Change 4 (Day 4 − Day 3): 0
- Change 5 (Day 5 − Day 4): +1
Split into gains and losses
For each change:
- Gain is the positive part; Loss is the magnitude of the negative part.
- Gain list: [2, 0, 2, 0, 1]
- Loss list: [0, 1, 0, 0, 0]
Average gain and average loss (simple)
- Average gain = (2 + 0 + 2 + 0 + 1) / 5 = 5 / 5 = 1.0
- Average loss = (0 + 1 + 0 + 0 + 0) / 5 = 1 / 5 = 0.2
Compute RS
- RS = 1.0 / 0.2 = 5
Convert to RSI
- RSI = 100 − (100 / (1 + 5))
- RSI = 100 − (100 / 6)
- RSI = 100 − 16.666… = 83.33 (approx.)
So, under these explicit assumptions, the RSI value is about 83.33.
Limitations and failure modes (material risks)
-
Calculation depends on data choices. If you use a different price series (for example, different timeframes, different source prices, or different intervals), the gains and losses change, and the RSI number changes.
-
Implementation differences can produce different RSI values. Some common RSI implementations use a smoothed averaging method for average gains and losses. If you reproduce the example with smoothing instead of simple averages, you may not get the same final RSI.
-
Edge cases can break intuition. If average loss is zero, RS becomes undefined or effectively infinite, and RSI approaches 100. Similarly, if average gain is zero, RSI approaches 0. These situations can occur with monotonic price moves over the lookback.
-
RSI is not a standalone prediction tool. Even when RSI is high or low, RSI only summarizes recent changes; it does not guarantee future direction. Outcomes also vary with market conditions and any execution costs, but those factors are outside the RSI calculation itself.
Verification and next question
To independently verify RSI from scratch:
- Write down the exact price list used.
- Choose the lookback length N.
- Decide and state whether you use simple averages or a smoothed method.
- Compute gains, losses, averages, RS, then apply the RSI formula.
If you want the next worked example, choose either:
- a case with both gains and losses more evenly mixed, or
- a case with consecutive declines to see how RSI moves toward 0.