RSI and relative strength are different tools: RSI measures a stock's momentum against its own recent gains and losses on a 0–100 scale, while relative strength divides a stock's price by a benchmark's price to show whether it is beating the market. One asks "has this run too far, too fast?" The other asks "is this a leader or a laggard?" At the October 16, 2026 close, the two flatly contradicted each other on tech: $XLK printed an RSI of 33.3, near textbook "oversold," while its price ratio to $SPY was up 10.6% over 126 sessions. Same words, opposite stories.
Same words, two different tools
The confusion is baked into the vocabulary. J. Welles Wilder introduced the Relative Strength Index in his 1978 book New Concepts in Technical Trading Systems, according to StockCharts ChartSchool. Despite the name, RSI compares a security with nothing but itself. Its internal "RS" term is average gain divided by average loss over a lookback window, 14 periods by Wilder's default.
Relative strength in the comparative sense, which ChartSchool also calls the price relative or Relative Strength Comparative, is a ratio chart: one security's price divided by another's. When the line rises, the stock is outperforming its benchmark. When it falls, the stock is lagging, even if its own price is going up.
Put bluntly: RSI is a speedometer. Relative strength is the race standings. A car can be braking hard and still be in first place.
How to calculate RSI
- For each session, record the gain (close minus prior close, if positive) or the loss (the size of the drop, if negative).
- Average the first 14 gains and the first 14 losses.
- After that, smooth: new average gain = (prior average gain × 13 + current gain) ÷ 14, and the same for losses. ChartSchool likens this to an exponential moving average.
- RS = average gain ÷ average loss. RSI = 100 − 100 ÷ (1 + RS).
Worked example, $XLK at the October 16, 2026 close: running Wilder's method on daily closes from January 2, 2025 gives an average gain of 0.7210 and an average loss of 1.4421. RS = 0.7210 ÷ 1.4421 = 0.50. RSI = 100 − 100 ÷ 1.50 = 33.3. Smoothed losses were running about twice the size of smoothed gains.
The standard convention, per ChartSchool: above 70 suggests "overbought," below 30 "oversold." That convention is where most of the trouble starts.
How to calculate relative strength
Relative strength = ETF (or stock) close ÷ benchmark close, on the same date. The level of the ratio means little on its own. Its trend is the signal, so traders measure the ratio's change over a window: RS change = (ratio now ÷ ratio N sessions ago) − 1.
Worked example from Financial Modeling Prep (FMP) end-of-day closes:
- October 16, 2026: $XLK 183.47 ÷ $SPY 763.25 = 0.24038
- April 17, 2026 (126 sessions earlier): $XLK 154.35 ÷ $SPY 710.14 = 0.21735
- RS change: 0.24038 ÷ 0.21735 − 1 = +10.6%
Cross-check: $XLK rose 18.9% in price over those 126 sessions while $SPY rose 7.5%, and 1.189 ÷ 1.075 − 1 is also about 10.6%. Relative strength is simply outperformance, compounded. Over the last 63 sessions, back to July 20, 2026, the ratio gained only 1.5%. Most of tech's lead was built between April and July.
The October 16, 2026 scoreboard: 15 sector ETFs
We computed both measures for the 11 Select Sector SPDR ETFs against $SPY and four iShares TSX sector ETFs against $XIU.TO, using FMP daily closes through October 16, 2026. RS columns show the change in the price ratio over 63 sessions (about three months) and 126 sessions (about six months). $SPY's own RSI(14) was 43.8; $XIU.TO's was 59.8. These are closing-price ratios, not total-return ratios, so dividends are left out, which slightly understates high-yield sectors such as utilities.
| ETF | Sector | Benchmark | Close | RSI(14) | RS 63d | RS 126d |
|---|---|---|---|---|---|---|
| $XLK | Technology | $SPY | 183.47 | 33.3 | +1.5% | +10.6% |
| $XLE | Energy | $SPY | 68.24 | 69.1 | +14.5% | +15.4% |
| $XLF | Financials | $SPY | 55.52 | 54.6 | -3.7% | -1.5% |
| $XLV | Health Care | $SPY | 176.71 | 66.0 | +7.9% | +10.5% |
| $XLI | Industrials | $SPY | 167.76 | 41.3 | -8.4% | -10.0% |
| $XLY | Consumer Discretionary | $SPY | 112.46 | 51.8 | -4.6% | -13.1% |
| $XLP | Consumer Staples | $SPY | 87.62 | 71.9 | +0.4% | -1.1% |
| $XLU | Utilities | $SPY | 42.21 | 60.9 | -8.7% | -14.9% |
| $XLB | Materials | $SPY | 50.92 | 58.3 | -1.0% | -8.7% |
| $XLRE | Real Estate | $SPY | 42.16 | 51.7 | -9.4% | -11.8% |
| $XLC | Communication Services | $SPY | 110.32 | 45.6 | -3.2% | -13.8% |
| $XEG.TO | TSX Energy | $XIU.TO | 31.24 | 71.8 | +19.0% | +12.8% |
| $XFN.TO | TSX Financials | $XIU.TO | 92.51 | 52.0 | -7.5% | +3.4% |
| $XIT.TO | TSX Info Tech | $XIU.TO | 83.18 | 56.0 | +13.7% | +6.7% |
| $XGD.TO | Global Gold (TSX) | $XIU.TO | 55.34 | 46.5 | +22.1% | -13.9% |
Read across the rows and the point is plain: RSI and relative strength do not move in lockstep. Health care ($XLV) and energy line up on both. Tech, staples and gold miners do not.
When they disagree: four worked examples
$XLK: an "oversold" leader
RSI 33.3, a hair above the 30 line. Yet the ratio to $SPY was up 10.6% over 126 sessions and 1.5% over 63. Read together, that is a leader having a pullback, not a broken sector. The 63-day number is the caution flag: the lead has flattened, and leaders do lose leadership.
$XLP: "overbought" with no lead
Staples carried the highest RSI among the US sector funds at 71.9. Relative strength? +0.4% over 63 sessions and −1.1% over 126. Over six months, staples have roughly matched $SPY. The 70-plus RSI reflects a short burst while $SPY's own RSI sat at 43.8, not a durable trend of outperformance.
$XLE: strong on both
RSI 69.1, RS +14.5% over 63 sessions and +15.4% over 126. Canada's $XEG.TO looked similar: RSI 71.8, RS +19.0% and +12.8% against $XIU.TO. Anyone selling purely because RSI neared 70 would have been fighting the strongest relative trend among the US sector funds. ChartSchool notes that momentum oscillators can become overbought and remain so in a strong uptrend.
$XGD.TO: the answer depends on the window
The gold-miner ETF's ratio to $XIU.TO was up 22.1% over 63 sessions but down 13.9% over 126. The ratio went from 1.1865 on April 16 to 0.8365 on July 16, then back to 1.0216 on October 16, 2026. Its RSI of 46.5 says nothing either way. One RS number is a snapshot of a chosen window; change the window and the verdict flips.
How trend traders combine them
A common practice, not a recommendation: use relative strength to decide what to watch, and RSI to judge when. Traders rank sectors or stocks by RS over three and six months and focus on the ones with rising ratios. Within that shortlist, they use RSI to spot pullbacks.
The thresholds shift in trends. ChartSchool cites Constance Brown's observation that RSI tends to range between 40 and 90 in a bull market, with the 40–50 zone acting as support, and between 10 and 60 in a bear market. So a trend follower watching a leader may treat a dip toward 40 as a pullback worth studying, rather than waiting for 30.
The flip side matters as much. A low RSI on a laggard is not a bargain sign. $XLI had an RSI of 41.3 with RS of −8.4% and −10.0%: middling momentum, sliding relative performance. Breadth adds a third lens; see how many stocks sit above their 200-day average. More indicator explainers live in our Analysis section.
Common mistakes
- Calling RS "RSI." A price ratio against $SPY is not the Relative Strength Index, and RSI's internal "RS" is not a market comparison. Mixing them up means reading the wrong chart.
- Treating 70 as a sell signal in a trend. $XLE and $XEG.TO sat near or above 70 with double-digit relative gains. In a strong trend, overbought can stay overbought.
- Treating 30 as a buy signal. RSI near 30 on a leader like $XLK and RSI near 30 on a laggard are different setups. Check the ratio first.
- Comparing a Canadian ETF to $SPY without currency. $XEG.TO trades in Canadian dollars and $SPY in US dollars, so their ratio also moves with the exchange rate. Compare TSX funds with $XIU.TO, or convert to one currency first.
- Trusting a single lookback. $XGD.TO's RS was +22.1% or −13.9% depending on the window. Check at least two.
Both indicators are built from past prices only, and this snapshot is a single date; neither says what comes next. For daily market coverage, visit Traders On Trend.
Bull/Bear Verdict
Bull Case: Where it helps: used together, the two tools separate a leader's pullback from a laggard's bounce. At the October 16, 2026 close, $XLK's 33.3 RSI with +10.6% six-month relative strength told a very different story from $XLP's 71.9 RSI with flat relative strength.
Bear Case: Where it misleads: both are derived from price alone, both depend on arbitrary lookbacks, and neither predicts anything. $XGD.TO's relative strength flipped from +22.1% to −13.9% just by changing the window from 63 to 126 sessions.
FAQ
Is relative strength the same as RSI?
No. RSI, introduced by J. Welles Wilder in 1978, is a 0–100 momentum oscillator built from a security's own average gains and losses. Relative strength (comparative) is the ratio of a security's price to a benchmark's price, read by its trend.
What RSI level counts as overbought or oversold?
By convention, above 70 is overbought and below 30 is oversold, using 14 periods. In strong trends RSI can hold high readings for long stretches, which is why trend traders often watch the 40–50 zone instead.
Which benchmark should I use for relative strength?
Use one in the same currency and market: $SPY for US sectors and stocks, $XIU.TO for Canadian ones. A stock can also be compared with its own sector ETF to see whether it leads its group.
What lookback works for relative strength?
There is no standard. This article uses 63 and 126 sessions, roughly three and six months. Checking two windows exposes cases like $XGD.TO, where the answer depends on the period chosen.