Relative Strength Analysis for CMT Candidates: Ratios, Rankings, and Rotation

Relative strength needs a comparison

In comparative market analysis, relative strength (RS) asks how one security or group performs against another. It is distinct from the Relative Strength Index (RSI), a bounded momentum indicator calculated from one asset's gains and losses. The 2026 CMT Program Guide includes relative-strength definitions, ratios, ranking, and portfolio uses at Level I, with advanced applications and RRG concepts at Level II. Level III applies relative strength to portfolio construction and sector rotation. Those are verified public objectives, not a promise of any particular ratio's exam frequency.

A ratio line is normally asset price divided by benchmark price using consistently adjusted series. If it rises, the numerator is outperforming the denominator over that interval, even if both prices fall. Without the benchmark and starting date, “strong” has no precise meaning.

A worked ratio example

Suppose a sector fund starts at 50 while its benchmark starts at 100. Initial RS ratio = 50/100 = 0.50. Later the fund is 55 and the benchmark 105, so the ratio is 55/105 ≈ 0.5238. The fund gained 10%; the benchmark gained 5%. The ratio rose by 0.5238/0.50 − 1 ≈ 4.76%. That ratio gain is not the same as the simple five-percentage-point difference between the two returns. The methods describe related but different comparisons.

Now imagine the fund falls from 55 to 53 while the benchmark falls from 105 to 98. The new ratio is 53/98 ≈ 0.5408, higher than 0.5238. The fund is falling in absolute price but improving relative to the benchmark because the benchmark falls faster. A portfolio manager who needs absolute capital preservation should not confuse relative leadership with a positive return.

Price adjustments matter. A large dividend or split can change an unadjusted ratio abruptly even if total economic performance did not. The fund and benchmark should be converted to consistent currencies and time stamps. A ratio between unlike closing sessions may record a currency or calendar artifact.

Ranking a universe

Suppose four sectors have six-month total returns of +12%, +8%, +2%, and −3%, while the benchmark returns +5%. Their simple excess returns are +7, +3, −3, and −8 percentage points. Ranking by these values gives the same order as ranking by raw returns because the benchmark is common, but it does not tell you risk-adjusted quality, liquidity, or trend stability. A six-month window can favor a recent reversal differently from a one-month window.

One practical research process is to fix a tradable universe, benchmark, lookback, rebalancing date, and eligibility filter before calculating ranks. For example, rank 20 sectors monthly by six-month total return versus a common benchmark; consider only sectors whose absolute price is above its own long-term moving average; rebalance at the next scheduled close. That is an illustrative rule, not a recommendation or official CMT procedure. A filter can avoid some negative-trend leaders, but may also enter late.

Sector rotation is a portfolio decision, not merely a chart ranking. If the top two sectors each receive 25% of a 100,000 portfolio, each allocation is 25,000 and half the portfolio remains for the other specified holdings or cash. If the two sectors are highly correlated, that 50% can represent concentrated economic risk. A rank does not set an appropriate weight by itself.

RRG and momentum of the ratio

The public Level II objectives mention JdK RS-Ratio, JdK RS-Momentum, and Relative Rotation Graphs (RRG). These display relative level and relative momentum versus a benchmark. The four commonly named quadrants are leading, weakening, lagging, and improving. A point in “improving” signals a particular combination of relative readings within that display, not an automatic buy. Exact calculations involve proprietary normalization and smoothing details; avoid reverse-engineering the plotted value from a raw price ratio alone.

The benchmark choice can reverse a conclusion. A technology sector might outperform a broad equity index but underperform another growth-sector benchmark. State the comparison that answers the investment question. A global portfolio may also need currency-adjusted comparisons.

Risk and invalidation

A ratio making higher highs while absolute price falls can justify relative preference but cannot justify an unqualified bullish claim. A conditional sector-selection rule might require both a rising RS ratio and a positive absolute trend. It could be invalidated by the ratio falling below its prior swing low or by the sector failing the absolute-trend filter at the next review. The timing of the review and transaction costs matter. If a rule changes every day, turnover may consume any historical advantage.

Use relative strength for diagnosis as well. If an index rises but fewer sectors outperform it, leadership may be narrowing. That is a participation warning; see Market Breadth. Relative ratios can also be affected by the benchmark's large constituents, so pair rank tables with constituent and exposure checks.

Common mistakes and self-check

Do not call RSI a market-relative measure. Do not compare raw share prices across securities; price level is arbitrary after splits and share issuance. Do not mix price-only and total-return series. Do not claim “outperformance” without naming the benchmark and interval. A ranking backtest must avoid survivorship bias and include formerly eligible sectors or securities, not only today's winners.

Self-check: Asset A rises from 80 to 88, benchmark B from 200 to 220. Does A outperform? Answer: Both gain 10%. Their ratio is 80/200 = 0.4 initially and 88/220 = 0.4 later. A did not outperform over that period, even though its price rose.

Next, plot one adjusted asset/benchmark ratio and the asset's own price underneath it. Mark a period where the ratio and absolute price disagree, then write the distinct decision each chart supports.