Sector Rotation for the CMT Exam: Rank, Allocate, Rebalance

Sector rotation is a repeatable allocation rule

Sector rotation shifts exposure among industries or sectors as leadership changes. The 2026 CMT Program Guide explicitly includes Level III objectives on constructing and running relative-strength and sector-rotation strategies. Level II also uses relative strength and asset weighting. Business-cycle stories can provide context, but sectors do not follow a fixed sequence in every cycle. A tradeable process needs defined data, ranking, weights, and risk limits.

Start with a stable, investable universe: identify the sector instruments available at each historical date and a broad benchmark. Use total-return data if distributions matter. A relative-strength ratio is sector price or total-return index divided by benchmark on synchronized dates. A rising ratio shows outperformance over that interval, even when absolute price is falling. Add an absolute trend filter if the mandate wants to avoid overweighting a declining market.

Build the rule in stages

1. Define the review frequency and lookback before seeing results.
2. Rank eligible sectors by a specified relative-return measure.
3. Require minimum liquidity and any absolute trend condition.
4. Allocate a known tactical sleeve, subject to sector and asset-class caps.
5. Model the next executable price, spread, commission, and tax implications.
6. State when ranks or risk limits trigger a rebalance.

A one-week rank may react quickly but turn over heavily; a one-year rank can lag a change. Comparing several lookbacks is reasonable during research, but selecting only the one that best fits the sample creates data-snooping risk. Evaluate the rule out of sample.

Worked portfolio decision

A 500,000 portfolio has a 20% tactical equity sleeve, or 100,000. The rule allocates half that sleeve to each of the top two qualifying sectors, subject to a 12% total-portfolio cap per sector. Sector A ranks first and Sector B second. A already has 20,000 in the strategic portfolio; B has 5,000.

A proposed 50,000 tactical A position would produce 70,000 total A exposure, or 14%, breaching the 60,000 cap. Only 40,000 more A is permitted. B's proposed 50,000 would produce 55,000, or 11%, within cap. The remaining 10,000 of the tactical sleeve can go to a third eligible sector if the rule permits or stay unallocated. It cannot simply be pushed into B if that would breach B's cap or the predefined allocation rule.

Suppose A's sector index rises from 100 to 108 while the benchmark rises from 200 to 210. A's relative ratio moves from 0.50 to 108/210 = 0.5143, roughly 2.86% higher. It outperforms, but that observation alone does not change the cap. If its relative line rolls over for the specified review period or its absolute trend breaks, the tactical overweight is invalidated; the strategic holding may remain under a separate policy.

Cycles and conflicting signals

An economic expansion can favor some sectors, but its precise phase is often only clear in hindsight. Rates, commodity prices, credit conditions, and policy can influence leadership differently across episodes. Use a cycle label as a hypothesis, not a fixed timetable. If macro narrative says one sector should lead but its relative price weakens, document the conflict and decide which evidence governs.

Sector ETFs also differ in composition. A nominally diversified sector may be concentrated in a few companies. Check holdings, country and currency exposure, liquidity, and factor overlap. Rotation among highly correlated growth sectors may do little for portfolio diversification.

Avoid hindsight in the sector universe

Historical sector data can be deceptive if today's index membership or ETF choices are projected backward. An industry may have been reclassified; an ETF may not have existed; a sector's dominant companies may have changed. Test with instruments or constituent definitions actually available at each date. If a proxy is used, state it and assess whether its history is economically comparable.

A rotation rule can also spend long periods in the wrong sector. Look beyond average return: how often does it rebalance, what is its worst drawdown, and how does it behave when leadership changes abruptly? If the model always holds the top two sectors, it may still own weak absolute trends during a broad bear market. An optional absolute trend filter changes exposure and must be evaluated as a separate rule.

For the worked 500,000 portfolio, imagine A rises after purchase so that its total position becomes 63,000 without a new trade. The 12% cap equals 60,000 at the original portfolio value, but the current portfolio value may also have changed. Recompute the cap from current equity and apply the mandate's drift or rebalance policy. A rule that caps only new purchases is different from one that requires immediate trimming after market appreciation. State which one applies.

If two leading sectors belong to the same broad growth factor, splitting the tactical sleeve between them may look diversified by sector label while retaining substantial common exposure. Compare their return behavior and economic drivers, especially under stress. If the portfolio mandate limits a factor or country, that broader cap can bind before the sector cap. A rotation model should rank opportunities inside the portfolio's actual risk constraints, not rank first and ask about constraints only after the order is sent.

The rebalance date matters: a monthly ranking observed at month-end cannot be filled at that month's opening price. Use the next available execution point.

Failure conditions and measurement

A strategy should specify the benchmark and report performance after costs, turnover, drawdown, and exposure. A high backtested return without live-eligible constituents and realistic execution is unreliable. Periods without a qualifying sector need a preplanned cash or benchmark fallback. Watch for survivorship bias from using today's sector products in older tests.

Self-check: A 400,000 portfolio caps one sector at 10%. It already owns 25,000 of that sector. The rotation model calls for a 20,000 addition. What can be added? Answer: Cap = 40,000; remaining capacity = 15,000. The remaining 5,000 needs another authorized destination or stays unallocated.

Next, write a sector-rotation rule in six lines—universe, signal, filter, weight, review date, and invalidation—then test it on a fresh period. Use the official objectives to check the expected level of explanation.