Seasonality and sector rotation

Lesson 6: Seasonality and sector rotation

The market does not rise evenly. Sectors rotate depending on the interest rate cycle, inflation, economic growth, and the phase of corporate earnings. Seasonality is not calendar magic like “sell in May”; it is a measurable effect of flows, reporting, and institutional allocation. The ability to measure rotation is an edge in your plan, as long as you do not use it as a forecast.

Lesson goal

You will learn to measure monthly seasonality, earnings reporting, and sector rotation on the Warsaw Stock Exchange (GPW) and in the US, to tell the business cycle apart from noise, and to build a sector map with entry and exit thresholds instead of guessing “what will go up now”.

1. Calendar seasonality — facts and myths

We measure seasonality as the average rate of return in a given month or quarter over a long sample.

Seasonality measurement formulas:

Monthly return = (Price_end_of_month - Price_start_of_month) / Price_start_of_month
Seasonal average for month M = AVG (returns of month M over N years, N>=10)
Hit rate = Number of years with a positive return in month M / N
Market-adjusted seasonality = Sector return - Broad index return (WIG or S&P 500)

Measurement in TradingView: use “Seasonality” and compare a sector against SPY or the WIG. Always show the median, not just the average — the average is distorted by crashes.

2. Sector rotation — the economic cycle

Rotation is the flow of capital between sectors in different phases of the cycle.

A simplified four-phase model:

On the GPW the banking sector has a weight above 25% in the WIG20 and is a proxy for NBP rates. In the US the technology sector is above 28% of the S&P 500 — rotation between IT and the rest moves the whole index.

How to measure rotation:

Relative strength of a sector = (Sector index / Broad index) - 3M/6M trend
Rotation breadth: % of stocks in the sector > SMA50
PMI vs sector: 6M correlation between manufacturing PMI and the return of the Industrials sector
Yield curve: 10Y-2Y spread rising = positive for Financials, falling = negative

Table of GPW and US sectors:

3. Earnings season — the most important seasonality for stock picking

80% of lasting relative moves happen around quarterly reports. That is when forward EPS gets revised.

Surprise formula:
EPS Surprise % = (EPS_actual - EPS_consensus) / |EPS_consensus| * 100
Sales Surprise % = (Sales_actual - Sales_consensus) / Sales_consensus * 100
Consensus revision = (Consensus_new - Consensus_old) / |Consensus_old|

Quality of the result beats the size of the surprise. A company beats EPS by 2% thanks to a lower tax rate but cuts guidance — a negative reaction is not an anomaly, it is logic.

4. How to build a seasonality and rotation map into your plan

Step 1 — annual map:

• Mark in your calendar: Fed/FOMC meetings, the ECB, NBP, OPEC, WIG/S&P earnings peak weeks.
• Mark: the report dates of the companies on your watchlist (10-12 companies).

Step 2 — rotation dashboard (weekly):

In a spreadsheet:
Columns: Sector, RS 1M vs WIG/SPX, RS 3M, % > SMA50, % > SMA200, 10Y-2Y spread trend, PMI 3M trend
Rule: if RS 3M < -5% and % >SMA50 <30% — the sector is in relative weakness, do not add positions without a quality filter

Step 3 — decision thresholds, not forecasts:

• “If Financials RS 3M > +3% and the 10Y-2Y spread has been rising for 4 weeks, I check banks in WIG-Banki that have CET1 >15% and C/I <45%”.
• “If Technology RS 3M < -5% and 12M EPS revisions are down more than 3%, I do not open new growth positions without CFO >0 and ROIC >15%”.

When the plan says “do not trade”

• You do not trade a seasonal “Sell in May” thesis as the only reason to sell — the historical deviation is too large, and the transaction and tax cost is real.
• You do not trade sector rotation when you do not understand whether the sector’s move comes from 1 company (for example CD Projekt in WIG-Gry) — WIG-Gry is 70% CD Projekt plus 2 companies.
• You do not trade before FOMC/NBP/RPP (the NBP’s Monetary Policy Council) with leverage on rate-sensitive sectors (banks, real estate, growth tech) — volatility of 2-4% in 15 minutes is outside your control.
• You do not trade earnings season with a position above 1.5% of the portfolio without a defined scenario for the surprise and guidance, if the post-earnings ATR is more than 2x the normal ATR.
• You do not trade a sector map built on 2 years of data — seasonality needs at least 10 years, including at least one recession.
• You do not trade rotation with a leveraged sector ETF as a long-term position — decay and roll costs eat the rotation thesis.

The mistakes that cost the most

1. Confusing correlation with cause in sectors. A rise in oil does not always equal a rise in every refiner — the crack spread and hedging matter more than oil itself.
2. Ignoring the concentration of sector indices on the GPW. WIG-Banki is 3-4 banks, WIG-Energia is 3 companies. Rotation of a “sector” means rotation of 2 companies. Measure the weight.
3. Seasonality without the rate regime. Seasonality from 2010-2021 (0% rates) does not work in a 5% rate regime. Split history by rate regime, not just by calendar.
4. No breadth measurement. The broad index rises, but 70% of the stocks in the sector fall — that is not a sector bull market, only 2-3 leaders.
5. Trading the narrative instead of the revisions. “AI will change everything, so I am buying an AI basket” without checking whether the EPS consensus for the basket is rising — a narrative without revisions is hype.

Measurable exercise

Build a Rotation Dashboard for the coming month.

1. Choose 6 GPW sectors (Banks, Energy, Gaming, Apparel, Industrials, Info) and 6 US sectors (XLF, XLK, XLE, XLI, XLP, XLY).
2. For each one calculate: the 1M, 3M, and 6M return, RS vs WIG / RS vs S&P 500, and the % of stocks > SMA50 (TradingView).
3. Pull: the US and PL 10Y-2Y spread, and the last 3 readings of PL PMI and US ISM.
4. Mark the 2 relatively strongest sectors (highest RS 3M + breadth >60%) and the 2 weakest (lowest RS 3M + breadth <35%).
5. For the strongest and the weakest GPW sector take 3 companies each from your watchlist and check CFO/EBIT and net debt/EBITDA — is the strength or weakness backed by quality?
6. Write down a conclusion: “If next month sector X holds RS >Y, then I will check…” — with a numeric threshold.

Time: 90 minutes a week. Output: 1 dashboard spreadsheet + 2 sentences per sector on quality.

Test yourself

1. Why should you use the median and the hit rate in addition to the average to measure seasonality?
2. What does a widening 10Y-2Y spread mean for the banking sector, and why?
3. How do you tell a sector bull market apart from 1-2 leaders dragging the index up?
4. Why does the interest rate regime change how you interpret seasonality?
5. Give the difference between an EPS surprise and a revision of the 12M forward consensus, and say which one has the more lasting effect.

Frequently asked questions

Does sector rotation work on the GPW the same way as in the US? The mechanics are similar (the cycle, rates, PMIs), but the concentration of the WIG20 distorts the measurement. In the US 500 companies even out a sector, on the GPW 3 companies can be an entire sector — always check the weights.

How often should you change sector exposure in your plan? Less often than you think. Quarterly rotation with monthly rebalancing plus a review after earnings season is enough. Jumping between sectors every week generates costs and behavioral error.

Is seasonality a strategy in itself? No. It is a context filter. Use it as background for quality decisions and risk management (for example, reduce the position before FOMC), not as an entry signal based on the month.

How do you combine seasonality with stock picking? First quality and the balance sheet (ROIC, FCF, debt), then the rotation map (is the sector in a favoured regime), and finally reporting timing (is the earnings window approaching). The reverse order increases selection error.

---