The Earnings Strength Score (ESS) is the single number Earnings Compass uses to rank every name heading into its print. It blends seven underlying components — earnings quality, surprise strength, estimate momentum, market reaction, pre-earnings momentum, catalyst strength, and consistency — into a 0–100 score. When backtested across our Top 100 universe, names that clear the ESS ≥ 70 threshold have produced roughly 2× the baseline 5-day post-earnings return. This guide explains exactly how to read the score, where it shines, where it fails, and how to build a repeatable earnings-trading workflow around it.
What the Earnings Strength Score actually measures
ESS is not a price target and it is not a directional call. It is a probability score: how likely is this name to deliver a positive earnings surprise AND see a positive post-earnings drift, based on everything that is knowable before the print.
The score blends seven components, each weighted by how predictive it has been historically:
- **Earnings Quality (25%)** — 4- and 8-quarter EPS beat rate, consecutive beats/misses
- **Surprise Strength (25%)** — magnitude of recent surprises, weighted toward the latest quarter, plus acceleration
- **Estimate Momentum (10%)** — analyst revision breadth (up vs down) over the last 90 days
- **Market Reaction (15%)** — average 1-day, 5-day and 20-day post-earnings returns
- **Pre-Earnings Momentum (20%)** — 5- and 20-day price drift into the print
- **Catalyst Strength (5%)** — sector momentum, macro risk, news flow
- **Consistency** — surprise standard deviation (low-variance reporters score higher)
A name scoring 75 is not predicted to go up 7.5%. It is telling you: across the last several years, the bucket of stocks that looked like this going into a print outperformed the bucket that did not. That is the entire claim.
The three thresholds that matter
The ESS scale has three practical bands:
| Score | Read | What to do |
|---|---|---|
| 70+ | High-conviction setup | Position with defined risk; this is the bucket the radar surfaces |
| 50–69 | Mixed signal | Trade only if one or two sub-scores are exceptional; size smaller |
| Below 50 | Avoid or fade | The blended history is not in your favour — pass, or look for short setups |
The 70 threshold is not arbitrary. It is the cutoff where backtested 5-day post-earnings returns roughly double the equal-weight baseline on our Top 100 universe. Below 70, the edge compresses fast.
Live scores update overnight, so the radar surfaces the current set every morning. Check our Catalyst Radar for today's ESS ≥ 70 list.
How ESS compares with implied move and analyst dispersion
Most retail traders look at one of three things going into earnings: the implied move (from options), the consensus EPS number, or analyst dispersion. Each is useful and each is incomplete on its own.
- **Implied move** tells you what the options market is pricing — it is a magnitude estimate, not a direction.
- **Consensus EPS** tells you the bar the company has to clear — it ignores whisper numbers and revision trend.
- **Analyst dispersion** tells you how much disagreement exists — high dispersion = bigger move on either side.
ESS rolls these into one score and adds the dimensions options data and consensus cannot see: the company's own surprise history, post-earnings drift pattern, and pre-earnings momentum. Used together, ESS tells you whether to play the print at all; the Implied Move Calculator tells you how to size and structure it.
A repeatable ESS trading workflow
Here is the exact workflow we run on our own book:
1. **Tuesday evening before a heavy week** — open the earnings calendar and filter for ESS ≥ 70. Typically 6–15 names per week. 2. **For each name**, check the four sub-scores that drive the most variance: surprise strength, estimate momentum, pre-earnings momentum, and market reaction. If three of the four are above 70, this is an A-tier setup. Two of four = B-tier. One or zero = skip. 3. **Pull the implied move** for each A-tier name. If implied move ≤ 1.2× the 8-quarter average post-earnings absolute move, the options are relatively cheap — buy premium (long calls, call spreads, or ATM straddles depending on directional confidence). If implied move ≥ 1.5× average — options are expensive, sell premium (credit spreads, iron condors). 4. **Position size** so a two-sigma loss is survivable. ESS gives you an edge, not certainty. The single most common mistake retail traders make is going from 5% positions to 25% positions because the score is high. 5. **Set exits before the print.** Stop loss = below the lower implied move band. Take profit = upper band, or roll a portion if it gaps through. 6. **Hold for the 5-day post-earnings drift window.** The ESS edge does not disappear on the open after the print — it compounds over the following 3–5 sessions. Closing the morning after gives up most of the backtested return.
When ESS gets it wrong
No earnings score is right 100% of the time. ESS underperforms — or actively misleads — in these situations:
- **Major guidance changes** — the score is built from reported numbers and analyst revisions. A surprise capex announcement, a CEO exit, or a fresh M&A headline overrides everything.
- **Macro overrides** — during a Fed week, CPI print, or geopolitical shock, single-stock signals get drowned out by index-level moves. Trade smaller or sit out.
- **Low-float or low-coverage names** — fewer than 8 analysts means estimate momentum is noisy. The score caps confidence in these cases, but treat sub-scores with skepticism.
- **Sector rotations in progress** — if money is exiting tech wholesale, even a 90-ESS software name will struggle on the print. Always check the sector heatmap first.
The honest framing: ESS is a positive expected-value tool, not a free lunch. Some prints will lose. The edge is in repeatedly playing the bucket, not in any single trade.
How ESS is calculated under the hood
For the technically inclined, here is the math at a glance.
Each of the seven components is normalised to a 0–100 sub-score using historical percentile ranks across our coverage universe. The blended score is:
`ESS = 0.25·EQ + 0.25·SS + 0.10·EM + 0.15·MR + 0.20·PM + 0.05·CS + 0·CO`
Where: - EQ = Earnings Quality (beat rate, consecutive beats) - SS = Surprise Strength (weighted recent surprise + acceleration) - EM = Estimate Momentum (up-revisions ratio, net revisions) - MR = Market Reaction (avg post-earnings returns) - PM = Pre-Earnings Momentum (5d + 20d price drift) - CS = Catalyst Strength (news flow, sector momentum) - CO = Consistency (surprise stdev — currently zero-weighted)
Consistency is tracked but currently zero-weighted because the backtest showed it was already proxied by the other components. We re-fit weights every two quarters; the current version is ESS v1.2.
Confidence is reported separately from score. A score of 80 with low confidence (small analyst panel, missing surprise history) is materially weaker than 80 with high confidence — the radar shows both.
ESS vs Earnings Whispers and other earnings signals
Earnings Whispers popularised the idea of a single whisper number — a refined consensus EPS that tries to beat the official Wall Street consensus. That is one input. ESS is broader: surprise history, revision breadth, post-earnings drift pattern, momentum and macro context, blended into one probability score.
If you currently use a whisper number alone, ESS is a strict superset — surprise strength inside ESS captures the same idea, then layers in everything else. If you currently use implied move alone, ESS tells you whether the setup is worth playing before you size it.
We do not claim ESS is the only useful signal. We claim it is the most useful single number we have found, and the backtest is public in our methodology.
Putting it all together
The fastest way to use the score:
1. Open the Catalyst Radar — the ESS ≥ 70 list is the day's short-list. 2. Cross-check the four high-variance sub-scores. Three-of-four above 70 = A-tier. 3. Size with the Implied Move Calculator — cheap options = buy premium, expensive = sell. 4. Hold through the 5-day post-earnings drift window. Do not exit on the open. 5. Track results. The edge is statistical — it shows up across 20+ trades, not the next one.
The score is a tool, not a prediction. Used as a filter on a watchlist and combined with defined-risk option structures, it is the highest expected-value workflow we have built into the platform.