When an asset's backtest earns an A but the order book can't support the fills that A implies, CoinRoc caps the grade at B-. That cap is the system doing its job — not penalizing the asset, but being honest about what the market can deliver.
Think of it like...
A road test score that looks perfect — until the examiner checks your car's brakes. The driving was flawless. The score is accurate for the conditions of the test. But if the brakes can't reliably stop the car in real traffic, the license gets a restriction attached.
The grade cap works the same way: the backtest score stands on its own merits, but a thin-liquidity flag attaches a restriction — a ceiling — that reflects a real constraint the backtest didn't test for.
Here's how that maps back: the backtest is the road test, the order book is the brakes, and the B- cap is the license restriction.
How a grade is built — and where the gap lives
CoinRoc's composite grade combines five signals into a weighted score:
Liquidity carries only 5% of the composite weight. That's enough to nudge a score directionally — but not enough to pull a strong backtest below an A grade on its own.
That's why the grade cap exists as a separate, post-composite mechanism. It doesn't change the weights. It imposes a ceiling after the composite is computed — if liquidity is too thin, the final grade cannot exceed B- regardless of how high the other signals scored.
Why a hard ceiling instead of more weight?
Increasing liquidity's weight to 30% would punish thin assets but would also distort grades for assets where liquidity isn't the binding constraint. A ceiling is cleaner: it activates only when execution risk is specifically the problem, and it leaves all other grades untouched.
From raw score to capped grade
The cap triggers when two conditions are both true: the asset's liquidity score falls below 35 (approximately equivalent to under $5M/day global volume), and the composite score would otherwise yield A or B. When both conditions are met, the composite score is capped at 64 — which maps to a grade of B-.
A-
Raw backtest grade
→
B-
Grade after liquidity cap
Cap B- · Thin liquidity
B- cap
FDC-CC+B-BB+A-AA+
Raw composite (A-)After cap (B-)B- ceiling
The raw composite scores at 76/100 — an A- grade. Strong backtest performance, solid risk-adjusted returns, positive sentiment.
The liquidity score is 28/100 — below the 35 threshold. The cap fires. The final grade is clamped to 64/100 (B-).
The asset isn't "bad" — it's correctly rated for what it can actually deliver under real execution conditions.
What you see on the Discovery page
The grade cap is disclosed at two levels: the grade itself renders with a visible Cap B- label, and the asset card carries an amber thin liquidity badge. Neither is a block — the asset remains visible in Discovery. They are disclosures.
Thin liquidity — fills may slip · treat backtest return as an optimistic scenario reference, not a performance guarantee
Grid Return (sim.)
+18.4%
Liq. Score
28 / 100
24h Vol (global)
$2.1M
Grade: B- Raw composite was higher; capped due to thin liquidity.
Cap B- label Signals this is a ceiling, not the raw grade outcome.
Amber badge Plain-English disclosure of execution risk. Always present when thinLiquidity flag fires.
Simulated Grid Return shown The backtest number is still displayed — honestly labeled as simulated.
Design principle: transparency over suppression
Thin-liquidity assets are not hidden from Discovery. They are disclosed. A user with a small position size may find a thin-liquidity asset entirely tradable — removing it silently would be paternalistic. The cap and badge give users the information to decide for themselves.
By the numbers: 32 of 97 assets flagged
After the 2026-06-13 liquidity scoring deployment, the full 97-symbol catalog was scored. The thin flag fired on 32 assets. None of the major high-liquidity assets were wrongly capped.
32
of 97
Assets flagged thin — grade capped at B- or already below
0
major assets
BTC, ETH, SOL, LINK, AVAX — all scored deep liquidity, no cap
$16
worst case
SKY — $16 confirmed global daily volume. Genuinely untradable.
The thin flag is not a surprise — it's concentrated in small-cap and micro-cap assets where thin order books are structurally expected.
Five of the 32 are unmapped tickers (no CoinGecko slug in the registry yet). Their thin flag reflects a data availability gap, not a confirmed verdict — those will be resolved when the registry is updated.
The cap is not a block
B- capped assets are still visible in Discovery. They're not hidden — they're honest.
Key takeaways
The grade cap is a post-composite ceiling, not a weight adjustment. Liquidity carries 5% of the composite weight — but when the liquidity score falls below 35, a hard B- ceiling overrides the raw grade, regardless of how well the other signals scored.
B- is not "bad" — it is "trade smaller and expect imperfect fills." The cap acknowledges a real constraint without removing the asset from consideration. A small position size on a thin-liquidity asset may be entirely viable.
32 of 97 assets are flagged thin. None of the major assets (BTC, ETH, SOL, LINK, AVAX) are capped. The flag is concentrated exactly where it should be: small-cap and micro-cap assets with structurally thin order books.
Simulated / Illustrative — for research discussion only. Not investment advice. Past simulated performance does not indicate future results. All grade outputs, liquidity scores, and return figures referenced in this explainer are derived from simulated backtesting applied to historical data (retail-binance-us fee tier, Year 2 blind forward test, specific grid configuration). Grade outputs are the result of a quantitative methodology applied to simulated data — they are not investment recommendations, suitability assessments, or endorsements of any specific asset. Asset names used in illustrations are for conceptual clarity only. Actual liquidity scores, grades, and flags for specific assets are generated at runtime and may differ from figures shown here. Results will differ materially at other fee tiers, capital sizes, or grid configurations. CoinRoc is a strategy analysis and simulation tool, not a registered investment advisor. Digital assets are highly volatile and speculative. Consult your own financial, legal, and tax advisors before making any investment decision. RES-EXEC-GRADING-01 · 2026-06-13