Halal screening you can audit, line by line.
Every other screener hands you a green tick and asks you to trust it. Sepcrium shows the four ratios, the thresholds they were measured against, and the SEC filing each number came from. When our answer disagrees with another screener's, we show you exactly which number caused it.
Free forever for screening. No signup to see a verdict.
source 10-Q · accession 0000000000-26-000000
screened June 2026 · next September 2026
Illustrative layout. Not a verdict on any real security.
Two reputable screeners, the same company, opposite answers.
This is not a bug in anyone's product. The published Shariah standards genuinely disagree about what goes on the bottom of the debt ratio — some divide by market capitalisation, others by total assets. A company can sit on one side of the line under one standard and the other side under another, on the same day, with the same balance sheet.
Almost no consumer app tells you this. You get a verdict, not a basis. So when two apps disagree, you have no way to find out which number moved — and no way to decide which one matches your own understanding.
We publish the basis
Sepcrium screens on a total-assets denominator at 33.333%, consistent with the published FTSE Yasaar and MSCI Islamic criteria. It's stated on every verdict, not buried in a help article.
We publish the arithmetic
Four ratios, each with its numerator, its denominator, its threshold, and the accession number of the filing it was computed from. You can re-derive every figure yourself.
We publish the disagreement
The same company scored both ways — total assets and market capitalisation — side by side, so you can see when the denominator alone decides the verdict.
Four ratios and an activity test. Published in full, applied identically to everything.
| Ratio | Threshold | What it's measuring |
|---|---|---|
| debt / total assets | < 33.333% | interest-bearing borrowing as a share of the business |
| (cash + interest-bearing securities) / total assets | < 33.333% | how much of the company is an interest-earning portfolio |
| (receivables + cash) / total assets | < 50% | how much of the company is monetary rather than real assets |
| non-compliant income / total revenue | < 5% | inclusive of all interest income, not just the obvious lines |
Plus an activity test excluding conventional finance and insurance, alcohol, pork and non-halal food, gambling, tobacco and vaping, adult entertainment, weapons and defence, cinema and music, and hotels.
A ±5 point buffer applies to the three balance-sheet ratios, so a company doesn't flicker in and out of compliance on noise — status only flips after two consecutive quarters outside the band. The screen re-runs in March, June, September and December.
Sometimes the honest answer is "the filing doesn't say."
Segment reporting under ASC 280 follows however management chooses to run the business. Walmart reports "Walmart U.S. / International / Sam's Club" — there is no line in that filing from which anyone can extract a dollar of alcohol revenue. Apple's segments are geographic. Most non-financial filers never separately tag interest income at all.
Every screener faces this. The difference is what they do about it. Assuming the missing number is zero manufactures a pass out of nothing. We'd rather tell you the data isn't there.
All four ratios inside their thresholds, activity test clear, and the underlying figures were actually reported.
A named ratio is over its threshold, or the company's primary business is on the excluded list. We tell you which one.
The filing does not contain what the screen needs. Not a pass, not a fail — a flag that this one needs human review, with the missing item named.
The same balance sheet. Two standards. Two answers.
Here is the mechanism, worked through on a hypothetical company so you can follow the arithmetic. A firm with $100m of debt, $300m of total assets, and a $500m market capitalisation:
Total-assets basis
FTSE Yasaar · MSCI Islamic · Sepcrium
Market-cap basis
AAOIFI SS 21 · Dow Jones Islamic · S&P Shariah
Nothing changed except the denominator. The company didn't borrow a dollar more; the market simply valued it at more than its book. That is the entire reason two honest screeners hand you opposite answers — and it moves with the share price, which means a market-cap verdict can flip on a rally with no change to the business at all.
Two halves of one tool. The screening half is free, permanently.
Halal stock investing
The screen itself, with nothing held back and no signup gate on the verdict.
- Ticker → pass, fail, or insufficient disclosure
- All four ratios as numbers, with thresholds and source filing
- Disagreement view — total assets vs market cap, side by side
- Purification calculators, both per-share and dividend methods
- Zakat calculation on holdings you enter
- Screener with your own criteria, not ours
- CSV export — your data leaves with you
Halal swing trading
The workbench nobody has built: technical tooling that only ever operates over the screened universe.
- Charts and indicators across compliant names only
- A backtester you define and you run
- Statistically honest results — deflated Sharpe, probability of backtest overfitting, purged walk-forward validation
- Compliance-drift warning — flags any backtest that held a position through a change in screening status
Compliance drift is the one feature that needs both halves of this product. A conventional backtester can't see it and a screener has nothing to apply it to.
The screening tools are how you find us and they will never cost anything. The workbench is the only thing anyone pays for. We think neither is worth much without the other, which is why we're saying so up front rather than paywalling the screen once you're attached to it.
What Sepcrium will never do.
Not a promise we might quietly revise. These are product boundaries, and they are the reason we can put a real name on this instead of running it anonymously like every signal service in this niche.
An arithmetic screen doesn't settle the questions people actually argue about.
The published ratio standards say nothing about whether a defence contractor, a company operating in an occupied territory, or a firm with a controversial supply chain belongs in a Muslim's portfolio. Screeners get attacked over exactly these questions, and "we just apply the standard" has never once been an adequate answer.
A named scholar, before we take money
Sepcrium will not sell a subscription until a named Shariah scholar has reviewed and signed off on the methodology, and is identified on this site. Not for legal cover — so that a verdict is a named scholar's position you can weigh, rather than a developer's opinion you can't.
The ethical policy is published, not implied
Our position on boycott, defence, and the activities the quantitative standards don't reach is written down and public. Silence on these is itself a position, and pretending otherwise is how trust gets lost. Read the policy →
One price. Not seven overlapping tiers.
Screening
Verdicts, all four ratios, disagreement view, purification, zakat, custom screens, CSV export. No account needed to see a verdict.
- No card
- No trial that expires
- No paywall added later
Workbench
Everything free, plus charts and indicators over the screened universe, the backtester, honest statistics, and compliance-drift warnings.
- One tier — no add-ons, no seat maths
- Same price on the web and in the app
- Purchasing-power pricing from day one for Indonesia, Pakistan, Bangladesh, Egypt and Nigeria
- Cancel in one click
Not yet on sale. The workbench ships after the scholar review is complete — waitlist members get first access and the launch price.
Fifty-nine percent of the world's Muslims live in the Asia-Pacific region. Charging all of them a New York price is a decision, and most of this industry has made it by default. We're not going to.
Tell us which half you actually want.
This second question is the one that decides what gets built first, so please answer it honestly — including if the answer is "just the verdict, I'd never use charts."