Track record
Every forecast, written down.
And every result checked, in the open, against the index funds anyone could simply hold.
The forecast engine
Every forecast, written down.
Every result, checked.
Most systems tell you what they think will happen. MARBIN also keeps score. Each night it makes its forecasts, records them, and weeks later marks them against what the market actually did. Signals only keep their influence while their record earns it.
- 1
Forecast
A forecasting model estimates how each company may do against the market.
- 2
Record
Every signal behind a decision goes into a dated ledger.
- 3
Check
After one, three, six and twelve months, each forecast is marked against what actually happened.
- 4
Re-weight
Signals with a working record gain influence; ones that stop working lose it, and the risk matrix is recalibrated.
The model's own record, month by month
62%
of 32 out-of-sample months pointed the right way
Out-of-sample means each month's forecast was made using only data from before it. On average, the fifth of stocks the model liked most beat the fifth it liked least by 2.4% a month.
It is a modest, uneven edge: strong through 2025, close to nothing in 2024. That is exactly why nothing relies on it blindly. Research figures on past data, tested on today's list of companies (which flatters them), not a forecast of future returns.
Earned, not assumed
A signal only gains influence after a long enough record of results, with a margin that isn't luck.
Anchored
Every decision stays anchored to rules tested on years of history. Evidence can tilt the decisions; it can't take them over.
Bounded
The risk assessment adjusts each factor by how well it actually predicted falls — within strict limits.
Measured against the index
The index funds are the bar.
Here is how MARBIN differs.
Holding an index fund is cheap, simple and hard to beat, and we say so. MARBIN isn't built to promise more return. It is built for things an index can't offer, and it measures itself every night, in the open, against the S&P 500 and against the Shariah index funds a Muslim investor could simply buy instead (S&P 500 Shariah, FTSE USA Shariah and MSCI World Islamic). Those are the fairest comparison: the same kind of screened companies, with no one making decisions. MARBIN's own base is such a fund, so the question it answers every night is simple: are the core and the dip adding anything on top of the index?
| Holding the S&P 500 | A Shariah index fund | MARBIN | |
|---|---|---|---|
| What you own | All 500 companies, banks, insurers, brewers and defence contractors included | Companies that pass the index provider's Shariah screen, reviewed on its schedule | Only companies that pass the Shariah screen, re-checked every night |
| When markets fall | You fall with them, all the way | You fall with them, all the way | Stops rest with the broker; new buying halts after losses; smaller positions when the market weakens or fear runs high; long-term holdings stay invested to recover |
| Concentration | A handful of giant technology companies drive much of the index | Often more concentrated still: without banks, a few technology giants weigh even more | Holdings spread across sectors, with limits per sector, currency and exchange |
| Decisions | None: you hold everything, always | None beyond the provider's rules | Written rules for every buy and sell, tested on history before use |
| Learning | None | None | Every forecast recorded and checked; signals keep influence only while they keep working |
| Costs | Very low | Low: a yearly fund fee, a little above a plain index fund's | Higher: trading, currency conversion and UK stamp duty, all counted in our tests |
Whether MARBIN ends up ahead of or behind these funds in any year, it will be for everyone to see against the same benchmarks, dividends included and over the same dates. MARBIN's own record counts real-money accounts only: practice accounts, which trade pretend money, are never added to it. Capital is at risk, and past results, real or tested, are no guide to the future.