Uninvestable — and that's the point.

Declined at first filter, on four independent grounds

Any one is sufficient. Sub-scale: £2,000 of capital — minimum tickets exceed the entire book by three orders of magnitude. No track record: three years, self-administered, self-reported, unaudited, no independent NAV — that's a log, not a track record. Fails operational due diligence outright: one person, one machine, manual execution into a retail app, no administrator, no custodian relationship, no auditor, free data sources. Wrong wrapper: it's a personal pension; there is no vehicle to invest into. None of that is a judgement on the strategy — the strategy question never gets asked.

Capacity — the constraint that survives fixing everything else

The UK liquidity floor admits genuinely tiny names. Holding twelve of them, rebalancing monthly, and needing clean exits puts realistic strategy capacity in the low single-digit millions before the screen's tail becomes untradeable — and the design forbids relaxing the very gates that would need relaxing. For most allocators, capacity below ~$100m isn't a small opportunity; it's not an opportunity, because the diligence cost is fixed and the fee pool can't cover it.

The question that has to be answered

Why is this not just a smart-beta ETF with extra steps? Long-only, unlevered, monthly, equal-weight, systematic value-quality-momentum in small caps — that product exists, audited and daily-dealt, at 25–40bp. The system's own chosen benchmark is a small-cap ETF, which concedes the framing honestly. The answer would have to be concentration — twelve names versus a thousand — and the honest reading is that this is a higher-variance expression of the same bet, not a different one. Higher variance isn't edge.

What genuinely survives scrutiny

Alignment is perfect: 100% own capital, no fee, no asymmetric payoff. And the pre-commitment quality is unusually high — frozen parameters with a fixed amendment window, amendments written before their returns are checked, backtests barred as evidence, a stated retirement rule, and pre-registered expectations per market regime for the 2029 review. Most managers' "process discipline" slide is aspirational; this one has a date on it.

The inversion worth noticing

The reason an allocator passes is capacity. Capacity is constrained because the universe is microcap. And microcap is precisely where factor premia have most plausibly survived — because institutions can't fish there. Limits to arbitrage are the mechanism. So the thing that makes this uninvestable is the same thing that makes the underlying premium credible. A £2,000 account has an advantage no fund has, and it is the only advantage it has.

What this lens is actually good for is the evidence standard. An allocator looking at what exists in 2029 would say: self-reported, no independent verification, no attribution, a benchmark-relative result inside one standard error of noise. Every one of those is fixable now, cheaply, except the sample size — reconciling broker contract notes against the printed sheet monthly, and computing simple attribution, is the difference between a record that can be examined and one that can only be asserted.