A proprietary trading firm that trades its own book. Investors don't hand over an account — they take equity in the company itself and compound alongside a decade of market scars and a refusal to repeat the same mistake twice.
KuroAlpha Ltd is a proprietary systematic trading firm. It trades its own book — capital owned by the company — across major liquid markets via CFD instruments on a regulated retail-prime broker. The book runs as two diversified portfolios with independent mandates, executed at the leverage profile typical of professional retail-CFD operations, targeting 1.5–3% net monthly returns, compounded (≈20–42% annualised), with bounded drawdown.
Investors participate by taking equity in the company, not by handing over a managed account. Their capital becomes the trading book; their return is a pro-rata claim on the company's trading profits and the growth of its enterprise value. Third-party asset management — running outside investors' own accounts under mandate — is a deliberate later phase, opened once a regulated, audited live track record is in place.
The systematic sister firm to an established trading operation co-founded by the principal. Multiple years of infrastructure development. Nine-plus years of active market participation since 2016. The platform is now coming online: live deployment is underway, with the validation pipeline still being hardened and refined as it runs — deliberately, before capital scales.
The thesis is not a clever model. The thesis is: domain expertise is the moat — software is how that expertise scales without proportional cost. What an institutional desk replicates with hundreds of people, this firm replicates from a single mind refined by ten years of being wrong, learning why, and rebuilding.
Built and operated by a single principal — a software engineer and quantitative trader actively trading since 2016. The painful lessons of a decade in the markets are not just remembered; they are systematically embedded into how the system reasons, sizes, gates, and reviews itself.
The trades that taught the most weren't the winners. They were the ones that broke the prior version of the system, exposed a hidden assumption, and forced the rebuild. This is what training data and clever maths cannot replicate.
The founder's compensation is structured entirely through dividends — the same pro-rata claim every shareholder holds. No fixed salary. No expense reimbursement from the trading book. Complete alignment: the founder earns only when the company earns, on identical terms to outside investors.
Anyone with capital can wire software to a broker API. The question is what's inside the system. The internal logic embodies a specific operator's risk discipline, market intuition, and refusal to deploy anything that hasn't survived a hard validation process. The moat is the discipline to make every gate hard, not soft.
A good idea is the input, not the verdict. Multiple stages of review — in-sample, out-of-sample, simulated forward, small live, full live — sit between any candidate and the live book. Skipping a stage is not allowed. At professional retail-CFD leverage, the gap between simulated performance and live performance is where capital dies; that ordering is non-negotiable.
"The secret isn't the software. The secret is a decade of being wrong, learning why,
and refusing to repeat the same mistake twice."
This is what an institution would need to replicate: not the maths, but the operator who knew what to encode, what gates to make hard, and which mistakes never to repeat. That's the part that takes a decade. The software is how that scales.
A boutique quant fund with the same capability surface — a broad strategy library, in-house research and validation pipeline, full live execution, ongoing model maintenance — would carry a permanent fixed-cost base in the millions per year. Senior portfolio manager, quant developers, ML engineers, infrastructure ops, regulatory function, support staff. The fund pays them whether the book makes money that month or not.
KuroAlpha runs from one person. Every basis point of return flows to the company and its shareholders, not to an org chart. This is not a phase to grow out of — it is the design.
The cost asymmetry is roughly two orders of magnitude. A bad month at the boutique still costs six figures of fixed payroll. A bad month here costs the cloud bill — and that bill is borne by the founder, not the trading book. Operational risk sits with the founder. Invested capital is deployed for trading, not for keeping the lights on.
The system shipping today is a snapshot of one operator's knowledge today. It is not the ceiling. Five years ago, none of this existed. Five months from now, parts of what's here will be obsolete — replaced by something better.
Stagnation is the only thing not aspired to. Compounding applies to skill before it applies to capital. The operator's job is to be wrong slightly less often each month, encode that improvement, and let the system propagate it across the entire book within hours instead of years.
What's pitched here is not the final form. It's the current form — the most defensible version of the firm that exists today. Investors who participate at this stage compound alongside the operator, not behind him.
The internal stack is a vertically integrated research-to-deploy pipeline: data ingestion, strategy generation, validation, paper deployment, live execution, and continuous oversight — all under one roof, all owned. There is no third-party platform we are reselling or wrapping. The infrastructure is the product the founder spent five years building.
Implementation specifics are intentionally not detailed in this document. The capability surface is real and verifiable in diligence under NDA. Investor materials are not the right venue to publish the recipe.
For investors who want a deeper technical picture, a separate diligence document is available under NDA. It demonstrates capability without publishing the recipe.
| Metric | 2025 Actual | Target |
|---|---|---|
| Net Monthly Return (avg) | ~2.4% | 1.5–3% |
| Net Annual Return | 32.25% | 20–42% (compounded) |
| Maximum Drawdown | 8% | <15% |
| Sharpe Ratio (est.) | 1.8+ | >1.5 |
| Win Rate | ~62% | >55% |
| Avg Holding Period | ~4 hours | Intraday |
| Months Positive | 10 / 12 | ≥75% |
Status — July 2026: the platform is coming online. Every strategy passes through live-demo validation before touching the company book, and the operator is still hardening gates and refining the pipeline as it runs. The 1.5–3% monthly target is deliberately set below the 2025 realised pace — it is the number the firm intends to defend through refinement and live conditions, not the ceiling it hopes to hit.
The uncomfortable fact of the asset-management industry: the large majority of professionally managed money does not outperform the S&P 500. Across long horizons, roughly nine in ten actively managed large-cap funds underperform the index — before their fees are even counted. Funds, firms, institutional desks, and the retail products built on top of them are, in aggregate, selling underperformance of a benchmark anyone can buy passively.
That is the bar this firm set out to clear — not by a fantasy multiple, but deliberately and repeatably, by roughly one percentage point per month ahead of the index, with a tighter risk envelope than the index itself. Compounded, one point a month is not a rounding error; it is the difference between an index tracker and a top-decile product.
Stated plainly: our sample is small next to a century of index history, and we treat it that way. What the running average demonstrates is not a guarantee of 2.91% — it is that the process is already achieving its intended goal: sitting ahead of the benchmark that most professional money fails to beat, on a risk profile that is actively governed rather than passively endured.
| Risk Profile | KuroAlpha | S&P 500 |
|---|---|---|
| Expected drawdown envelope | 10–15% — anything beyond is an outlier that triggers intervention | -25% (2022), -34% (2020), -57% (2008) |
| Who manages the drawdown | Continuous oversight, hard caps, kill-switches — a bad run is stopped, not endured | No one — the index holds through the full fall |
| Recovery | Bounded losses are recoverable in months | Index drawdowns have taken years to recover |
The products this firm actually competes with for an investor's next pound — and where each one falls short:
| Product | Typical Return | The Catch |
|---|---|---|
| Bank savings accounts | ~4–5% AER at best | Below or barely above inflation; return of capital, not on it |
| Index ETFs | ~10% / yr long-run | Full market drawdown risk — -30% to -57% in bad regimes, no risk manager |
| Active funds & firms | Mostly below the index | ~90% underperform over long horizons, and charge fees for it |
| Loan-backed investments | Up to ~30% / yr advertised | Concentrated credit risk, illiquidity, opacity — the return is the default premium |
| Product schemes (e.g. EV chargers) | Varies | Operational & counterparty risk, often unregulated, no liquid underlying |
| KuroAlpha | 1.5–3% / mo target, 2.91% running | Systematic, risk-governed, liquid underlying instruments, aligned cost base |
Global professionally managed assets exceed $100 trillion. UK households alone hold well over a trillion pounds in cash savings earning bank rates. This firm does not need a significant share of any of it: a fraction of a fraction of that market, run through this system, generates millions in revenue — against running costs that are already defined and borne by the founder, a salary that is performance-based only, and a founder who has staked his own capital to refine the system before asking for anyone else's.
This is a proprietary book, not a managed account — so there is no performance-fee split between manager and client. 100% of net trading profit is retained by the company and accrues to its shareholders pro-rata to their stake. There is no management fee, no carried interest, and no third-party allocator skimming the return. The fee drag that defines the asset-management model is simply absent here, because that model is deferred to a later phase.
Invested capital becomes the trading book. A shareholder's return has two components: a pro-rata claim on annual trading profit (distributable as dividends) and appreciation in the value of the company as the book compounds and the live track record matures. Broker revenue share on executed volume accrues to the company, improving the net.
| How shareholders earn | Mechanism | Basis |
|---|---|---|
| Trading profit | Pro-rata share of net annual P&L on the company book | Ownership % |
| Enterprise value | Share price grows as book NAV and audited track record compound | Ownership % |
| Dividend policy | Distributions from realised profit at the board's discretion; reinvestment compounds the book | Per share |
| Fee leakage | None — no management fee, no performance split, no carry | 0% |
| Operating costs | Borne by the founder personally — not charged to the trading book | Founder |
The founder bears all operational costs — infrastructure, compute, electricity, software, regulatory fees — personally, and takes only dividends on the same terms as every other shareholder. Invested capital is deployed solely for trading. 100% of the book's net return belongs to its owners.
Read this before anything else. KuroAlpha is a proprietary trading firm. It trades its own book — capital owned by the company, its founders and its shareholders. When you invest, you do not hand over money for us to manage on your behalf; you take equity in the company and become one of its owners. The capital being traded is the company's own capital, staked by the people who own the company.
We do not, and will not at this stage, trade external client money. No managed accounts, no pooled client funds, no discretionary mandates over anyone else's assets. That distinction is deliberate: trading your own book as a company does not require FCA authorisation, whereas managing outside client money does. The regulated asset-management business is a later phase that follows an audited live track record — it does not precede it.
The equity round opens once we close out the current financial year. Until then this page is informational only — no subscriptions, enquiries or contact are being taken. The indicative structure and cap table shown below are for context; final pricing and subscription terms are confirmed when the round formally opens.
What will be on offer: ordinary shares in the company (SEIS/EIS eligible), with the founder retaining majority control and every shareholder — founder included — earning on identical, pro-rata terms. Your capital becomes part of the proprietary book; your return is a share of the company's trading profit and the growth of its enterprise value. No management fee, no performance split, no carry.
Registrations open after year-endIn one line: you own a piece of the company, the company trades its own money, and none of it is external client money under management. That is what makes this a prop firm and not a fund — and it is why the invitation to participate is coming soon rather than open today.
The seed round is being raised against a conservative pre-money valuation, with the explicit understanding that future rounds price into a fairer band as the live track record matures. Independent valuation work supports a fair-value range of £4.5M–£6M pre-money on the firm's Y2 earnings potential; the seed entry is offered below that band as a deliberate alignment with first-cheque investors.
| Ownership | Stake | Notes |
|---|---|---|
| Founder (retained) | 51% minimum | Majority control retained permanently |
| Available for investment | Up to 49% | Released in tranches across rounds |
| Seed round (current) | 10–15% | £100k–£150k at £1M pre-money |
| Advisory / Board | 2–3% | Compliance & strategic advisors |
| EMI option pool | 10–15% | Reserved pre-Series A; locked until first hire |
| Future rounds | Remaining | Higher valuations as the book and live track grow |
Three UK Ltd entities, structured for liability isolation, SEIS/EIS eligibility, and clean future fundraising:
| Phase | Period | Capital ($M) | Milestones |
|---|---|---|---|
| Launch | 2026–2027 | $0.5M–$2M | Live deployment, seed equity closed, proprietary book funded |
| Validation | 2027–2028 | $2M–$5M | FCA Appointed Rep, audited track record, multi-broker — own capital only |
| Growth | 2028–2030 | $5M–$18M | Asset management begins — managed accounts open. Asset-class expansion. |
| Scale | 2030–2032 | $18M–$45M | Full FCA authorisation, institutional allocators, external AUM |
| Maturity | 2032–2035 | $45M–$120M | Multi-asset global. Capacity-aware mandates. |
The infrastructure is generic to liquid markets by design. The same stack trading today's instruments trades adjacent ones tomorrow. Every process scales on compute and data, not headcount. The cost asymmetry on the previous page compounds with scale.
| Risk | Detail |
|---|---|
| Key-man | Single-operator firm. Mitigation: the system is documented and versioned. Succession is the operator's responsibility from day one. |
| Market | Leveraged CFDs. Past performance (32.25% in 2025) is not indicative of future results. The firm is designed for drawdown discipline first, return second. |
| Regulatory | Trading the company's own capital does not require FCA authorisation. This raise is an equity offer to qualifying investors (SEIS/EIS). The later asset-management phase — managing outside client money — will require FCA registration, which is planned and not yet in place. |
| Counterparty | The company's trading capital is held in a corporate account at a regulated retail-prime broker. Broker default or restriction is a risk; mitigated by broker selection and, over time, multi-broker diversification. |
| Liquidity | Equity in KuroAlpha is a private, illiquid, high-risk early-stage investment. There is no public market for the shares. Investors should be prepared to hold long-term and to lose the entire amount invested. |
| Technology | System failures possible. Owned infrastructure with redundancy planned. Auto-restart wrappers, kill-switch chain, continuous monitoring. |
| Model | Regime changes may cause underperformance. Continuous monitoring of live-vs-simulated divergence; underperforming mandates are paused, not waited out. |