Research Note / Operational Due Diligence
What a verified quant track record should mean, how different kinds of evidence compare, and a five-question quant fund diligence checklist.
- Mandate focus
- Track record verification and quant fund diligence
- Audience
- Allocators, family offices, diligence committees
A track record is only as good as the evidence behind it. For quantitative funds the gap between a polished backtest and a verified record is wide, and allocators who do not close that gap in diligence close it later with their own capital. This note sets out what a verified quant track record should mean in practice, how different kinds of evidence compare, and how a quant fund diligence process, or a quant fund diligence platform, should structure its questions. It is educational, written for professional investors, and is not an offer or solicitation.
1. How a Record Is Checked, in Order
Qlumina checks a record in a fixed order. The steps below are the same for every program, and none is skipped.
- 01Collect the raw evidence. Broker statements, timestamped trades and backtests, plus audited accounts or independent administrator NAV letters where they exist.
- 02Gate 01: is the edge real?. The record is taken apart: the data, the backtests and the capacity, point by point.
- 03Gate 02: can it be run safely?. Execution, risk systems and custody, checked the way an allocator would.
- 04Gate 03: does it hold up live?. A live pilot allocation, observed before anything scales.
- 05Attach the record to the program. The diligence record stays with the program, and program statistics unlock after investor verification.
The full method, including which records count as evidence, is set out on the quant fund diligence page.
2. What “Verified” Should Mean
“Verified” is used loosely in fund marketing. In diligence it should mean one thing: the numbers can be traced to a source the manager does not control. A return stream that exists only in the manager's own spreadsheet is a claim, not a record.
No single source settles the question. Each kind of evidence shows something and hides something else, so a credible verification combines several of them.
| Evidence source | What it shows | What it does not show |
|---|---|---|
| Backtest | What a model would have done on historical data under stated assumptions. | It is a simulation. It does not show real fills, and it is exposed to overfitting and look-ahead bias. |
| Administrator NAV statements | Official reporting packages produced by a regulated fund administrator. | They describe the vehicle and period reported, not how the strategy would behave in a different account. |
| Direct prime broker statements | Trade confirmations and daily equity runs sourced directly from the clearing broker. | They evidence one account over one period. They say little about capacity or about regimes not yet seen. |
| Timestamped tick-level trade data | Every individual execution, analyzed for slippage, latency and fill veracity. | It tests execution quality. It does not by itself establish that the source of edge persists. |
| Live pilot capital | Real capital deployed and observed before anything scales. | A pilot window is finite, so it confirms execution but cannot prove the strategy across every regime. |
3. Recognized Routes to a Verified Record
Allocators will meet several established routes to a verified record. Each covers something different, and each leaves something out. Providers of attestation and GIPS services describe the first four, for example KPMG's track record verification services and EisnerAmper's overview of GIPS and other options. The summaries here are general descriptions, not legal or accounting advice.
| Route | What it covers | What it does not cover |
|---|---|---|
| Accountant attestation of a record | An independent accountant attests to a historical performance record that is not prepared under the GIPS standards. | Limited to the record and the criteria attested; it is not a view on the strategy. |
| GIPS verification | An independent verifier assesses whether a firm's performance policies and procedures follow the GIPS standards and are applied firmwide. | A review of process across the firm; it does not endorse a strategy or predict results. |
| Examination engagement | An accounting firm gives an opinion, with reasonable assurance, on whether a track record presentation is fairly stated under defined criteria. | Limited to the presentation and the criteria used. |
| Agreed-upon procedures | An accounting firm performs specified procedures and reports its factual findings. | No opinion or conclusion is given; findings are limited to the procedures agreed. |
| Administrator NAV letters | Official net asset value from an independent third party, with assets under management confirmed. | Describes the vehicle and period reported, not behavior at another size. |
| Direct broker statements | Trade confirmations and equity runs sourced from the clearing broker. | One account over one period; says little about capacity or unseen regimes. |
| Live pilot | Real capital deployed and observed before anything scales. | A finite window; it confirms execution but cannot cover every regime. |
Qlumina's review is not an audit and not a GIPS verification. It reads the raw evidence directly and is one more route alongside the others, not a replacement for any of them.
4. Why a Backtest Is Not a Track Record
A backtest answers a different question from a record. It asks what a model would have done, and the answer depends on every choice made along the way: which parameters were kept, which were discarded, and which period was looked at before the rules were fixed. Marketing decks routinely report an attractive Sharpe ratio without disclosing the parameter trial lineage or the provenance of the tick data.
Quantitative research can be guarded against these problems with walk-forward cross-validation, out-of-sample stress testing across different market regimes, and combinatorial purged cross-validation to prevent data snooping and look-ahead bias, as explained in combinatorial purged cross-validation. Two companion notes go deeper: Causal Factor-Absence Placebo Testing and The 20-Year Blind Out-of-Sample Air-Gap. The practical rule is simple: a verified record starts where the simulation ends.
5. A Quant Fund Diligence Checklist in Five Questions
A structured process keeps diligence comparable from one manager to the next. The first three questions mirror the gates Qlumina applies before a program is shown to allocators: is the edge real, can it be run safely, and does it hold up live.
Allocators who want a longer audit list can use the 38-point forensic due diligence framework, which scores a systematic manager across research, risk, custody and governance pillars. For the step-by-step procedure, including the documents to request and how to reconcile them, see how to verify a quant track record.
6. Red Flags Worth Escalating
Several patterns justify slowing down or stopping a process, whichever manager is involved:
- Returns without a source: performance quoted without statements from an administrator or broker that you can request directly.
- Undisclosed search effort: a backtest presented without any account of how many variants were tested to reach it.
- Flattering denominators: returns calculated against capital that the strategy could not actually have deployed once margin is counted.
- No capacity estimate: a strategy described as scalable without a stated ceiling.
- Manager control of assets: any structure in which the manager, rather than a broker or custodian, can move your cash.
7. How Qlumina Applies the Standard
Qlumina, a BVI FSC approved investment manager, admits a program to the Apex Marketplace only after a quantitative review, an operational review and a live pilot allocation. The diligence record stays attached to the program, and program statistics unlock after investor verification. Manager evaluation includes algorithmic verification of timestamped fills, with checks for overfitting, hidden regime bias and tail risk, and portfolio allocations and risk parameters require human fiduciary sign-off.
Read how Qlumina verifies a quant strategy for the full method. Managers who want to understand what is examined can read the portfolio managers page and the diligence FAQ. Allocators can see how evidence, fees and risk are explained in the investor guide, and browse programs through the program catalog or the Apex Marketplace. For the network side of the picture, see how an allocator network for systematic managers works.
Frequently asked questions
What is a verified quant track record?
A record whose figures trace to sources the manager does not control, such as administrator NAV statements, direct prime broker statements and timestamped trade data. A verified record starts where the simulation ends.
Why is a backtest not a track record?
A backtest is a simulation. It does not show real fills, and it is exposed to overfitting and look-ahead bias.
What does a quant fund diligence checklist cover?
Five questions: is the edge real, can it be run safely, does it hold up live, are the fees and terms what they appear to be, and who is the counterparty.
Is Qlumina's review an audit or a GIPS verification?
No. Qlumina's review is not an audit and not a GIPS verification. It reads the raw evidence directly and is one more route to a verified record, alongside accountant attestation, GIPS verification, examination or agreed-upon-procedures engagements, administrator NAV letters, broker statements and a live pilot.
What does a live pilot prove?
It confirms execution with real capital before anything scales. A pilot window is finite, so it cannot prove the strategy across every regime.
How should performance fees be structured?
Performance fees should be earned only on net new profits above a high-water mark, and the fee basis should be confirmed against account statements you can verify yourself.
Trace the Numbers to a Source You Can Ask
A verified quant track record is one whose figures trace to statements and trade data the manager does not control, whose backtests disclose how they were built, and whose strategy has been observed live before capital scales. A diligence process that asks the same structured questions of every manager is what makes programs comparable.
Investing involves substantial risk of loss. The value of investments and the income derived from them can fall as well as rise, and investors may not recover the amount originally invested. Past performance is no guarantee of future returns. This note is for professional investors only and is not financial, legal, tax or investment advice. See the risk disclosure.
Diligence Attached
Every program on the Apex Marketplace carries its diligence record. Professional investors can browse the live programs and request access.
Research notes are published for professional readers. Not an offer or solicitation. Risk disclosure
