01
Systematic
Rules, not calls. The specification that we research is the specification that trades — no human reaching for the wheel when it gets uncomfortable.
Judgement doesn’t scale. Rules do.
Signal Grove is a systematic trading firm. We write strategies as code, publish the record behind each one, and run that same code in research and in production — so what you evaluate is what actually trades.
Most edges die on contact with reality. The ones that survive get written down, measured, and then left alone.
- Written down A strategy is a specification before it is capital. There is no discretionary overlay on a live book.
- Measured Return, drawdown, and correlation are computed the same way on every page, and the definitions are published.
- Left alone A live strategy is not re-tuned because one month went badly. It either holds its thesis or it is retired.
Doctrine
What we hold to
02
Evidence first
Nothing is claimed that cannot be opened. Every strategy carries its full record, including the flat stretches and the worst drawdown it has ever taken.
03
Transparent risk
Risk is shown before it is taken. Drawdown, correlation, and exposure belong in the argument for a strategy, not in the fine print underneath it.
Research process
How a strategy earns its place
-
01
Mechanism
It begins as a hypothesis we can state in one sentence: who is on the other side of this trade, and why they keep showing up. If we cannot name the mechanism, we do not spend compute on it.
Fails here: a pattern with no explanation -
02
Robustness
It has to survive neighbouring parameters, adjacent instruments, and periods it was not designed on. An edge that only appears at one setting is a description of the past, not a rule about the future.
Fails here: one lucky configuration -
03
Risk
We size it by what it loses, not what it makes: the adverse path within each trade, how its exposure overlaps the rest of the book, and how it behaved in the worst month on record.
Fails here: a tail nobody priced -
04
Implementation
Then it runs on paper against live data, compiled by the same code that will place the orders. Fees, spread, and partial fills all get a vote before anything is called production.
Fails here: the edge dies in execution
Most candidates stop at step two, and that is the point of having the steps. The arithmetic behind each gate — how drawdown, Sharpe, and portfolio caps are actually computed — is written down in the methodology rather than summarised here.
Coverage
One method, four asset classes
A rule that only works on one market is usually a rule about that market’s last few years. Running the same research process across different instrument types is the cheapest robustness test available — and it is what makes a combined book worth composing rather than a single bet with extra steps.
The library started in digital assets and now spans four classes. It will keep growing. The process it has to pass will not change.
US equities
51 strategies
Large- and mid-cap single names, each traded on its own bar structure.
Digital assets
26 strategies
Majors and selected alts, on perpetual and spot venues. Where the library began.
Commodities
7 strategies
Metals and energy exposure taken through listed instruments.
Index & sector ETFs
4 strategies
Broad and sector baskets, used for lower-beta sleeves in a book.
Counts are entries in the public library at this deployment, not a performance claim. Each one has a tearsheet you can open.
Founder
Peter Risch
Founder
FRM
Algorithmic trading
Market making
Peter Risch has spent more than ten years in algorithmic trading and market making.
He built a market-making operation and took it through to exit. Market making teaches a specific kind of discipline: you are always quoting, the risk is continuous, and the only way to survive is to know your exposure before the market tells you.
The work now is quieter. Systematic research across digital assets and equities, and the execution infrastructure that carries that research into production without a person in the loop — which is the same problem as before, on a longer clock.
- Background
- Algorithmic trading and market making, ten-plus years
- Certification
- Financial Risk Manager (FRM)
- Previously
- Founded a market-making operation, through to exit
- Now
- Research, risk, and execution infrastructure at Signal Grove
- Base
- Switzerland
Direction
Where this is going
Today Signal Grove is a research and execution platform: you pick the strategies and compose the book. That is the foundation, not the destination.
Today
A research platform
The library, the composer, and the execution path are live. Subscribers build their own book from production strategies and deploy it, with the same tooling we use internally to watch it.
In development
Managed fund products
The same engine run as real fund products rather than self-directed subscriptions. We are building toward that deliberately: the research standard and the reporting have to be institutional before the structure is worth having.
Nothing on this page is an offer of a fund or a solicitation to invest, and no fund product is available today. We are not publishing launch dates, assets under management, or a client list — we would rather say nothing than say something we cannot evidence.
Start with the record
Every strategy in the library ships with its full history. Read one before you read us.