The problems we are brought in to solve.
Returns look acceptable in aggregate but nobody can say which strategy, venue or execution decision produced them, so you cannot double down or cut.
The tested edge disappears once real orders hit real venues. Slippage, partial fills and latency were never modelled.
Exposure sits across brokers, accounts and asset classes with no consolidated position, and reporting is assembled by hand each month.
Named deliverables, not capabilities.
Code-level review against your own historical data: Sharpe, Sortino, information ratio, beta to benchmark, drawdown and turnover, benchmarked properly.
Order routing, intelligent slicing and distributed execution across brokers, with per-fill attribution so cost is visible rather than inferred.
Rebalancing, allocation and hedging logic that runs unattended, integrated with the systems you already operate.
Consolidated position, exposure and P&L reporting with the ratios and weights an investor or committee expects to see.
Korn Capital Management
We took Ryan from backtesting and optimisation through to live deployment, integrating automation with his existing systems and building consolidated reporting across brokers and venues. Latency fell materially and he gained full visibility over live positions and performance.
"Variance are the best in the business. I would not consider looking elsewhere."
Projects like yours.
Drawn from live engagements. Client names are withheld except where a client has agreed to be named.
Consolidated a book held across Interactive Brokers and a Korean broker into one reporting engine: Sharpe, Sortino, information ratio, standard deviation, beta to an MSCI World benchmark, max drawdown, sector and geography weights, gross and net returns and turnover ratio.
Backtesting through optimisation to live deployment for a manager running $500M, integrated with the systems already in place, with latency reduced materially and full visibility over live positions.
Position-level analytics and performance attribution over a live book, built in Python against the manager’s own historical data rather than a vendor dataset.
Took an algorithm already trading, established where the edge had decayed, and rewrote the signal and execution layers around the part that still held.
Listing dates, issuer names and first-day volume assembled into a clean screenable dataset for a systematic strategy.
Where a product is public we show it: CF Trading, Ticker Tape and iDividendStocks all have real screens on their case studies. Client trading systems are a different matter, and for those we will walk you through the architecture, the approach and the outcome on a call.
Read the full case studies.
Backtest through optimisation to live deployment, integrated with systems the manager already ran, with reporting consolidated across venues.
A client’s own indicator set reproduced measure for measure, so research output and live execution produce the same numbers.
Allocator-grade performance analysis across two brokers.
Jess Xu runs an incubator fund with a global technology portfolio on a buy-and-hold strategy, holding 15 to 17 positions at low turnover, with accounts at Interactive Brokers and eBest in Korea. The requirement was reporting an allocator would accept, computed automatically from both accounts.
The specification is reproduced here because it shows the standard the work is held to. This is not a dashboard, it is a defined metric set with a named benchmark attached, computed automatically from both accounts.
Four steps, fixed price before any build.
Free, with an engineer. If there is no defensible edge, or no viable product in it, we say so then.
Architecture, milestones and a fixed price, agreed before any build.
Milestone demos against your own data. Nothing touches capital untested.
A documented codebase your own developers can maintain. Retainer optional.
Common questions.
Yes, on request and before the first call. Client strategies, data and IP stay confidential and are never reused.
You do, outright on completion. That is contractual. We retain no licence over your strategy or system.
Usually, yes. We integrate with what you run rather than asking you to migrate, and the specification names every integration before any build starts.
Tell us what you're working on.
A specialist reads every enquiry, not a sales team. Bring your rules, your current setup and your constraints, and we will tell you what we would build and roughly what it takes.
Get in touchQuants and engineers building trading and investment systems, infrastructure and products.