16 articles
What we know about systematic trading.
We write mostly about automated trading systems (AOS) and strategy testing — backtesting, walk-forward, robustness, risk management. And about everything behind it and around it: our own engine, software, GPU and edge clouds, market psychology and the opportunities we watch. No marketing — the workflows and views we actually use.
Walk-Forward and the Million-Dollar Questions
Walk-forward is a re-fitting rhythm, not a rubber stamp. Four questions nobody agrees on — and the boundary past which you stop adjusting and switch the strategy off.
Read →Who's selling to you when you jump on a moving train
When a move is "obvious", it's already late: the biggest edge went to whoever bought when it was boring. Why you supply liquidity in the hype — and why a system has no FOMO. The hardest discipline? Leaving while it's still good.
Read →Don't count strategies, count independent bets
Diversification has a ceiling: each added strategy smooths less, and in a crisis correlations rise toward +1. Where you stop playing against the market and start playing against yourself — and why to count independent bets.
Read →Order execution: what happens to your order from click to fill
Behind a "buy" click is a whole world: order types, the order lifecycle, order-book mechanics, filling in pieces and built-in safeguards — and why it decides whether a backtest tells the truth.
Read →Why Backtests Lie: Fees, Spread, and Fills
Where exactly a backtest lies — fees, spread, slippage, the fill queue — and the deeper trap: your order isn't in the historical data. How to get as close to reality as possible.
Read →From idea to production: how we test trading strategies
The ten phases of our testing protocol: from an idea through the settings landscape, walk-forward and Monte Carlo to the portfolio and zero-delta deployment. A map of the whole journey.
Read →Instrument universes: why today's stock list lies about the past
Survivorship bias in plain words: a backtest on today's index membership only counts the survivors. An honest universe has rules instead of names — and point-in-time membership.
Read →Genetic optimization: searching a million combinations
An exact grid, or evolution: population, generations, mutation and elitism. Why every candidate is a full backtest and why the result is a landscape, not a single winner.
Read →Metrics: how to tell a good strategy from a lucky one
Return, risk and their ratio: profit factor, drawdown, Calmar, Sharpe, Sortino and the pessimistic PROM — what they say, when they lie, and why nothing is judged below 30 trades.
Read →Strategy portfolios: one pool of money
Smooth equity is not tuned — it is composed. Why a portfolio is not the sum of two backtests, how non-correlation cuts drawdowns, and why one shared pool of capital decides everything.
Read →How much to bet: Kelly, risk of ruin and time in the market
The Kelly criterion, fractional betting and an under-watched metric: the share of time in the market. Signals decide whether you make money — sizing decides whether you are still there.
Read →A backtest that behaves like the live market
Event-driven, real-time like: one strategy codebase from testing to the exchange — with order books, fill queues, slippage, fees and latency. Precise, fast, scalable.
Read →Monte Carlo: a distribution instead of a single number
One equity curve is a sample of size one. Hundreds of markets that almost happened reveal a fan of outcomes — and whether your backtest just got lucky with the data.
Read →Black swans, fat tails and Fat Tony
Markets are not normal — extremes decide the fate of an account. What Taleb teaches, why the bell curve lies about markets, and why every strategy here gets grilled by Fat Tony.
Read →Walk-forward analysis: why one backtest is not enough
An optimized strategy knows the answers to the exam it just passed. Walk-forward tests it on material it has never seen — and only that result counts.
Read →Overfitting: why a beautiful backtest often fails live
The prettier the optimized equity curve, the more caution it deserves. How overfitting happens, how to spot it — and how we defend against it.
Read →