// What we do
JKHD develops quantitative trading systems — managing capital provided by Proprietary Trading Firms.
“Quantitative” means rule-based and data-driven. The system makes decisions devoid of gut feeling, emotion, or discretion. Every action — from entry to exit — strictly follows pre-defined mathematical parameters.
A Prop Firm provides qualified traders with capital. Those who comply with their rules trade the firm’s accounts and receive a profit split. JKHD currently manages accounts across seven providers: FundedNext, Topstep, Apex, Lucid, The5ers, TPT, and FTMO.
The aggregated account volume stands at €28,000,000 (individual accounts ranging from $50,000 to $300,000). We trade these accounts through a unified central system. This requires the system to master two core capabilities:
- Identify pattern signals that are statistically significant beyond random chance.
- Enforce every rule of every firm in real time and without exception.
The Project
From the outside, it looks like automated trading software. Behind it lies a four-part system architecture:
Capital Provided by Prop Firms
The partner firm provides the capital, sets the framework, and bears the financial market risk. We supply the execution logic — and bear the operational risk: A rule violation leads to immediate account forfeiture along with all accrued profits. Therefore, rule compliance strictly supersedes profit.
One System, Multiple Accounts
Each firm uses distinct contracts. A trade permitted by Provider A might lead to a ban at Provider B. Consequently, the system evaluates signals independently per account. Across all accounts, a global hedging prohibition applies: Opposing positions on two accounts within the same market are systemically blocked.
Exposure and Leverage
Due to the leverage inherent in futures contracts, the active market volume corresponds to a multiple of the deposited capital. At maximum position deployment, the system controls a trading volume of up to €560,000,000 (over $1 billion USD in order throughput for a complete buy-and-sell cycle).
Because many prop firms match orders internally or execute via proprietary B-book models, this volume is not fully passed through to the live exchange order book. For our execution, this is an advantage: Exceptionally large position sizes generate zero adverse market impact or negative slippage.
Shadow Operations Before Live Deployment
Before a system manages real capital, it undergoes shadow trading. It executes all calculations and logic under live market conditions without capital commitment. Only when every API interface and rule threshold is fully validated does the system go live.
Prop Firm Rules
A rule violation results in immediate account loss — regardless of whether it stems from market volatility, software errors, or human oversight. The system is engineered to eliminate human and technical errors completely.
Loss Limits (Daily & Max Drawdown)
Before placing any order, the system calculates the maximum drawdown risk of the position. If a trade threatens to breach a daily or total loss limit, it is prevented from generating. Trailing drawdowns are calculated in real time at the tick level.
Consistency Rules
For firms with consistency clauses (e.g., maximum single-day gain limited to X% of total profit), the system dynamically throttles position sizing so no single trading day dominates the account.
Parameters & Trading Hours
Maximum contract limits, permitted asset classes, and fine-grained execution windows. The system automatically closes positions ahead of high-impact news events, market closes, or holidays. Differing daylight saving time shifts between the US and Europe are hardcoded.
Restrictions & Prohibited Strategies
Automatic exclusion of latency arbitrage, cross-account hedging, or micro-contract exploitation. On platforms that prohibit automated trading systems, signal generation remains automated while final execution is handled manually.
Payout Logic
The system tracks minimum trading days, buffer thresholds, and withdrawal windows. Prior to submitting a payout request, an automated audit script verifies all eligibility criteria.
Automated Compliance Monitoring
Rulebooks and terms change unannounced. An automated watcher daily scrapes and checks the terms of all partner firms on a structural and textual level. If a change is detected, new trades for that provider are locked until a manual review is confirmed.
Path of an Idea
Every strategy passes through a multi-stage filtering process. The goal is not to force approval, but to discard flawed hypotheses as early as possible.
Hypothesis
Formulation of a market anomaly (e.g., momentum inefficiencies or mean reversion).
Out-of-Sample Validation
Parameter optimization is performed strictly on historical in-sample data. Performance validation is conducted exclusively on untouched out-of-sample datasets.
Statistical Significance
Execution of Monte Carlo simulations and data-mining bias testing. An idea is rejected if its performance is not significantly superior to random execution.
Cost & Execution Modeling
Deduction of realistic commissions, exchange fees, and market-standard slippage.
Shadow Operations
Execution in a live, zero-capital environment to verify processing latencies and system integration.
System Monitoring
Automated heartbeat signals monitor process health. If a heartbeat drops, an external watchdog process intervenes and reverts the system to a safe state.
Fail-Safe & Security
Enforced Risk Limits
If a required risk configuration parameter is missing, the system refuses to launch (Fail-Closed principle).
Controlled Emergency Stop
Activating the emergency stop halts new trade creation. Existing positions are not recklessly dumped into the market, but exited in an orderly manner according to pre-set parameters.
Strict Read-Only Intervention
Manual overrides can reduce or shut down system risk at any time, but can never manually increase risk exposure.
No Inference-Based Decisions
If market data is corrupted, incomplete, or ambiguous, the system does not interpret or guess; it remains in a secure, flat state with zero exposure.
Risk Profile
JKHD’s risk is mathematically determined and strictly capped.
The system knows its maximum risk exposure every millisecond down to the exact cent. Risk management at JKHD does not rely on hope for an exit, but on setting immovable loss limits prior to order placement.
Where human market participants rely on intuition, JKHD operates within hard, pre-computed system boundaries.
Write to us.
One sentence is enough: who you are and what you would like to talk about. E-mails are answered by a person, not a program.
Get in touch kontakt@jkhd.de