StatGuardia is a trading workstation and research lab built on mathematics, probability and statistical analysis. Specialist engines read the same market from different angles. A single guard decides what is allowed to act. Every call is recorded, so it can be checked against what actually happened.
Markets do not reward a single clever formula. StatGuardia is built as a coordinated desk. Readers notice. Interpreters explain. Fusion compares. The guard decides. The record remembers.
No engine is trusted on its own, and no signal is allowed to become the whole decision.
30
engines in one workspace
5
questions asked of every market
1
final authority
Thirty engines, five questions, one guard. Illustration.
The desk
Five questions, asked of every market, every time.
01
Read
What is moving, and where is pressure building?
Market readers watch price, flow, activity and key levels as they happen, and point attention where it has been earned.
02
Interpret
What kind of structure may be forming?
Setup interpreters name the shape of the market: a trap, a squeeze, a pattern, a familiar shape from history.
03
Fuse
Does the evidence agree?
Fusion layers weigh every view against the others, so one loud signal never becomes the whole decision. Agreement is measured, not assumed.
04
Guard
Is action allowed, and how much?
Capital protection is the final authority. It admits, sizes down, protects or blocks. No signal, alert or agent can go around it.
05
Record
What actually happened?
Every call, every block and every outcome is written down, replayed and compared, so the desk learns from evidence instead of memory.
How it works
From signal to record.
Connect
The desk checks its data and its connections before it reads anything.
See
Prices, levels, flow and context arrive in one shared picture.
Interpret
The engines read that same picture, each from its own angle.
Fuse
Agreement and disagreement are weighed, not counted.
Guard
Admit, size down, protect, or block.
Act or stand aside
The operator decides. Doing nothing is a valid outcome.
Record and replay
The result is written down. Changes are replayed and held to data they were not built on.
Principles
The rules it reads by.
Heat is attention, not permission.
Unusual activity earns a look. It does not earn a trade.
Confidence is evidence, not certainty.
A confidence value describes a model's output. It still has to survive freshness, calibration and review.
Confluence is agreement, not truth.
Several engines can agree for the wrong reason. Outcomes decide who was right.
The guard is the final authority.
No recommendation, alert or pattern bypasses capital protection.
If it cannot explain itself, it is not finished.
What the system saw, why it alerted and why it blocked are part of the product.
Replay is research, not a guarantee.
A good backtest compares ideas. It does not promise tomorrow.
Method
Trust is earned on data it has never seen.
Search long enough and randomness will hand you a masterpiece. So nothing is promoted on the strength of its backtest. Every change is replayed, benchmarked against the current desk, and judged on data it was not built on.
The experiment here is live. Prices are fair coin flips, so no rule can have a real edge. A thousand simple rules are tested on the first ten years, and the luckiest is shown ten more years it has never seen.
We publish our process, not our edge.
SimulationCoin-flip prices
Seen, ten years+0
Never seen, ten years+0
Profit in ticks of the best of 1,000 random five-day rules. On fair coin flips every rule's true edge is zero, and ten years of chance alone moves a rule by about 50 ticks either way. Each run draws new coin flips in your browser. Hypothetical and illustrative.
Foundations
Built on first principles.
The classical tools of the trade are the shared vocabulary everything here is built from. The art lies not in knowing them, but in knowing when each is the right one to reach for.
The normal distribution organizes mass symmetrically around a mean μ, with spread controlled by a single parameter σ. About 68% of the mass lives within one σ of μ, about 95% within two, and about 99.7% within three: the canonical 68/95/99.7 rule.
It is the most-invoked, most-misapplied distribution in all of quantitative work. Knowing when not to reach for it is as important as knowing the formula.
The posterior probability of a hypothesis θ after seeing data D is the prior belief about θ, multiplied by how well θ explains D, normalized by how likely D was overall.
It is not really an equation. It is an instruction for how a rational mind should respond to evidence. Every statistical model is, at heart, a wrapper around this rule.
The asset price S evolves in continuous time with a drift term (μ S dt) and a random shock proportional to the price itself (σ S dW). The multiplicative form is what keeps prices positive and what makes log-returns tractable.
Nearly every closed-form pricing result in quantitative finance stands on this equation, as do all the caveats about what it quietly fails to capture.
Sharpe ratioS=\frac{\mathbb{E}[R-R_f]}{\sigma_R}
Excess return per unit of standard deviation. A scalar that lets you compare strategies with wildly different scales on the same axis: how much reward did you earn for each unit of risk you took on?
Powerful because it is simple; dangerous for exactly the same reason. Sharpe ratios computed on short windows or skewed distributions can mislead badly.
Shannon entropyH(X)=-\sum_i p_i\log p_i
The expected information content of a random variable, measured in bits when the log is base 2 and nats when base e. Low entropy means predictable; high entropy means the world is allowed to surprise you.
Entropy shows up everywhere: decision trees, model compression, market microstructure, risk concentration. Whenever you need to quantify how much you don't know, reach here first.
Black-Scholes callC=S_0N(d_1)-Ke^{-rT}N(d_2)
Under a specific set of assumptions (continuous trading, constant volatility, log-normal returns, no arbitrage), the fair price of a European call is the expected discounted payoff, which collapses to this closed form.
The formula is less interesting than the cultural achievement it represents: the moment pricing became something a human could write down, critique, and build upon, rather than negotiate.
The smallest loss ℓ such that losses exceed ℓ with probability no greater than (1 − α). A summary statistic, not a risk measure in the modern sense: VaR tells you where the tail begins, but says nothing about how far it extends.
We use it, but always alongside complementary measures: Expected Shortfall, scenario analysis, stressed sensitivities. Never in isolation.
Given a prior estimate of hidden state x, a new noisy measurement z, and a model of how measurements relate to state (H), the optimal updated estimate is a weighted blend, where the weight K (the Kalman gain) depends on how noisy the measurement is versus how uncertain the prior was.
It is the mathematical formalization of a core epistemic virtue: trust evidence in proportion to its quality, and revise prior beliefs in proportion to how poorly they explained what you observed.
Contact
Start a conversation.
For research, advisory and engineering inquiries. We reply by correspondence.
StatGuardia provides statistical modeling, quantitative research, and software engineering services. StatGuardia is not a registered investment adviser, broker-dealer, or financial planner, and nothing on this website constitutes investment, legal, tax, or accounting advice or a solicitation, recommendation, or offer to buy or sell any security or financial instrument.
All content is provided for informational and educational purposes only. Any illustrations, simulations, or sample outputs shown are stylized and hypothetical; they do not represent actual trading results or any specific strategy. Past performance is not indicative of future results. Statistical models carry inherent limitations and all quantitative estimates are subject to model, data, and estimation error.
Visitors should consult qualified licensed professionals before making any financial, legal, or tax decision. Use of this website and any correspondence is subject to the Terms of Use and Privacy Policy.