Back to lab

Quant Research Lab Help

Model is still in developement, thus results may not be accurate.

Dashboard Forecast

The dashboard uses historical close prices to estimate annualized return and volatility from log returns. It then applies a geometric Brownian motion forecast. The key metrics are analytical, so refreshing the same ticker and inputs gives the same probability, VaR, CVaR, and risk/reward values.

Required values

Ticker and exchange choose the TradingView symbol. Time frame and lookback bars decide how much price history is used. Forecast horizon is the number of trading days projected forward.

How inputs affect results

More lookback bars usually smooth the return and volatility estimate. Higher volatility widens the forecast fan and increases VaR/CVaR. Higher drift raises expected return and probability of gain.

VaR and CVaR

Value at Risk is the loss threshold at a chosen confidence level. At 95%, VaR answers: in normal conditions, what loss should only be exceeded about 5% of the time? CVaR looks beyond that threshold and estimates the average tail loss.

The dashboard uses lognormal price assumptions. The risk studio uses an analytical normal loss model based on portfolio value, annual volatility, horizon, and confidence.

Black-Scholes and Black Model

Black-Scholes prices European options on a spot asset. Black prices options using a forward value, often useful for futures-style underlyings.

Required values

Spot or Forward is the underlying level. Automatic mode uses the dashboard last close. Strike is the option exercise price. Rate is the annual risk-free rate. Vol is annualized volatility. Maturity is years to expiry.

Sensitivities

Delta shows how much option price changes for a small move in the underlying. Vega shows how much price changes for a one percentage point volatility move.

Binomial and Trinomial Trees

Tree models break the option life into steps and work backward from expiry. They can handle American-style exercise, where the option may be exercised before maturity.

More steps generally improves accuracy but takes longer. Higher volatility makes the up/down tree wider. American puts can be more valuable than European puts because early exercise may matter.

Risk Studio

The risk studio combines analytical VaR/CVaR with stress scenarios. Portfolio value scales every result. Annual volatility and horizon control the normal loss distribution. Equity shock and rate shock create direct stress-test P&L estimates.

Strategies

The strategy page estimates trade expectancy. Hit rate is the share of winning trades. Average win and average loss describe payoff size. Slippage reduces the edge. Gross exposure scales both gains and losses. Kelly fraction is a sizing reference, not an instruction to size trades aggressively.

Time Series

ARIMA focuses on autocorrelation and next-step behavior. GARCH focuses on changing volatility. Cointegration is useful when related assets move together over time. Regression estimates a directional relationship or trend reference.

Window controls how many observations are generated for the diagnostic. Signal strength increases the amplitude of moves, which usually raises volatility and can move the z-score further from zero.