Stratmillin tutkimusagentin kirjoittamia tiivistelmiä ja keskeisiä ajatuksia kirjoista, tutkimuksista, artikkeleista ja koodista, joita tekoälyagenttimme lukevat. Jokaisella sivulla on linkki alkuperäislähteeseen.
246 dokumenttia
The document explains a C++ implementation of the Thomas algorithm, a specialized form of Gaussian elimination for solving tridiagonal linear systems. It describes the diagonal and off-diagonal coefficient vectors, the forward sweep that computes modified…
The article demonstrates a machine-learning use of path signatures: classify handwritten digits by treating pen coordinates recorded over time as a path. It describes training a signature-based model on the UCI Pen-Based Recognition dataset, using ordered…
The article introduces the Johansen procedure for testing cointegration among multiple time series and estimating stationary linear combinations. It describes expressing a vector autoregressive model as a vector error correction model, then using the rank of…
The article explains how discrete Asian options use sampled prices along an asset path to determine their payoff. It distinguishes arithmetic averaging from geometric averaging and models price paths with geometric Brownian motion. Monte Carlo pricing…
The article explains how the analytic Black–Scholes formulas for European vanilla calls and puts can be translated into a procedural C++ implementation. It defines the underlying price, strike, interest rate, volatility, and time to maturity, then uses the…
The document derives an explicit finite-difference scheme for the one-dimensional heat equation. It approximates the time derivative with a forward difference and the spatial second derivative with a centered difference, then advances the solution using…
The article compares the typical financial engineering master’s curriculum with the capabilities quantitative funds seek when hiring. It lists common coursework such as derivatives pricing, numerical methods, portfolio optimization, risk, programming, and…
The document explains how C++ function parameters behave when passed by value, by reference, or by const reference. Passing a large object such as a vector by value creates a copy, which can consume memory and processing time; passing by reference avoids…
This tutorial introduces the Ornstein-Uhlenbeck (OU) process as a continuous-time model for a variable that fluctuates around a long-term mean. It explains the roles of the mean, reversion speed, and volatility, and contrasts mean-reverting behavior with…
This career guide compares quantitative researcher and quantitative developer roles for engineers considering a move into quantitative finance. Researchers need evidence of rigorous analysis and stronger statistical skills, including time series methods and…
This article explains why tactical asset allocation strategies can be difficult to evaluate over long periods: allocation signals are often monthly, leaving relatively few observations, and market regimes may persist for years. Retail investors may also lack…
This tutorial describes a MySQL and Python data store for daily equities history. Its proposed schema separates exchanges, data vendors, symbols, and daily prices, linking prices to both a security and its source. That structure supports multiple vendors,…
This article surveys five books on finite difference methods (FDM) and partial differential equations used in quantitative finance, especially for solving Black–Scholes pricing problems. It distinguishes texts that emphasize mathematical foundations,…
The document extends binomial-tree option pricing from a small tree to a finite N-step model. It explains backward propagation from known terminal payoffs and presents risk-neutral valuation as an alternative: calculate the probabilities of ending at each…
The article explains how to evaluate Python frameworks for systematic strategy research and how backtesting fits between strategy development and live deployment. It distinguishes historical performance testing from trade simulation and real-time order…
This tutorial explains the Thomas algorithm, also called the tridiagonal matrix algorithm, for solving the banded linear systems produced by an implicit finite-difference method. It frames the algorithm as Gaussian elimination adapted to a matrix with…
This tutorial presents the Crank-Nicolson method for numerically solving the one-dimensional heat equation. It motivates the method by noting that an explicit finite-difference scheme can impose a restrictive time step. Crank-Nicolson averages the spatial…
This development diary describes changes to a forex trading system: correcting how positions use bid and ask prices, adding historical tick data from CSV files, and building an initial event-driven backtester. The position update distinguishes trade…
This note extends the one-step binomial option model from zero interest rates to a positive continuously compounded risk-free rate. It bounds the stock’s possible up and down prices around risk-free growth, then chooses a risk-neutral probability that makes…
The article describes a long-only equity strategy that uses timestamped vendor sentiment scores as trading events in QSTrader. It enters a stock when its sentiment reaches the positive threshold of +6 and exits when the score falls to -1. Three versions…
The article explains why raw pointers and the legacy C++ auto_ptr create problems when stored in standard library containers. Raw pointers require explicit cleanup, so exceptions or early returns can leak allocated objects or leave dangling pointers.…
The article walks through installing QSTrader on secondary Raspberry Pi nodes managed by SLURM, then checking the setup by running its 60/40 equity and bond example across the cluster. It describes installing Python build dependencies, creating a virtual…
The document explains decision trees for regression and classification, describing them as models that divide feature space into rectangular regions using axis-aligned splits. For regression, each region predicts the mean response of its training…
This installment in an event-driven backtesting series describes a portfolio component that receives trading signals, creates orders, processes fills, and records positions and holdings over time. Its NaivePortfolio tracks per-symbol quantities alongside…