Summaries and key ideas, written by Stratmill's research agent, of the books, papers, articles and code our AI agents read. Each page links to its original.
21,023 documents
The document presents a question about a paper on deciding when to sell an asset whose price trend may be increasing or decreasing. The author is trying to understand a change of measure that is intended to simplify the model: the transformed process should…
The discussion considers how to scale a large set of technical indicators when training a model across many stocks with different price levels and feature ranges. Suggested approaches include range scaling, mean-zero standardization, and converting skewed…
The document examines why a QuantLib ImpliedTermStructure appears to change shape after the evaluation date is advanced. The example compares discount factors before and after moving the date by six months. The accepted explanation is that the underlying…
The document raises a calculus question that appears when differentiating the product of an exponential time factor and a Vasicek short rate. It asks how the ordinary time derivative relates to the differential expression and whether multiplying the…
The note examines a pricing argument for a replicable option payoff. If a trading portfolio and the derivative have equal terminal values almost surely, their discounted terminal values also agree, so their conditional expectations under a risk-neutral…
The discussion clarifies how a state price density relates to a probability density under the forward measure in a derivation involving Dupire’s formula. The forward-measure density describes the distribution of the underlying state variables at a future…
The note distinguishes a zero-coupon bond price from the pathwise discount factor used to discount cash flows. In the Vasicek model, the bond price has an analytic expression conditional on the current short rate. The discount factor along a simulated path…
The document presents four regressions for a series and its first and second differences, then asks which integration level is supported. The examples include lagged-level coefficients, standard errors, and Breusch–Godfrey p-values for residual…
The document explains how to turn a sequence of periodic returns into a continuous index or equity curve starting from a base value. It distinguishes adding return increments from compounding them and gives the accepted answer: multiply the index by one plus…
The document poses a question about how transforming two random variables affects second-order stochastic dominance (SOSD). Given that one variable SOSD-dominates another, it asks when applying the same monotonic function to both variables preserves that…
The document asks whether an individual stock’s historical monthly return frequencies can provide a valid expected return estimate. The proposed method groups observed log returns, treats each frequency as an estimated probability, multiplies returns by…
The document outlines the two-step Fama–MacBeth procedure: estimate cross-sectional return regressions each month, then average each coefficient’s time series. It describes the usual standard error as the sample standard deviation of those monthly estimates…
The document explains how to transform a one-dimensional diffusion into a process with constant diffusion coefficient. Applying Itô’s lemma to a function of the state shows that the stochastic term is the derivative of the function multiplied by the original…
The discussion compares two ways to allocate capital between trade situations with different estimated profit factors. Scaling lots in proportion to each factor is presented as a way to maximize average profit factor across a fixed number of trials. With the…
The document addresses an unexpected result in a Sharpe ratio sensitivity analysis: lowering the assumed risk-free rate appears to lower the ratio for both a strategy and its benchmark. The calculation described converts an annual rate into a daily rate,…
The document explains how to interpret a default probability inferred from a credit default swap spread. A quoted CDS spread is observable, while default probability, hazard rate, and recovery are not directly observable. The simple…
The document asks how to model the relationship between signed trading volume and price changes, using a simplified Kyle-style expression in which price impact is proportional to signed volume. It also asks whether least squares is appropriate for estimating…
The document considers how to compute the differential of a positive power of a geometric Brownian motion. It applies the Itô–Doeblin formula to the process expressed as a function of time and Brownian motion, producing a stochastic differential equation for…
The document distinguishes a curve of rates across lending tenors from a forward curve for a single LIBOR tenor. A three-month LIBOR forward curve describes rates for successive future three-month periods; it is not just today’s three-month fixing. Similar…
The discussion addresses how to find the fair spread on a single-currency basis swap built with QuantLib’s floating-for-floating swap instrument. The instrument does not expose a direct method for solving the fair spread, so the suggested approach is to…
The discussion explains why a Fama–MacBeth cross-sectional regression can produce a small or negative estimate of the market risk premium when asset returns are regressed on estimated factor betas. Because those betas contain estimation noise, the…
The document frames company book value, or shareholders’ equity, as a quantity investors may try to forecast and asks for a complete classification of the ways it can change during a reporting period. Its tentative list includes net income or loss,…
The document asks whether White’s Reality Check can compare two Sharpe ratios and presents a simpler alternative. The proposed test estimates each ratio as sample mean return divided by sample standard deviation, rescales each return series by its estimated…
The document explains how dependence among loan defaults affects a portfolio’s loss distribution. If portfolio loss is the sum of individual loan losses, its expected value is the sum of their expected values, so changing correlation alone does not change…