No description available

Using dynamic Nelson-Siegel model to fitting and forecasting the yield curve.

A repository to mess around in for my research-focused implementation of the Dynamic Nelson–Siegel (DNS) model using a state-space framework and Kalman filtering to estimate and forecast the yield curve.

We forecast the U.S. Treasury yield curve using the Dynamic Nelson-Siegel model in a state-space framework. A Kalman Filter estimates latent factors, and a GARCH(1,1) extension captures time-varying volatility. The KF-GARCH model improves uncertainty quantification, especially for long-term yields.

We develop an adaptive bond ladder strategy using optimal stopping, yield curve bootstrapping, and the Dynamic Nelson Siegel model with Kalman Filter. By decomposing cash and bond positions, we track value changes and identify optimal reinvestment points using real U.S. Treasury data.

DynamicYield models the U.S. Treasury yield curve using the Dynamic Nelson-Siegel framework with Kalman filtering, extracts latent factors (β₀, β₁, β₂), forecasts the 2s10s spread, & builds macro trading signals. It demonstrates real-world quant research techniques including state-space model, time series forecast &signal driven strategy simulation