Using K-Nearest Neighbors to Match Macro Regimes with Asset Performance

Summary

The document proposes a framework that uses macroeconomic and market variables to find historical periods resembling the current environment, then identifies which asset class, investment style, or strategy performed better in those periods. Candidate inputs include short- and long-term interest rates, rate differentials, equities, foreign exchange, and volatility. The suggested matching method is k-nearest neighbors (KNN), and the author asks for advice on variable selection, quantitative design, and related research.

The only response points to a discussion of time-series matching with dynamic time warping as a related approach. No implementation details, selected features, validation results, or portfolio outcomes are provided. Similarity-based selection would require careful choices about scaling, distance, lookback windows, and how historical performance is measured; the post leaves these open and offers no evidence that KNN improves asset allocation.

Key ideas

• The proposed framework matches current market conditions to similar historical observations.
• Candidate predictors span interest rates, equities, foreign exchange, and volatility.
• K-nearest neighbors is suggested to identify historically successful asset classes or strategies.
• Dynamic time warping is mentioned as a related time-series matching approach.
• The document gives no tested feature set or evidence of investment performance.

Tags

# Analog - Pattern Recognition model using KNN


# Analog - Pattern Recognition model using KNN












I'm building a pattern recognition model for my master thesis. The idea is to build a framework with some Macro variables (long/short term rates; rates differential; equity; fx; vix) in order to find wich asset class (or investment style or strategy) would perform better on the current period, based on similarities with historical data. For that purpose I am using the K-nearest neighbour algorithm. I would like to ask sugestions regarding not only the quantitative method (KNN) but also the most significant macro variables to use.

I also would like to ask if you know any relevant literature regarding this or any similar theme?

Thanks in advance

## Answer by Richi Wa (score 1)

https://quant.stackexchange.com/a/14798

Such an approach is done by the systemic investor blogger in his blog Time Series Matching with Dynamic Time Warping.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.