Using the Gaussian Mixture Models to Identify Market Regimes

Characterizing the market is an important step in trading system development. Currently, there exist a couple of approaches for identifying market regimes such as using trend and/or volatility filters, machine learning techniques, etc. Reference [1] proposed an approach that uses the Gaussian Mixture Models to identify market regimes by dividing it into clusters.

In statistics, a mixture model is a probabilistic model for representing the presence of subpopulations within an overall population, without requiring that an observed data set should identify the sub-population to which an individual observation belongs. Formally a mixture model corresponds to the mixture distribution that represents the probability distribution of observations in the overall population. However, while problems associated with "mixture distributions" relate to deriving the properties of the overall population from those of the sub-populations, "mixture models" are used to make statistical inferences about the properties of the sub-populations given only observations on the pooled population, without sub-population identity information. Read more

Using the Gaussian Mixture Models, the market was divided into 4 clusters or regimes,

1. Cluster 0: a disbelief momentum before the breakout zone,
2. Cluster 1: a high unpredictability zone or frenzy zone,
3. Cluster 2: a breakout zone,
4. Cluster 3: the low instability or the sideways zone.

As an application, the authors used the regimes to analyze the performance of triple moving average trading strategies,

This research work has demonstrated that conventional Triple simple moving average and Triple exponential moving average trading strategies cannot produce desirable profits throughout all market regimes. As a result of this inefficiency, we identified the best market regime where each of the strategies can be used to achieve better trading portfolio returns.

In short, the triple moving average trading systems did not perform well. However, the authors managed to pinpoint the market regimes where the trading systems performed better, relatively speaking.

We observed the following,

• Using more complex trading systems doesn’t necessarily yield better results. Simpler moving average trading systems can give better risk-adjusted returns.
• It’s interesting to use the Gaussian Mixture Models to divide the market into regimes and analyze the trading systems’ performance. However, the analysis is after the fact. Without developing an efficient mechanism to detect the regime change and incorporate it into a trading system, characterizing the market after the fact is of little use.

References

[1] F. Walugembe, T. Stoica, Evaluating Triple Moving Average Strategy Profitability Under Different Market Regimes, 2021, DOI:10.13140/RG.2.2.36616.96009

Originally Published Here: Using the Gaussian Mixture Models to Identify Market Regimes