Simple Merrill Clock Asset Model

Article BigQuant

Summary

This document outlines a simplified Merrill Clock approach that classifies economic conditions using the direction of year-over-year GDP and CPI changes. It applies moving averages to reduce short-term noise and delays the economic data by one period in backtests to account for publication lags and limit look-ahead bias. The resulting regimes are used to guide allocations among stocks, bonds, commodities, and cash.

The summary reports historical backtests on U.S. and Chinese data. It says the model’s stock and bond signals were generally more accurate, while commodity results were weaker, possibly because commodities differ substantially from one another. Strategies incorporating the model were described as relatively steady and as producing positive absolute and risk-adjusted returns under assumptions that disallow shorting and leverage, but no performance figures or detailed methodology are included here. The authors caution that the model lacks additional leading indicators, captures only incremental macro information, and defines economic regimes differently from conventional economic classifications. Its regime calls should therefore be treated as model-specific, not definitive economic diagnoses.

Key ideas

The model classifies regimes from the direction of year-over-year GDP and CPI changes.
Moving averages are used to smooth economic data, which are delayed by one period in backtests.
The summary reports more reliable stock and bond signals than commodity signals in U.S. and Chinese historical tests.
Strategies using the model were described as steady under no-short-selling and no-leverage assumptions.
The approach omits additional leading indicators, and its regime labels are model-specific.

A Simple Merrill Clock Model Using GDP and CPI Trends

Summary

Key ideas

Tags