MRI Are Important in Avoiding Market Losses

---

The Market Resilience Indexes® (MRI) are relevant to avoiding market losses. They are measures of the return acceleration of an index and better reflect the emotional stance of investors compared to price changes. Understanding the physics-based drivers of investor sentiment provides insight into the possible path for the MRI and makes it easier to implement our loss-avoidance algorithms. I believe this knowledge will also benefit a range of professional investors using different strategies.

In early 2008, while working at Nikko Asset Management in New York City as a Global Tactical Asset Allocation fund manager, I developed algorithms designed to avoiding stock market losses. These algorithms are designed to trade at the inflection points of our MRI, rather than at the inflection points of the stock index. The algorithms also consider measures of price trend. 

We tested our techniques across various stock indexes, starting with the Russell 1000, due to my involvement in developing the Russell/NRI indexes for Japan. We later expanded our tests to include the S&P and MSCI indexes. Among the indexes we tested, the Dow Jones Industrial Average (DJIA) generated the most reliable signals, likely because of its long history of consistent decision rules that result in relatively stable stock weights over time. 

The figure above shows the performance of the loss-avoidance signals for the DJIA from 1918 through Sep 27, 2024. The DJIA price level is represented by the brown line, while the initial traded series, depicted by the green line, reflects the performance of being either long or short based on the MRI inflection points. Our algorithms were finalized in 2008, and the real-time performance pattern of the signals is displayed.

Since 2014, CPM has provided these signals to individual investors for use in their investment strategies. On average, the initial traded series generated 6.5 trades per year from 1918 to 2024. However, the trades are not evenly spaced throughout the year; instead, they cluster around major market inflection points, sometimes resulting in weekly trades for several consecutive weeks. This clustering of trades increases the stress one feels during market inflection points. 

To address the clustering issue, we adjusted the algorithms by introducing a one-week delay, executing a trade only if the signal was not reversed in the following week. The performance of this modified algorithm is represented by the orange line. As expected, the modified algorithm resulted in fewer trades and slightly lower returns, but the performance pattern remains attractive.

The table below shows the performance of both the original and modified (low trading) algorithms.

While the resilience cycles vary considerably from year to year, we found that the number of cycles per ten-year period has remained surprisingly consistent over the past 100+ years. On average, there are about two Micro MRI cycles and roughly 0.25 Macro MRI cycles annually since 1918. Many are familiar with the 6-month and 4-year cycles, but we observed how these varied from year to year through the MRI. 

We sought to identify the forces behind this long-term stability, while also explaining the year-to-year differences. The outcome of this research is the physics-based sentiment drivers.

The MRI Applied to Bonds  

The MRI are also relevant to other major markets, as demonstrated by the signals for the U.S. Treasury Futures contract, "TY1."  Algorithms for this index were finalized in 2008 as well. The figure below covers from 1983 through Sep 27, 2029.

We have not attempted to reduce the level of trading for this set of algorithms but instead use the performance pattern as validation for considering the MRI. Our Treasury bond decisions are based directly on the MRI for TY1.

We have applied the MRI and related algorithms to other indexes. We find the techniques work best on indexes for major markets like the US stock, bond, and commodities markets.