Minimum Turbulence Portfolio
Highlights
- Economic shocks cause systemic structural changes in financial markets which are expressed in changes in connections between financial assets.
- Mahalanobis Distance enables us to estimate stock market turbulence.
- I Know First can help to find the most appropriate assets according to the current macroeconomic environment and systemic structural changes in financial markets.
Measuring Financial Turbulence
For more than two years we have been living in a situation of high volatility because of the pandemic, economic lockdowns, interruptions in supply chains (for instance the chip shortage), inflation, and finally, the war situation in Eastern Europe. Economic shocks cause systemic structural changes in financial markets which are expressed in changes in connections between financial assets (in one of our previous articles, we checked hedge and safe-haven assets and identified correlation changes between the S&P500 and TNX in sixty years data frame). Structural changes between financial assets cause asset weights in a portfolio, which were identified as optimal before structural changes. These asset weights may be far away from optimal in a new economic environment. In periods of such events, actuality in a minimum turbulence portfolio increases.
(Figure 1: VIX Weekly Dynamic from July 14, 2014, to August 8, 2022)
The turbulence of a financial asset can be evaluated by estimating unusualness in financial market data. The unusualness of financial data can be estimated by the method of Euclidian distance. This method enables us to estimate the magnitude of an asset return deviation from its usual by calculating the squared difference from its expected return μ to the asset’s return variance in Figure 2. The higher the ratio in Figure 2, the more unusual is the level of return in a period of time in the sense that the instantaneous return variance exceeds its long-term average.
(Figure 2: Euclidian Distance for a Magnitude of Asset Returns)
However, the Euclidian distance does not take into account any dependence (correlation) between financial assets. At the same time, Mahalanobis Distance takes into account information about correlation connections between financial assets. Originally the Mahalanobis distance measure was developed to identify the similarities of human skulls in 1927. In the context of the financial market the Mahalanobis distance, as the measure of financial turbulence, enables us to identify multivariate unusualness in financial market data. For example, the Mahalanobis distance allows us to determine the following unusualness in data returns: let us assume two highly correlated stocks whose returns show a strong positive correlation, but on someday both returns are at the same distance from their means where one return is still above its mean, whereas the other one is now below its mean. In other words, the squared Mahalanobis distance takes into account unusualness in correlation between financial assets.
(Figure 3: Mahalanobis Distance for Estimation of Financial Turbulence)
In the context of portfolio management, we must take into account asset weights in an investment portfolio.
(Figure 4: Portfolio Financial Turbulence)
The Mahalanobis distance for a portfolio should react less strongly if assets which show unusual behavior have only small weights in the portfolio. Figure 4 presents the equation for portfolio turbulence taking into account asset weights in a portfolio.
Financial Turbulence of the US Stock Market Sectors
Let us estimate the financial turbulence of a portfolio which includes different sectors of the US economy and the bond market.
| Sector ETF | Description |
| XLF | XLF tracks an index of S&P 500 financial stocks, weighted by market cap. |
| XLI | XLI tracks a market-cap-weighted index of industrial-sector stocks drawn from the S&P 500. |
| XLK | XLK tracks an index of S&P 500 technology stocks. |
| XLV | XLV tracks health care stocks from within the S&P 500 Index, weighted by market cap. |
| XLB | XLB tracks a market-cap-weighted index of US basic materials companies. The fund includes only the materials components of the S&P 500. |
| XLP | XLP tracks a market-cap-weighted index of consumer-staples stocks drawn from the S&P 500. |
| XLU | XLU tracks a market-cap-weighted index of US utilities stocks drawn exclusively from the S&P 500. |
| XLE | XLE tracks a market-cap-weighted index of US energy companies in the S&P 500. |
| XLY | XLY tracks a market-cap-weighted index of consumer-discretionary stocks drawn from the S&P 500. |
| XLRE | XLRE tracks a market-cap-weighted index of REITs and real estate stocks, excluding mortgage REITs, from the S&P 500. |
| IEF | IEF tracks a market-value-weighted index of debt issued by the US Treasury with 7-10 years to maturity remaining. Treasury STRIPS are excluded. |
| IGIB | IGIB tracks a market-value-weighted index of USD-denominated, investment grade corporate debt with maturities between 5-10 years. |
(Table 1: Sector ETFs and Bond ETFs Descriptions)
In Figure 5, we can notice the Financial Turbulence of the Equal-Weighted ETF Sectors Portfolio. The portfolio financial turbulence is estimated based on daily logarithmic stock returns of close prices provided by Yahoo Finance for the period from August 4, 2017, to August 5, 2022 (5 years). Overall, we can estimate the financial turbulence of the Equal-Weighted ETF Sectors Portfolio by calculating the daily mean which is equal to 90.73%.
Table 2 presents estimations of financial turbulences for three portfolios.
| ETF | Equal-Weighted ETF Sectors Portfolio | Minimum Turbulence ETF Sectors Portfolio | Minimum Turbulence ETF Sectors and ETF Bonds Portfolio |
| XLF | 10% | 9.67% | 2.66% |
| XLI | 10% | 9.37% | 1.60% |
| XLK | 10% | 9.04% | 1.30% |
| XLV | 10% | 9.44% | 0.70% |
| XLB | 10% | 9.21% | 1.12% |
| XLP | 10% | 10.71% | 0.78% |
| XLU | 10% | 12.41% | 1.18% |
| XLE | 10% | 9.85% | 1.36% |
| XLY | 10% | 9.15% | 1.44% |
| XLRE | 10% | 11.14% | 1.36% |
| IEF | X | X | 28.28% |
| IGIB | X | X | 58.23% |
| Portfolio Turbulence | 90.73% | 90.53% | 68.91% |
According to Table 2, we can notice that the Equal-Weighted ETF sectors portfolio provides a reasonable level of financial turbulence of 90.73% which can be decreased to 90.53%. At the same time, including government and corporate bonds in the portfolio allows us to significantly decrease the portfolio financial turbulence to 68.91%.
Stock Picking with I Know First in the Turbulence Market
The concept of portfolio turbulence allows us to identify an optimal portfolio from the risk management point of view in a period of macroeconomic changes. However, a period of market turbulence is also a period of long-term investment opportunities and it is where I Know First can help to find the most appropriate assets according to the current macroeconomic environment and systemic structural changes in financial markets.
I Know First is one leading company that has been effectively using machine learning and AI-based algorithms to provide daily forecasts and facilitate trading for over 10,500 financial instruments. More importantly, I Know First’s algorithm can fulfill the idea of discovering “fractals” and patterns using a more accurate way through AI and machine learning without involving any human judgments. The algorithms can present historical price patterns based on the data inputs, testing the performance on years of market data, and validating them on the most recent data to prevent overfitting. If an input does not improve the model, it is “rejected”, and another input can be submitted. I Know First provides different forecast packages based on the AI algorithm which allows us to select the most promising stocks (you can access them here). For example, below you can see the investment result of our ETF package which was recommended to our clients on July 20, 2022.
Package Name: ETF’s Forecast
Recommended Positions: Long
Forecast Length: 14 Days (7/20/22 – 8/3/22)
I Know First Average: 10.19%
During the 14 Days forecasted period several picks in the ETFs Forecast Package saw significant returns. The algorithm had correctly predicted 10 out of 10 returns. The highest trade return came from TQQQ, at 25.23%. TAN and SSO had notable returns of 12.62% and 11.3%. The package’s overall average return was 10.19%, providing investors with a 4.64% premium over the S&P 500’s return of 5.55% during the same period.
Conclusion
We have been living in a period of high uncertainty and macroeconomic changes since 2020. The method of Mahalanobis Distance enables us to estimate stock market turbulence and construct an optimal portfolio. At the same time, a period of market turbulence is also a period of long-term investment opportunities and it is where I Know First can help to find the most appropriate assets according to the current macroeconomic environment and systemic structural changes in financial markets.
To subscribe today click here.
Please note-for trading decisions and use the most recent forecast.
