Is Japanification good or bad? by Peter Chatwell
China’s 30y yield is below 30y JGB; cue many statements about how awful this is.Read on Substack
Macro Advisory With Deep Learning

Is Japanification good or bad? by Peter Chatwell
China’s 30y yield is below 30y JGB; cue many statements about how awful this is.Read on Substack
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
The FTSE factor with the largest residual is Momentum. To find the best stock to buy or sell to position for a correction of that factor, I will use the sorted residuals from the getEquityIndexMembersSortedResiduals action.
Based on the sorted residuals, I recommend selling AutoTrader Group plc (AUTO.L) as it has a positive residual, indicating it is overvalued. Buying GlaxoSmithKline plc (GSK.L) is recommended as it has a negative residual, indicating it is undervalued.
Based on the residuals to the Momentum factor, SGE.L is the most undervalued.
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
latest residual % newsSentiment
EZJ.L -17.20857 0.051292
SGE.L -11.350435 0.139439
BT-A.L 11.170838 0.206277
OCDO.L 9.399289 0.100000
CNA.L 9.047567 0.121313
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
Based on the sorted residuals and news sentiment, I have identified the following stocks:
1. BT-A.L with a residual of 16.66% and news sentiment of 0.1053
2. EZJ.L with a residual of -16.24% and news sentiment of 0.0513
3. STAN.L with a residual of 11.46% and news sentiment of 0.0396
4. SGE.L with a residual of -10.75% and news sentiment of 0.1122
5. CNA.L with a residual of 10.32% and news sentiment of 0.0993
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
User: Please run the equity index model. Negative residuals suggest the stock is cheap, positive residuals suggest the stock is expensive. Investigate the recent news and news sentiment of the top 5 companies shown in the results.
Miura: I will first run the equity index model action to identify the top 5 companies with negative residuals, indicating potential value investments. Then, I will perform web searches and sentiment analysis on the latest news related to these companies to inform the user about their current situation.
Residuals =
EZJ.L -15.433207
STAN.L 13.698465
CNA.L 13.418233
OCDO.L 12.800308
SGE.L -12.632389
…
SSE.L 0.051412
SMT.L 0.047375
SHEL.L 0.03648
HLN.L -0.031766
NG.L -0.001827
…
Based on the latest residuals and news sentiment analysis, the companies with the most promising opportunities for investment are EZJ.L and SGE.L, as they have the most negative residuals (-15.43% and -12.63% respectively). The negative residuals suggest that these companies are currently undervalued.
Let’s take a closer look at the news and sentiment analysis for these companies:
For EZJ.L:
– News: The sentiment analysis shows a positive trend, with the most recent sentiment score being 0.232407.
For SGE.L:
– News: The sentiment analysis shows a mixed trend, with some positive and negative sentiment scores. The most recent positive sentiment score was 0.478571, while the most recent negative sentiment score was -0.056970.
Based on this information, it appears that both EZJ.L and SGE.L could be worth further investigation for potential investment opportunities. However, it’s important to conduct more in-depth analysis and consider other factors, such as financial statements and industry trends, before making a final decision.
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
User: Thoroughly analyse Ocado PLC, ticker OCDO.L, a member of the ftse index, using all of the tools available to you. Tell me whether to go long or short.
Miura: To assist the user, I will first gather relevant information about Ocado PLC using the getHistoricalData action for financial product data and the getEquityIndexFactors action for equity factors. I will then perform a multiple linear regression analysis using these factors to determine if I should recommend going long or short for OCDO.L.
…
Based on the multiple linear regression analysis, the value, size, and yield factors have a significant impact on the OCDO.L stock price. The regression equation is: OCDO.L = …
To make an informed decision on whether to go long or short on OCDO.L, you should consider the current values and future movements of these factors. For instance, if you believe that the value factor is expected to increase, the OCDO.L stock price would also increase. Conversely, a decrease in the size factor and yield factor would lead to a decrease in the stock price.
To summarize, based on the multiple linear regression analysis, an increase in the value factor is expected to lead to an increase in the OCDO.L stock price, while a decrease in the size factor and yield factor would lead to a decrease in the stock price. However, it’s important to consider the current market conditions, economic indicators, and other relevant factors before making a final decision.
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
User: Retrieve the ftse equity index factors, store the results in a variable named ftse_factors. Then perform a PCA on these ftse_factors. Tell me what the results mean.
Miura: To retrieve the FTSE equity index factors, I will use the getEquityIndexFactors action with the index set to “ftse” and the factors set to [“Value”, “Quality”, “Size”, “Vol”, “Yield”, “Momo”]. I will store the results in a variable named “ftse_factors”. After that, I will perform a Principal Component Analysis (PCA) using the calculatePca action with the tickers set to the output of the getEquityIndexFactors action. The results of the PCA will provide insights into the underlying structure of the data and help identify patterns or relationships among the factors.
Data…
forecast previous residual residual% ticker
0 1.105525 1.105308 -0.000217 -0.019624 value
1 1.129802 1.128800 -0.001002 -0.088762 quality
2 1.106535 1.106950 0.000415 0.037516 size
3 1.092824 1.091434 -0.001390 -0.127329 vol
4 1.116901 1.118090 0.001189 0.106302 yield
5 1.199155 1.201612 0.002457 0.204492 momo
…
Miura: The PCA analysis reveals that the six equity factors – value, quality, size, volatility, yield, and momentum (momo) – can be represented as linear combinations of the principal components (PC1, PC2, and PC3). The principal components explain a significant portion of the total variance in the data. In this case, PC1 explains approximately 90.13% of the variance, while PC2 and PC3 explain smaller proportions of 5.88% and 3.23%, respectively.
The factor loadings represent the relationship between each factor and the principal components. A high absolute value indicates a strong relationship between the factor and the principal component. For example, the value factor has a strong negative relationship with PC1, while the size factor has a strong positive relationship with PC3. Understanding these relationships can help in interpreting the results and identifying the underlying structure of the data.
The feature vectors represent the transformed data points in the new coordinate system defined by the principal components. These vectors can be used for further analysis, such as clustering or classification. The PC1:2 neutral weightings provide the optimal weights for each factor to achieve a neutral position in the new coordinate system. These weights can be used as a benchmark for evaluating the performance of investment strategies based on these factors.
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
Based on the information gathered, I can identify several structural factors that could potentially push the neutral interest rate higher in the US economy. These factors include:
1. Low Unemployment Rate: As the unemployment rate decreases, there is less slack in the labor market, which can lead to upward pressure on wages and inflation. This can, in turn, push the neutral interest rate higher. In the regression results, the coefficient for the unemployment rate is positive, indicating a positive relationship between the two.
2. Productivity Growth: Productivity growth is a key driver of economic growth and inflation. If productivity growth is strong, it can lead to higher wages and inflation, which can push the neutral interest rate higher. The regression results show a positive coefficient for productivity growth.
3. Gross Domestic Product (GDP) Growth: GDP growth is another important factor that can influence the neutral interest rate. Strong GDP growth can lead to higher inflationary pressures, which can push the neutral interest rate higher. The regression results indicate a positive relationship between GDP growth and the neutral interest rate.
4. Inflation: Inflation is the rate at which the general level of prices for goods and services is rising. If inflation is high, it can lead to upward pressure on interest rates as the Federal Reserve tries to keep inflation in check. The regression results show a positive relationship between inflation and the neutral interest rate.
5. Leading Economic Indicators: Leading economic indicators are a set of economic variables that can provide an early indication of future economic conditions. If these indicators are strong, it can signal that the economy is growing strongly, which can lead to higher inflationary pressures and a higher neutral interest rate. The regression results indicate a positive relationship between leading economic indicators and the neutral interest rate.
Based on this analysis, I recommend focusing on these structural factors when considering what could move the neutral interest rate higher in the US economy.
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
¶¶¶¶¶
The US economy grew at a slower pace of 1.6% in the first quarter of 2024, according to the latest GDP report. This slowdown was driven by a decrease in business investment and a decline in exports. However, the Personal Consumption Expenditure (PCE) inflation component came in hotter than expected, with a year-over-year increase of 5.3%. This unexpected inflation data has raised concerns among investors and economists. The sentiment towards GDP growth has been mixed, with some positive readings and some negative readings. The sentiment towards inflation expectations has been consistently negative, indicating that investors and analysts are concerned about the potential for higher inflation in the coming months. The sentiment towards the war in the Middle East has been volatile, with some negative readings due to escalating tensions and positive readings due to positive developments. Here is a breakdown of the sentiment scores for each topic:
– GDP growth sentiment: The sentiment towards GDP growth has ranged from -0.32 to 0.50, with an average sentiment score of 0.08.
– Inflation expectations sentiment: The sentiment towards inflation expectations has ranged from -0.075 to 0.50, with an average sentiment score of -0.03.
– War in Middle East sentiment: The sentiment towards the war in the Middle East has ranged from -0.10 to 0.18, with an average sentiment score of 0.05.