Map · Shock simulator
Where a shock given to one stock reaches over the days, according to the propagation matrix learned by A · Graph diffusion.
Shock simulator
What it is: A computation run by choosing a stock, a shock size and a number of days, and the list of the most affected stocks.
What data feeds it: /api/analysis/shock: the sparse VAR(1) matrix W in the A · Graph diffusion result of the latest analysis run. Only links that pass the two-period test enter W: the relationship between one stock's residual return today and the other's residual return the next day must have t of 2 or more in the same direction in both halves of the training period. The shock is given as a multiple of the stock's own residual return standard deviation and carried forward with ε(t+1) = W·ε(t). Stocks whose median daily TL trading volume over the last year is in the lowest quarter are marked as thinly traded.
What it shows: Day 0 is the shock itself. For each following day, the number of affected stocks and the most affected ones are listed. The bar shows what fraction of the initial shock the effect is; all days are on the same scale, and a decaying shock gets smaller. The number on the right is the effect on the daily return with market and sector movement removed. At the top, a plain sentence states how many stocks the shock passes to the next day, on which day it dies out, and how much of the next-day returns the model explains.
What can be concluded from it: This is not a forecast but the propagation the model implies. If the model's explanatory power is below 5%, the screen gives a warning: propagation is not reliably visible in prices. If W is largely zero, the shock does not go beyond a few neighbours; this is a direct indication that lagged propagation is weak in this data. The screen states how many possible links were tested, how many passed and how many would be expected by chance; it warns if the number passed is less than twice the expected number. The selected stock's links are listed with the t values of the two periods. In a thinly traded stock, prices form with a delay, so its links may be spurious.
Propagation map
What it is: The fabric map coloured with the shock values on the selected day.
What data feeds it: The result of the shock simulator and the map coordinates of the latest run.
What it shows: It shows which clusters the shock reaches and how it dies out as the days pass. The size of a point is its share of the initial shock; the colour shows the direction.
What can be concluded from it: A shock jumping outside its own sector suggests a lagged link between two different sectors; whether this link is significant should be checked in the lagged relationships table in the Lab > A tab.
The investment information, comments and recommendations given here are not within the scope of investment advisory services. Investment advisory services are provided under an investment advisory agreement to be signed between a client and brokerage houses, portfolio management companies, or banks that do not accept deposits. The signals here are produced from historical data with statistical models, are shown the same to everyone and are not personalised; they may not suit your financial situation or your risk and return preferences. Therefore, making investment decisions based solely on the information given here may not produce results that meet your expectations.