New Product Launch: What-If Scenario Response Modeling, a New Component of the Algorithm


This article was written by Esther Hanon, a Financial Analyst at I Know First.

I Know First Introduces New Product: a What-If Analysis Scenario Model


  • What-If Analysis provides a new means to which I Know First can find new ways to take capabilities of AI and expand on it.
  • To this end, the “What-If Scenarios” tool was intended for institutional investors
  • How the “What-If” tool utilizes AI
  • Method of gauging market sensitivity to certain events and measuring their responses.

What Is the “What If Scenario” of I Know First?

Our latest forecasting product is the “What If Scenarios”, a tool to simulate different market conditions. These scenarios are based on our current forecasting algorithms and endeavor to see beyond tomorrow’s actions of major market players. They provide macroeconomic perspectives to monetary policies and market fluctuations. For example, say tomorrow oil prices go up by 10% — what will be the market forecast in this case? It is just one example of the immense possibilities we can exploit to make investment decisions. How does it work? Interest rates, oil prices, and currency rates are in the hands of a few major market influencers, such as governments and oil producers. Assuming different possible actions are possible by these big players, we can forecast the outcome of their actions. Additionally, the market sentiment can change due to some unexpected event, which may affect the major indices. Our algorithms will assess the current global events and their effects on the market. This will allow you to know which market will react, how strongly it will react and in what direction. The answer to such eventualities is not obvious. It all depends on the circumstances, on current market conditions, and what precedes them.

Thus, we will also measure the market sensitivity to such moves, and even assign a probability, for instance, of anticipated Fed or ECB actions, and what the market reaction will be, based on these simulations. This market sensitivity, a measure of global systemic risk, should concern every trader and policy maker and is a subject of our current research. I Know First’s latest forecasting product is the “What If Scenarios”, a tool to simulate different market conditions. These scenarios are based on our current forecasting algorithms and endeavor to see beyond tomorrow’s actions of major market players. They provide macroeconomic perspectives to monetary policies and market fluctuations. Market response to any event depends on the circumstances, on current market conditions, and what preceded them. This service is meant for institutional investors who desire more detailed scenarios for their investment plan though it can also be adapted to other needs. Read more, here.

Image result for what if scenario

We are constantly striving to improve the services our company provides by looking for new ways in which we can take the capabilities of AI increasingly further in order to best approach the stock market. Evidently, no one can quite look into the future, but we are persistent in making this process as accurate as possible using our current AI technology to account for an entire host of possible factors such as interest rates, oil prices, currency rates, market movers (governments & big corporations), global events, supply shocks, etc. We use our AI algorithms to assess all of these different ongoing movements and come to a few strong conclusions about which market will react, how strongly it will react, and in what direction it will react. Part of being able to do this is developing and measuring a metric we call market sensitivity, which measures how likely is X to happen and how will Y market react. This is a metric that has a pulse on global systemic risk and is one we are conducting in-depth research in. In the face of all this advancement and evolution in how we are using AI, we want to reiterate what makes us so confident in AI in the first place and why we believe it is the way forward for financial investment.

What-If Analysis Highlights the Top 10 Contributors to Market Changes – AI is Not a Black Box

Although AI results in complex quantitative formulas that are not easily explained in plain English, we can still derive very tangible results from the system. For example, what is affecting the market, the weight, or importance of each event, and the top 20 contributors to the forecast. We want to reaffirm that AI is a complex algorithm, but it is not a black box–we have a strong understanding of its capabilities and it shouldn’t be seen as beyond comprehension, but rather as an extremely advanced and capable tool in the development of your investment technique and portfolio. Read more about AI and its effect on What-If Scenarios, here. 

Stock Market Algorithm

What is a “What-If Scenario” Tool, and How Does it Work?

Scenario Analysis evaluates the expected value of a proposed investment or business activity. The statistical mean is the highest probability event expected in a certain situation. By creating various scenarios that may occur and combining them with the probability that they will occur, an analyst can better determine the value of an investment or business venture, and the probability that the expected value calculated will actually occur. Determining the probability distribution of an investment is equal to determining the risk inherent in that investment. By comparing the expected return to the expected risk and overlaying that with an investor’s risk tolerance, you may be able to make better decisions about whether to invest in a prospective business venture.

Historical performance data is required to provide some insight into the variability of an investment’s performance and to help investors understand the risk that has been borne by shareholders in the past. By examining periodic return data, an investor can gain insight into an investment’s past risk. For example, because variability equates to risk, an investment that provided the same return every year is deemed to be less risky than an investment that provided annual returns that fluctuated between negative and positive. Although both investments may provide the same overall return for a given investment horizon, the periodic returns demonstrate the risk differentials in these investments.

Strict regulations over the calculation and presentation of past returns ensure the comparability of return information across securities, investment managers, and funds. However, past performance does not provide any guarantee about an investment’s future risk or return. Scenario analysis attempts to understand a venture’s potential risk/return profile; by performing an analysis of multiple estimates for a given stock and denoting a probability for each scenario, one begins to create a probability distribution (risk profile) for that particular portoflio.

Scenario analysis can be applied in many ways. The typical method is to perform multi-factor analysis (models containing multiple variables) in the following ways:

  • Creating a Fixed Number of Scenarios
    • Determining the High/Low Spread
    • Creating Intermediate Scenarios
  • Random Factor Analysis
    • Numerous to Infinite Number of Scenarios
    • Monte Carlo Analysis

Pic 2: Deep Learning and ANN

Many analysts will create a multivariate model (a model with multiple variables), plug in their best guess for the value of each variable and come up with one forecasted value. The mean of any probability distribution is the one that has the highest probability of occurrence. By using a value for each variable that is expected to be the most probable, the analyst is, in fact, calculating the mean value of the potential distribution of potential values. Although the mean has informational value, as previously stated, it does not show any potential variation in the outcomes.

Risk analysis is concerned with trying to determine the probability that a future outcome will be something other than the mean value. One way to show variation is to calculate an estimate of the extreme and the least probable outcomes on the positive and negative side of the mean. The simplest method to forecast potential outcomes of an investment or venture is to produce an upside and downside case for each outcome and then to speculate the probability that it will occur. The figure below uses a three scenario method evaluating a base case (B) (mean value), upside case (U) and a downside case (D).

For example a simple two factor analysis:
Value V= Variable A + Variable B, where each variable value is not constrained. By assigning two extreme upside and downside values for A and B, we would then get our three scenario values. By assigning the probability of occurrence, let us assume:50% for Value (B) = 200
25% for Value (U) = 300
25% for Value (D) =100
When assigning probabilities the sum of the probabilities assigned must equal 100%. By graphing these values and their probabilities we can infer a rather crude probability distribution – the distribution of all calculated values and the probability of those values occurring. By forming the upside and downside cases we begin to get an understanding of other possible return outcomes, but there are many other potential outcomes within the set bounded by the extreme upside and downside already estimated. The figure below presents one method for determining the fixed number of outcomes between the two extremes. Assuming that each variable acts independently, that is, its value is not dependent on the value of any other variable, we can conduct an upside, base and downside case for each variable. In the simplistic two factor model, this type of analysis would result in a total of nine outcomes. A three-factor model using three potential outcomes for each variable would end up with 27 outcomes, and so forth. The equation for determining the total number of outcomes using this method is equal to (YX), where Y= the number of possible scenarios for each factor and X= the number of factors in the model.


In the figure above, there are nine outcomes but not nine separate values. For example, the outcome for BB could be equal to the outcome DU or UD. To finalize this study, the analyst would assign the probabilities for each outcome and then add those probabilities for any like values. We would expect that the value corresponding to the mean, in this case being BB, would appear the most times since the mean is the value with the highest probability of occurring. The frequency of like values increases the probability of occurrence. The more times values do not repeat, especially the mean value, the higher the probability that future returns will be something other than the mean. The more factors a model possesses and the more factor scenarios one includes, the more potential scenario values are calculated resulting in a robust analysis and insight into the risk of a potential investment.

How can I get this “What If Scenario”?

This service was meant for institutional investors who desire to see more detailed scenarios for their investment plan, though it can also be adapted to other needs.

Step 1: Decide which scenario is the one you have expectations for

Step 2: Send an email to [email protected] with your inquiry to get your quote.

To subscribe today click here.

I Know First Algorithm Heatmap Explanation:


This indicator represents the predicted movement direction/trend; not a percentage or specific target price. The signal strength indicates how much the current price deviates from what the system considers an equilibrium or “fair” price.

The signal strength is the absolute value of the current prediction of the system. The signal can have a positive (predicted increase), or negative (predicted decline) sign. The heat map is arranged according to the signal strength with strongest up signals at the top, while down signals are at the bottom. The table colors are indicative of the signal. Green corresponds to the positive signal and red indicates a negative signal. A deeper color means a stronger signal and a lighter color equals a weaker signal. The sign of the signal tells in which direction the asset price is expected to go (positive = to go up = Long, negative = to drop = Short position), the signal strength is related to the magnitude of the expected return and is used for ranking purposes of the investment opportunities.

An analogy with a spring: The signal strength is how much the spring is stretched. The higher is the tension the more it’ll move when the spring is released.


This measures the importance of the signal. The predictability is the historical correlation between the prediction and the actual market movement for that particular market. For each asset this indicator is recalculated daily. Theoretically, the predictability ranges from minus one to plus one. The higher this number is the more predictable the particular asset is. If you compare predictability for different time ranges, you’ll find that the longer time ranges have higher predictability. This means that longer-range signals are more important and tend to be more accurate.

Predictability is the actual fitness function being optimized every day and can be simplified explained as the correlation-based quality measure of the signal. This is a unique indicator of the I Know First algorithm. This allows users to separate and focus on the most predictable assets according to the algorithm. Ranging between -1 and 1, one should focus on predictability levels significantly above 0 in order to fill confident about/trust the signal.

To subscribe today and receive exclusive AI-based algorithmic predictions, click here.