The risk model is a core component of the advice engine in Deep Alpha Platform (DAP). The objective of the risk model is to classify the risk class of the investor and thereby be able to match this risk score with a suitable portfolio. By answering a set of questions, we will enable you to map the user into a risk class. The risk model is configurable, and allows you to adjust it to fit your investment philosophy and routines of mapping the risk score of an investor. Read more about this in the section under "Configuration possibilities"
The power of the risk model is that equal investors are equally assessed. This is highly appreciated by our customers, as it assists their advisors and ensures full transparency on how each individual investor has been classified.
The default risk model consist of the following look-up tables:
- Initial risk tabel
- Adjustment tabels
The initial risk tabel consist of time horizon input and up to two risk questions
Time horizon: The time horizon is given by the investment goal. This is typically expressed with the following question "What is the time horizon for the investment?" In the default risk model we operate with 4 time horizon levels:
- Less than 1 year
- 1 to 3 years
- 3 to 5 years
- 5 to 10 years
- More than 10 years
In the default risk model we ask two risk questions, each question has four answer options, refering to different risk levels:
Risk 1: Generic risk question to assess their "What is your expectation of risk?"
Risk 2: Scenario question, where different investment scenarios are presented .This is used to cross-check the answer from the investor: "Where would you invest your money?"
- 2 % return with 13 % upside and -11 % downside
- 4 % return with 16 % upside and -13 % downside
- 6 % return with 20 % upside and -17 % downside
- 8 % return with 24 % upside and -26 % downside
Initial risk score
The input taken from time horizon and risk is used to map the initial risk score per investment goal. This is done by a matrix that look like this;
Assuming that the investor has timehorizon = 3 and answer risk 1 = 3 and risk 2 = 4, the mapping will be as illustrated in the picture below:
The initial portfolio will be defined as the avarage of the two risk scores. In the example above, the initial risk score of the investor is 6. This initial risk score is kept for lookup in the adjustment questions.
After the intial risk score is defined, we apply adjustment questions. These questions and answers will adjust the initial risk score classified in the step above. In our default model, the questions are covering the following topics:
- Adjustment 1 = Loss aversion
- Adjustment 2 = Ability to bear losses
- Adjustment 3 = Knowledge and experience
The question we use is the following: "If your investments fall 20 % in value, how would you react?". The answer options presented to the investor are as follows:
- I would use this as an opportunity to invest more
- I would not make any changes
- I will get uneasy but not sell my investment immediately
- I would sell my investments
The matrix in our default risk model for this question looks like the table below. As you can see we reduce the overall risk if the investor demonstrate loss aversion (answer 3 and 4), we also adjust this down more for higher risk class down . The adjustment effects can be configured.
In the table above, you will see that we are working on the column that correlates to the initial risk score of the investor. Assuming that the investor answer "I will get uneasy but not sell my investment immediately", the adjustment 1 = -1.
The question we use is the following: "How much can your investment fall in value before it affects your financial situation?". The answer options presented to the investor are as follows:
- This investment is not important for mye financial situation
- 40 %
- 20 %
- 10 %
The matrix in our default risk model for this question looks like the table below. As you can see we reduce the overall risk if the investor demonstrate low ability to bear losses (answer 3 and 4), for the high risk classes. At the same time, we reward having a large ability bear losses for lower risk portfolios (answer 1 and 2). The adjustment effects can be configured.
Assuming that the investor answer "10 %", the adjustment 1 = -1.
The adjustment 3 question is a cross-matrix that we use to adjust the risk level based on the input of knowledge and experience of the investor. The questions we have in our default solution is as follows:
Knowledge: "What knowledge and experience do you have with investing in funds?"
- I have no experience
- I have education in economics and/or finance
- I have work experience that makes me understand financial products and services
Experience: "How many time have you traded stocks and funds over the last five years?"
- Less than 5
- Between 5 and 10
- More than 10
The default cross-matrix we use looks like this:
In the example above the investor has neither education or work experience but demonstrates some level trading experience. That input combined
Final risk score
Based on the examples above, the adjustment scores are as follows
- Adjustment 1 = -1
- Adjustment 2 = -1
- Adjustment 3 = -0.25
Final risk score for the investor is then given by the following formula: Final risk score = Intial risk score + adjustment 1 + adjustment 2 + adjustment 3 Final risk score = 6 - 1 - 1 - 0.25 = 3.75. After rounding the final risk score of the investor = 4.
We allow for configuring the risk model, to fit your way of advising investors. Configuration of the riskmodel consist of two steps:
- Configuration of the risk model in the advice engine
- Adjusting questions in the Purpose and Risk section and Knowledge and Experience section of the advisor solution
The default risk model comes with 5 time horizons. However, you can apply as many time horizons as you want.
You can select whether you want to use one or two risk questions. The minimum requirement is that you use one. This is the minimum requirement to be able to calculate the initial risk score
The risk adjustment questions are not required. You can select whether you want to use all risk adjustment questions or non of them.
In the advisor solution of DAAP, the risk questions used in the risk model are found in the following two pages:
The risk model is universal, across your product platform configurations and when you use multiple goals, each goal will get an individual risk score.