When comparing projects, in order to be able to choose the one with the highest expected return, it is essential to have the proper tools for such an evaluation. And again, the finance theory has some excellent tools/concepts/models for this purpose.

In earlier days, assessments were only base on assigning some values for the expected development of key variables, and an expected result was calculated based on these static figures. This is all good and well, it’s just that all these assumptions are uncertain, so called stochastic. And to try to find ONE value for such a variable is like trying to pin jelly to the wall. Can you picture it? It is very difficult to pin jelly to a wall.

So what we have to do is to not only express the expected return, but also the risk it is subject to.

In order to accomplish this, a more fine-meshed net is needed. This net is obtainable in using a **simulation model**. Meaning that you map all the variables, and the probability distribution they can be expected to have. An example:

Variable A has an expected value of 1. However, after careful consideration, we think that it is possible that this value can become 0, and it is possible that it can become 2. In this very moment, dear daters, we have defined a probability distribution for variable A. We do this for all relevant variables, and in addition we assess the relationship between the variables, either by defining an equation, called a regression, or by defining the way in which they move in the same or in opposite directions, called a correlation. We define such target variables mathematically, build a sophisticated model which we then run through the wringer, **known as Monte Carlo simulation** 1). The name proves also that statistics/finance nerds aren’t as unsexy in their perspectives as you would think; we have indeed understood that fast cars are something special. Monte Carlo simulation was developed for the real big bangs, and that is exactly what daters are looking for. The Big Bang.

So, dear daters, let us make a model. Let us call the model the **Standard model for date-selection (SMDS)**. With this model we determine who are worth dating. The target function is: Expected return of date. The model will give estimates of both expected return and risk. Note that the dater must set values for all variables, as they are individual based on preference.

To simplify we define all variables to have possible values between 0 and 100, where the normal distribution approach gives and expected value of 50. Or to put in differently: the average dating material has a value of 50. Then we define variables who will describe a successful date:

1. **How does the person look**? We have of course seen pictures.

2. **How talkative is the person?** Have we spoken on the phone? In general it is a very good idea to speak on the phone before meeting.

3. **Does the person have a sense of humour which matched out own**, indicating that we have a fun evening ahead?

4. **Does the person seem honest?** This is a very difficult variable to assess. We have to rely for the most part on intuition, but daters who have been around the block a few times usually have good intuition.

5. **Is the person in the forward market?** This is of course essential information, see «Finance applied to dating (1)»

6. **Are you in the forward market?** It is of course very important that the two parties in the dating game have the same duration in mind.

In addition correlation between variables 2 and 3 is assumed. I have rarely met someone who is speechless and at the same time has a good sense of humour. And it really doesn’t help to have a great sense of humour if you’re not able to say much.

Our simple model then can be formulated as:

**AVERAGE**(V1,V2,V3,V4,V5,V6) given **CORR**(V2,V3).

Let us assume that there are two men I consider dating, but for various reasons I can only date one of them. Let us call them **Andy and Pete**. I use the **SMDS-model **to find out the following:

1. Does Andy or Pete have the highest expected dating value?

2. How risky is a date with Andy and Pete? We measure this by the calculated standard deviation of the expected dating value, giving us a number which measures risk.

3. If both daters get an expected value below 50, they are both expected to be below average dates. Indicating that it might be a good idea not to date any of them.

We put information about Andy and Pete into the model:

The table shows that Andy seems to be an above average expected date, whereas Pete seems to be somewhat below average. However, we need to run the model in order to get more information about Andy and Pete.

**Simulated results:**

**Plotted:**

And voila! All of a sudden we have a lot of interesting information about Andy and Pete. For example:

Andy still has the highest expected dating value, but it is lower than in the base case. Using finance lingo – risk adjusted return is lower than expected return. Pete’s risk adjusted return is a bit higher; he is now on average.

Andy has a higher standard deviation than Pete. This is because Andy is more risky. Even if we consider the fact that Pete’s average is lower, Andy is still more risky (12,7%) than Pete (11,5%).

Andy’s minimum value is lower than Pete’s. Thus, at date with Andy can be Less of a success than a date with Pete. Very interesting, since he has a higher expected value. Without question Andy is more risky.

Andy’s maximum value is substantially higher than Pete’s. There is much more of an upside in dating Andy.

What the model owner needs to do, when these results are in, is to look at her own risk carrying ability. And if she can carry the risk Andy represents, choose to date him since his expected return is much higher. If she cannot carry the risk Andy represents, she should consider dating Pete, with the lower level of risk he represents. However he is an average date, with little upside potential.

**I would probably choose to date Andy.** Because of the potential upside, I can live with the downside. Yes, dating Andy can be less of a success than dating Pete (28 to 30), but my God! It can be much more of a success as well (82 to 69).

The model can of course be built out and refined. After all we live in a complex world, and lord knows humans are more complex than most other creatures on this planet. Models can be tailor made to fit the individual dater, by contacting Iskwew@dating-simulationmodels.com, and against a small fee. A fee which by no means will be exorbitant, considering what you get out of it.

Isk

*just saw a business opportunity*

1) The term Monte Carlo was introduced during World War II as a code name for simulation of problems associated with development of the atomic bomb.

Originally written in Norwegian.

Iskwew ©

## Siste kommentarer