Klein Decisions Archives
The KleinPost

Feature Article

Inside K4

Model of the Quarter

Tips from the Klein Bottle

Ask Felix

The ReKleiner


More Tips

Basic Factor Groups

Importance Strategy

Filters For "Dummies"

Track Record - Simple but Critical

Simple Models Are Often The Best

Copy and Save Time

Pros and Cons of the Distinct Portfolio Filter

 

Click here to see all tips.
 

Sign Up Today!
* required
*

*


*

*


Powered by VerticalResponse

 

KleinDecisionsLogo

68 T.W. Alexander Drive

P.O. Box 13628

Research Triangle Park, NC 27709

919.233.6767

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Importance Strategy
Don't Be Intimidated By The Preference Exercise.
By Dr. Felix Kleinstein

One of the key strengths of Klein Decisions’ K4 Fund Selection and K4 Manager Selection is their ability to create weighted factor models. Unlike simple screening processes based on equally weighted pass/fail tests, the K4 tools enable you to place more emphasis on those factors you feel are more important, evaluate all products versus all factors, and then rank order rather than eliminate them. Obviously, the factor weights play an important role in determining the product rankings. Both K4 Fund Selection and K4 Manager Selection give you three ways to do this.

Equally Weighted – This is the simplest approach and may be appropriate if you want to create a quick scenario. With one click of the mouse, you can equally weight all factors and go right to the ordered results. Although easy, this approach doesn’t allow you to incorporate your intuitions about the factors’ relative importance. To do that, you’ll need to use one of the other two options.

Specific Factor Weights - Both K4 Fund Selection and K4 Manager Selection let you enter specific percentages for each factor. These represent the weights and must total 100%. This is a good choice for analysts who have already developed their own models and want to automate them with K4. But many users who believe some factors are more important than others don’t feel comfortable quantifying those beliefs. For them, the third option is more appropriate.

Preference Exercise – Through a process known as “adaptive conjoint analysis,” K4 Fund Selection and K4 Manager Selection will convert your intuitive beliefs into consistent quantitative factor weights. The process employs two simple sets of questions: “Importance of Difference” and “Trade-Offs.” The first set asks you to rate the importance of each individual factor, from “extremely important” to “not important.” Based on your responses, a second set of questions is generated and asks you to compare the relative importance of two factors at a time. Based on your responses to these simple questions, K4 then calculates specific factor weights.

Although the Preference Exercise is a simple process, it can be intimidating to novice users. This is particularly true in regard to the Importance of Difference questions. Some simply resort to giving them all equal importance and, by default, creating an equally weighted model. That’s certainly an alternative, but if that’s your goal, you can do it more quickly and easily by selecting the Equally Weighted option.

There’s a simple strategy that can help reduce and possibly even eliminate this anxiety. It doesn’t require you to spend a lot of time or do a lot of analysis or calculations. It can, however, provide some structure. First, it’s important to realize that none of your responses are written in stone. You can always come back and change them if you subsequently decide they weren’t quite right. Secondly, adding a little rigor to the way you respond to the Importance of Difference questions can be a big help. Again, this doesn’t have to be complicated.

Consider the factors in this illustration drawn from K4 Fund Selection:

  • Expense Ratio
  • 5-Year Total Return NAV
  • 5-Year Standard Deviation
  • 5-Year Down Market Ratio
  • 5-Year Sharpe Ratio
  • 5-Year Batting Average

Of those six attributes, one has to stand out as the most important. In this case, it might be the 5-Year Sharpe Ratio. If so, mark it as “extremely important.” If it’s a hard decision between the Sharpe Ratio and the 5-Year Total Return, decide which of them is most important (a simple trade-off) and mark the winner as “extremely important” and the other one notch below. If you can’t decide between the two, make them both "extremely important."

How Important is...

Now go to the other end of the spectrum. Of the remaining attributes, one or two must seem less important than the others. In this case, it might be Expense Ratio (since all returns are net anyway). If so, mark it “not important” or “somewhat important.” Standard Deviation’s already part of the Sharpe Ratio, so it’s probably less important than the other remaining attributes. Place it a little to the left of Expense Ratio by either designating it as “important” or a notch below “somewhat important.”

Now only two attributes remain and you know they should fall between “very important’ and “important.” Rather than having to find them a home among nine alternatives, you’re now down to three. What remains is a choice between two alternatives which should be relatively easy. In fact, when broken down in this manner, the entire process is very manageable even with a greater number of factors. Once you’ve assigned a relative importance to each factor, take a look at the overall distribution. Are they all bunched together? Do you need to add some space to reflect greater differences in importance? If so, go ahead and adjust them before moving on to the Trade-Offs.

Finally, consider your responses to the first set of questions when you answer the Trade-Offs. Think back to the level of importance you assigned to each. If they were relatively close, the degree of preference in the Trade-Offs should be close to indifferent. On the other hand, if one factor was “extremely important” while the other was only “somewhat” or “not important,” the first should be strongly preferred to any in the second group. Some users keep track of their responses to the Importance of Difference questions by numbering the response 1-9 from left to right. They then refer this list to help them answer the Trade-Offs. Unless you really want to, there’s no need to go to this length. By applying the principles of adaptive conjoint analysis, the K4 tools will come up with consistent results based upon your answers. No additional work is required on your part; just provide what you naturally think are the most important responses.

That’s one of the major advantages of the K4 decision tools. They provide a sophisticated and thorough analysis with simple and straightforward inputs. There’s no need to be intimidated by the importance rating exercise. Just relax, answer naturally, and let K4 Fund Selection and K4 Manager Selection do the heavy lifting.