Jan 17

Why we believe in visual learning

by David Harper, CFA, FRM, CIPM

The short answer: because Kathy Sierra says words + pictures > words alone Also, she says we should add graphics.

The long answer has three parts, visual learning is: faster, better, and more fun. In short, visual learning is bionic.

First, a picture is worth 895 words (our lab ran the numbers, turns out to be not quite 1,000 words). For example, if you are trying to learn Basel II, the first step is getting your head around disparate elements. That is, getting to know the terrain. The main document is 284 pages, so that's not so easy. You've got to figure out, for example, the approaches to credit risk within the first pillar. Which is easier, text or a picture? Here is the text...

This section of the Framework describes the IRB approach to credit risk. Subject to certain minimum conditions and disclosure requirements, banks that have received supervisory approval to use the IRB approach may rely on their own internal estimates of risk components in determining the capital requirement for a given exposure. The risk components include measures of the probability of default (PD), loss given default (LGD), the exposure at default (EAD), and effective maturity (M). In some cases, banks may be required to use a supervisory value as opposed to an internal estimate for one or more of the risk components. [Part 2. III. A. 211., International Convergence of Capital Measurement and Capital Standards, Basel Committee on Banking Supervision]

...compared to the same idea, really, conveyed in the picture below. Actually, the picture below contains more information than the paragraph above: it shows where supervisory values must be used in the Foundation IRB approach. The key differences between Foundation IRB and Advanced IRB can be summarized quickly in a chart:

Second, images are easier to remember. Most memory techniques rely on connecting concepts to images ("...use positive, pleasant images"). For example, an important idea (in the the FRM curriculum) is the idea of random or stochastic processes. As often happens in math, some of the processes are merely special cases of the others (or conversely, general cases on others). At first blush, the list of standard stochastic processes can be daunting; i.e., Wiener Process, the Markov Process, and Geometric Brownian. But these are related, and frankly, it's mostly a matter of learning one, the most general case. The others omit terms to become special cases. This is easier to understand with a diagram: