Monday, July 19, 2010

AG: Credit Risk Evaluator

Credit Risk Evaluator

The Library of Credit Portfolio Management

With the completion of Basel 2 financial institutions have established the prerequisite of modern credit risk management.

The Credit Risk Evaluator enables financial institutions to take the next step over the rating of single customers to an integrated risk and return management of the full bank portfolio.

Internal rating methods can create additional value for banks and financial institutions beyond a more economical equity allocation.

The Credit Risk Evaluator rests upon the data foundation laid by Basel 2 and links it to the modern credit portfolio models. It delivers and reports to the bank risk and return analyses on all portfolio levels from the single client or even the single deal over arbitrary aggregation levels to the full bank portfolio.

It creates the basis for the identification of the well diversified and the highly concentrated business and market areas up to the single client where the intended return on equity is reached or failed, respectively. From the results impulses can be derived for sales and management.

Design

The Credit Risk Evaluator is designed as an open library of credit portfolio models. It comprises today 7 models which can be applied in analyses by mouse-click:

Credit Metrics
The KMV model
A copula asset value model
Credit Risk+
Credit Risk++ - a generalization of Credit Risk+ for rating migrations, intersectoral correaltions and variable recovery models
A multi-period model that generalizes Credit Metrics and the KMV model.
A model that generalizes the copula asset value model to multiple periods.


Further models can be easily integrated.

All portfolio models can be combined with several recovery models to allow for a precise valuation of sub-portfolios and for the trading of credit risks.



Results

The Credit Risk Evaluator aims at an integrated view of risk and return. Under both aspects it delivers a number of key indicators on all levels of the analysis:

Value at risk, shortfall, standard deviation, expected loss
RAROC, economic value added (EVA), risk adjusted return, costs of risk, costs of equity

The results are visualized by a large number of selectable standard reports and linked to each other to make them accessible for interpretation.

Technical aspects

The Credit Risk Evaluator is componente oriented in design. Data storage, business logic and control and reporting are technically fully separated from each other and can be distributed to different computers.

The performance is model dependent. It is for the single-period models between 3-7 minutes per 100,000 clients and 10,000 simulation runs on a standard pc. For the multi-period models the runtime slightly higher. In all cases the run times scales linearly in the number of clients and simulation runs.

Analyses can be done with manual intervention as batch-processes. Reports are automatically generated and archived at runtime.

Correlations and parametrization

Credit portfolio management often requires data that is not immediately at hand.

We aid you to determine correlations and other model parameters relating to your portfolio so that you can make full use of the potential of credit portfolio models.

Your advantages

Additional usage of the investments into Basel 2
Bank-wide portfolio management also in large institutions possible
State-of-the-art credit portfolio and recovery models at hand
Extendible for bank owned portfolio and recovery models
Simultaneous consideration of risks and returns
Highly automatized processing with minimum manual intervention

AG: A new way to value the market

(Fortune Magazine) -- Are stocks cheap yet? That slippery, eternal question is worth a look right now because a remarkable new set of data has just become available, allowing us to analyze the market in ways we never could before. I wish I could tell you that this new trove of numbers reveals that stocks are a screaming buy. It doesn't. But it does suggest that, amid all the recent tumult, just maybe the market is being rational.

The new data are derived from the most fundamental, capital-based way of analyzing a company's finances and value. How much capital is a company using? What is its return on capital? How much does the capital cost? Those questions hold the key to corporate performance, but finding the answers in most financial statements isn't easy, and many executives don't know the answers themselves. The Stern Stewart consulting firm began popularizing these concepts more than 15 years ago with the term EVA (economic value added), and the new data come from EVA Dimensions, a firm that is now the source of Stern Stewart's EVA data.

EVA-based analysis has proven extremely valuable in analyzing individual companies. I almost never make calls on specific stocks, but in late 1999 the EVA analysis of AOL was so compelling that I wrote a column declaring flatly that the stock price could not possibly be justified. That column was published on Jan. 10, 2000, right near the overall market peak (and the very day that AOL announced it was using its insanely overvalued stock to buy my employer, Time Warner (TWX, Fortune 500) - but that's another story). I also used EVA analysis to write last summer that Google (GOOGLE) was overpriced at $540; that call looked wrong for a while, though as I write this the stock is at $501.

One thing you couldn't do with EVA analysis was use it to value the whole market. Compiling the data for a significant number of companies used to take months. But now, through the miracles of our networked world, EVA Dimensions can compile it every day for 2,669 companies in the Russell 3000 (those for which at least two years of data are available). This is essentially the U.S. stock market. So: Is it worth what it costs?

Look first at how well the companies are doing at their most basic task, which is earning a return on their capital that's greater than the total cost of that capital. Turns out they've been doing very well. The dollar difference between their return on capital and cost of capital (their EVA) was $375 billion over the past four quarters. It was only half that much in 2005, and in 2004 it was negative, which isn't surprising. Over time, for the broader market, EVA should be more or less zero since competition is always forcing high returns down toward the cost of capital, while companies that can't meet their capital cost will eventually go under. So America's publicly traded companies did great last year; in fact, with economic growth strong through the third quarter, it's safe to say that they were at or near the top of the business cycle.

Next question: How are they being valued? On a recent day when the Dow closed at 12,265, the 2,669 Russell 3000 companies had a total enterprise value of $29.8 trillion (equity plus debt). To judge whether that's a lot or a little, consider that over the past four quarters these companies produced after-tax operating profits of about $1.8 trillion. Even if we assume that earnings will only match, not exceed, that level in future years, then the companies' aggregate market value today would still be $22.5 trillion (note to finance wonks: that's their profits capitalized at their capital cost of about 8.1%), which is about 75% of their actual market value.

So now we reach the central question. About 25% of the current market value of these companies is based on expectations of future profits above and beyond the profits they earned last year, at the top of the business cycle. Does that seem reasonable? Actually, it just might. The math gets a bit tedious, but you can assume no profit growth for the next several years and very modest growth thereafter, and the valuation still looks okay.