TMethod + Metrics

TMethod + Metrics

This page includes the portfolio manager’s objectives and the TST value proposition, metrics categories, model framework and definitions of the TST QL+QT Ranking Model Template.   To provide context, we use a set of Morgan Stanley research reports as an example.  Seeing how Morgan Stanley used the TST QL+QT Ranking Model Template in a real case study will help you understand how all of the elements of the Template fit together.  The fundamental value proposition, objective, methodology, core structure, and five metric categories have not changed.   However, the number of metrics and their sub-categories and weights have changed as noted in the next section.

Using Metrics Ranking Model to Prioritize the Drivers

This TST Metrics Ranking Model helps you prioritize your variables (metrics) based upon the weights you assign them in their models.  This enables you to focus on the drivers of your stock selection process — which Malcolm Gladwell refers to as “the few variables that matter” (Blink).  Visualizing how you rank your metrics will often lead you to reduce your number of metrics and to re-define and re-weight them.

For example, the same senior research analyst used a different set of metrics and definitions in his models at Morgan Stanley and Evercore Partners below.  He also redefined “catalysts” to “short-term catalysts” and “x factors” to “black swans”.   Catalysts and x factors were the two highest weighted variables at Morgan Stanley and catalysts, risks and black swans the three highest weighted at Evercore.

Users often reduce the number of QT metrics when they see that they have intuitively attached greater importance to their QL metrics than to their lowest weighted QT varuables.   For example, the analyst reduced the number of QT metrics from 25 in his Morgan Stanley to 6 in his Evercore models — in part due to the ranking model clearly highlighting the lesser importance of the lowest weighted QT metrics.  He then culled the variables in orange in the Morgan Stanley model with 2% overall weight from his Evercore model.


3 Common Risk Factors

The  3 Common Risk Factors help you differentiate your stocks based upon a set of common risks.  To leverage this method effectively, you must understand the difference between common risks which are covered under the Risks category and stock-specific risks which are covered in catalysts and x factors.  For example, the Deepwater Horizon & Exxon Valdez were risks specific to BP and Exxon, whereas Oil Prices is a risk common to the whole group.

The common risks are subjective, customizable and dynamic.  Users differ on their opinions as to the 3 most important common risks and often change them based on evolving market conditions.  E.g., you can see how the common risks vary from user to user and industry to industry from the table below.


3 Common Risks for Oil Stocks Model

The 3 current common risk factors, in order of seriousness, are oil prices, businesses and geopolitical.  The portfolio manager makes a judgments as to whether each company has an above average or below average risk relative to the group.  If above average risk, the manager uses a drop down box to insert the letter abbreviation for that risk; if below average, the manager leaves the risk blank.

Oil Price risk is the risk higher or lower oil prices will have greater impact on increasing or decreasing the price of the stock.  The current assumption is that oil prices stay within a $40 to $60 range for the next 12 months.

Businesses risk is the risk that the company has a poor mix of businesses.  The best businesses are integrated; second best is services; and the third is E&P and drilling.

Geopolitical risk is the risk that the intergovernmental relations or the actions of non-governmental terrorists will impact the stock price.

Morgan Stanley Description of THINKSheet

Morgan Stanley’s definition of its TST model factors and the overall TST methodology is described below.  We place the greatest emphasis on the qualitative factors because the definitions and methodology for QT factors has been standardized in the investment world.

Model Objective

The aim of the exercise, as with our prior ranking system, is to identify the cheapest stocks with the best fundamentals that have a near-term, positive catalyst, low risk, and beneficial X-factors.

Model Definitions

Catalysts Definition

The investment world, the buy side and sell side alike, are in constant search for the burning “catalyst” for any particular stock story. Potential catalysts can be good,bad, or non-existent depending on the time-period.  Our model weights the catalyst factor heavily or 17.5% of the entire weighted sum score. We score the catalyst factor employing seven different scores between 100% (the most positive) and “0”.

Catalysts represent important factors which we believe not to be currently fully reflected in the stock price or fundamental operating numbers. These may consist of expected events, announcements, etc., that we anticipate will move the stocks up or down. This exercise is obviously highly subjective. In the intangibles breakdown we include descriptions of the catalyst (i.e. “tough comps”, or “higher interest rates”) and a symbol ranging from +++ to — which connotes excellent to poor. In our complete ranking overview, we limit our discussion of catalysts to the scale ranging from excellent to poor.

X Factors Definition

We similarly score X-factors employing seven different scores. We also weight the X-factor a high 17.5%. In essence, the X-factor is our judgment about some event, quality trait of the company, secular or cyclical trend or recent event (that we do not believe the market has yet factored in properly) which we believe is critical to ranking the attractiveness of the stock.

In this section, we evaluate a stock’s exposure to the three most common risks in the publishing sector. Those risks are Expectations (denoted by an X), which we view to be the most serious, stock Execution (E), and Acquisitions (A), which we view to be the least pernicious. It is possible for a stock to range from no exposure to these three most common risks, to exposure to all three (communicated as X&E&A).

Risks Definitions

We assign a 5% weighting to Risks. The risks we assess are straightforward and include Expectations, Execution, and Acquisitions. In essence, we score the stock for the risk that expectations are too high, for concern of operating execution, and for the risks attendant with the potential for stepped-up acquisition activity.

….  We evaluate a stock’s exposure to the three most common risks in the publishing sector. Those risks are Expectations (denoted by an X), which we view to be the most serious, stock Execution (E), and Acquisitions (A), which we view to be the least pernicious. It is possible for a stock to range from no exposure to these three most common risks, to exposure to all three (communicated as X&E&A).

Overall Methodology

Below, we describe in greater detail the factors, weights and inputs behind our five major categories.

Critical Issues Factors

The weights for the 3 Critical Issues factors total 40.0%:

  1. Catalysts 17.5%
  2. X-Factors 17.5%
  3. Risks 5%

Quantitative Factors

The weights for the 2 Quantitative Category and their 25 sub-categories total 60%:

Valuation (25.0% overall weighting).

We are using the following eleven criteria in our valuation ranking: 1)  2005E ModelWare P/E to growth rate; 2) 2005E ModelWare Enterprise Value to EBITDA; 3) implied upside to the theoretical stock price using a 10-year discounted cash flow analysis for each company to the current stock price; 4) implied upside to the private market value determined by breaking down each company’s business segments and valuing each cash flow (EBITDA) stream separately to the current stock price; 5) ModelWare 2005E free cash flow yield; 6) current dividend yield; 7) insider selling netted against purchasing (for the last six months) as a percentage of total shares outstanding; and the four elements that comprise our technical stock analysis provided by the Morgan Stanley Technical Strategy team: 8) support level, 9) first and 10) second levels of resistance, and 11) overall technical outlook. We weight each of these eleven criteria to arrive at an overall valuation ranking.

When talking about valuation, beauty tends to lie in the eyes of the beholder. Nonetheless, we have tried to assemble a relatively encompassing set of absolute and relative criteria capturing both current momentum and intrinsic measures and shaped to media sensitivities. For instance, we do not use price/earnings ratios in a vacuum. Rather, we look at P/E compared to earnings growth; in turn, growth is measured on a compounded basis from 1996 to our projected 2008(E). In addition, by equally weighting a number of criteria, we are not wedding ourselves to any one benchmark.

Valuation (Cont’d)

For instance, while many in the investment community believe discounted cash flow valuation is among the purest valuation measures, we note the large number of judgment calls involved in this methodology. These include the appropriate required return on equity that goes into the weighted average cost of capital as well as the long-term growth rate assumptions we set for over 20 stocks.

Why include technical analysis? Clearly, nothing in the stock market should be viewed in isolation, in our opinion. However, stock price movements accompanied by changes in volume patterns often signal a distant turn in the fundamentals. To ignore this input — to which we assign a small weighting of 5% — would be foolhardy, in our view.

Financials (35.0% overall weighting).

This category simply measures the fundamental operating performance of each company with emphasis on recent historical data as opposed to just projected results.

The fourteen criteria we include are: 1) 2005E organic revenue growth, which excludes acquisitions and divestitures; 2) a measure of overall profitability or estimated 2005 ModelWare EBITDA margins; 3) ModelWare Return on net operating assets (RNOA) defined as net operating profit after tax divided by net operating assets (beginning period); 4) expected EPS change for 2005 using our forecasts; 5) 1996–2008 EPS percent change (CAGR); 6) ModelWare 2005 estimated