Targeting Investors |
Aug 20,1998
Finding new investors for your company's stock is a form of marketing. Valuation Technologies gives you the information to establish quickly and efficiently the needs of the potential investors, and see how your company can meet these needs. To make most effective use of your time, you focus on those investors most likely to view your company favorably. Once you have narrowed the list, the next step is to determine which characteristics of companies are most important to each potential investor. With this knowledge, you can tailor your presentation to each investor, emphasizing features of your company of particular interest. You cut right to the issues each investor uses in discriminating among securities, avoiding wasting his time on items that he does not employ.
Valuation Technologies is the first to apply an understanding of modern investment methods to help you with both steps. We analyze each investor's stock selection process. By studying the claims of managers and the actual returns to securities, modern finance can isolate those criteria used in stock selection. Each investor has his own specific approach, yet different combinations of relatively few basic criteria explain the vast majority of investment decisions. Next, we see how well your company meets each investor's process. If you are a close match, the investor will likely be interested in a deeper conversation. If not, you can probably spend your time more fruitfully with other investors. Finally, we select the best matching investors, and provide easily understood reports, explaining why they should buy your stock.
Despite an air of mystery about money management, the basic steps are rather simple. An active manager, who is trying to outperform an index, is looking for good stocks to buy. He may be using a computer screening system, scanning sell side analysts' reports, reading the Wall Street Journal, or listening to his barber. He is looking for stocks that meet his criteria as a good investment.
Our approach is to figure out what criteria or combinations of criteria motivate each manager. Criteria are calculations used to separate those securities the investor expects to perform well from those he expects to perform less well. Examples of criteria are Dividend Discount Value, Price-Earnings Ratio, and Changes in Analysts Expectations.
We don't pass judgment on how a manager should choose stocks, nor do we ask him to tell us how he picks stocks, we simply observe what he actually does. We look at which stocks he is holding, which ones he is buying, and which ones he is selling. We then determine which criteria are important in explaining his process. Even though individual stock may come and go from his portfolio, the underlying process is usually very stable. For example, a manager may be a sector rotator, moving from the currently less favored sectors into higher performing sectors. The actual securities held could change dramatically, but the strategy remains constant.
The next step is to match your company against the pool of investment institutions. We determine how your company looks on each of these criteria. Since we now understand which criteria best explain managers' actions, we look at each investment institution, examine the criteria of most importance to that particular manager. We then see how closely your company matches these important criteria. We weight criteria based on its importance to each institution. Those institutions with a close match to your company are likely to be receptive to further conversation.
How do we know which criteria motivate investors as a group?
Modern Portfolio Theory has been developed by academics and practitioners seeking to understand and improve investment management. The earliest development, starting in the 1960's, was Efficient Market Theory, leading to index fund management. Since then, more detailed analysis and sophisticated econometrics have helped understand active portfolio management. BARRA, the premier supplier of analysis to investment managers, has been in the forefront of these developments.
BARRA provides investment management software employing 17 criteria for security selection. Some are based upon value concepts, such as Dividend Yield, Price-Earnings Ratio, and Dividend Discount. Others are based upon growth and past success. Momentum models look at securities with extraordinary price movements, or at sectors of the market that have done well. Another group of criteria examines the recommendations of sell side analysts. BARRA has analyzed a wide variety of formulations and security selection criteria, and have found these to best separate high performing securities from low performers. BARRA provides Valuation Technologies with a database of these calculations for essentially all firms publicly traded in the United States. Each month BARRA re-computes these values to reflect changes in each security and in the market as a whole.
Each investor has his own combination of factors that he thinks identifies desirable securities. He may look for a low price-earnings ratio, plus a history of growth in earnings, combined with a recent run-up in price. If a stock meets his criteria, he is likely to buy. If it doesn't, then he will look elsewhere, or if currently an owner, consider selling.
How can we know what criteria an individual manager uses?
We could ask him, but are unlikely to get a useful answer. Our approach is to infer from his actions what strategy he employs. We use a group of 17 criteria that are representative of the various strategies. This list and the underlying calculations have been developed by BARRA, the premier provider of quantitative security research. For every stock, we get from BARRA the value of each criteria. Thus for IBM and General Electric, and all other traded securities we know their current Price Earnings ratio, Dividend Discount Value, Earnings Momentum, etc.
Next, we look at each manager's current portfolio. We obtain a list of what they own and how much they own from Vickers. We also get a list of their recent trades. We then take each security in the current portfolio, look up the value of each criteria, and develop a profile of the portfolio. Thus if a manager owns 50 stocks, for each of the 17 criteria, we will know the criteria for each stock.
We determine the largest and the smallest value for each criteria, for all the stocks in the portfolio. We compute the average and the standard deviation. We look at the average for each criteria, determining how far the average stock in the manager's portfolio is away from the market average. Next, we use statistics to determine how important to the manager is each criteria. How likely is it that the results we see happen by chance? If very unlikely, we conclude that the results are intended. If the manager's portfolio has a Price Earnings ratio twice as big as the average, did this happen by accident, or is the manager targeting high P/E stocks?
The smaller the standard deviation of a criteria, when computed using the manager's holdings, the more likely the manager is focusing on a specific value of this criteria in selecting stocks. Similarly, the further the average of a criteria in his portfolio is from the average in the whole market, the more likely the manager is consciously aiming at an extreme value. Those who are interested in statistics may like to know that we use the T statistic and the Chi Square statistic to separate chance from intention.
How can we tell if your stock is a good fit with the manager?
First, we look at the average value of each criteria in the manager's portfolio. If your stock is close to the average, it is likely you fit in. However, we weight the decision on the importance of the criteria to the manager. You may not fit his yield profile, but if he doesn't care about yield, it doesn't matter. Next we look at the portfolio's trend for each criteria. Is he increasing the value of the criteria by buying securities with a higher value than his current holdings? If so, then if your value is higher than the average, a good fit is indicated. We compute an overall figure of fit and determine whether this manager has a high or low fit with your stock.
How do we know investors' portfolios?
Basics of Institutional Ownership Data We obtain from Vickers Stock Research a list of money managers and mutual fund portfolios. These portfolios indicate recent holdings, and trades from the previous quarter. Vickers gets this information by obtaining SEC filings and mutual fund prospectuses. They then put all of this information into a consistent format and deliver it to us.
Is the portfolio information up to date?
Filings do take some time to work their way to us. The portfolio filings may correspond to holdings that have changed. However, we focus on the strategy, not individual securities. For example, a manager might be a sector rotator. We look at the stocks he held as of his latest filing, say July 1997. We then look at whether the buys were in high performance sectors and the sells in low performance sectors, as of the filing date. His portfolio shows us that he buys firms in sectors that are doing well and sells out of favor sectors. Currently, entirely new sectors may be in favor. To determine whether this manager is a likely buyer for your company, we check whether your company is in a currently favored sector.
How about non-quantitative managers?
These days even the most eclectic of managers has access to company data for screening, or uses a spreadsheet for analysis. He will generally prefer a particular type of stock, low P/E, high yield, recently popular, or recently unpopular, high growth or high value. These strategies will be revealed by our reports.
And Index Funds?
Many index fund managers simply buy the assets in the index, such as the Standard and Poorís 500. For these managers, no targeting will be helpful. However, many index fund managers do not buy all the stocks in the index, but sample from the index. For indexes containing a large number of securities, like the Frank Russell 3000, sampling is a popular approach. These managers will care to avoid surprises, and will take an interest in investor relations. So-called "enhanced index" managers are trying to outperform an index, while taking little risk relative to the index. They use quantitative methods to pick securities, but also use their judgment. Again, assurance that you are doing a good job of keeping them informed will increase their interest in your company.
Can we detect trends?
Yes. If the average value for the recently purchased portfolio is higher than for the held portfolio, which in turn is higher than the recent sales portfolio, we see a trend to increase a particular criteria.
How is this different from traditional targeting?
Others look at peer companies to see which managers ought to buy your stock. Peers tend to be in the same industry, perhaps similar in size. We know that most managers look much deeper at potential investments than simply industry. We look at a wide variety of active stock selection techniques, rather than just a few.
Do I have to understand mathematics to use your service?
No. We provide information in graphical form and text, using the language of the investment manager.
How can I use your reports to understand the fund and better communicate with the manager?
Our models describe the fund's interests in ordinary investment language.
What are the criteria you use?
We use a series of models which are designed to select desirable investments. A model separates all companies into groups, those expected to outperform, and those expected to underperform. For an example, the graph shows how the Cash Plowback model has performed over the last ten years. The following calculations capture strategies used by the vast majority of investment managers:
Predicted Beta
Predicts the amount of market risk obtained by investing in this company. BARRA Beta is a prediction of future market risk based upon company fundamentals, not simply a description of past market risk. Beta of 1.0 is average, indicating that a 5% market move would be expected to result in 5% return to this company. A Beta of 2.0 is very high, predicting a 10% return from a 5% market movement.
Looks at the amount of earnings that the company has reinvested over the past year. Companies that tend to score well here are relatively cheap compared to earnings and have a low payout ratio. A value measure
The score is based on annual implied return from a three-stage dividend discount model. We use forecast earnings and forecast growth rate to project a stream of earnings for each company. The asset's and industry's payout ratios are combined, yielding a forecast dividend for each company every year. The score is based on the implied return that equates the current price with the present value of the projected dividends. A value measure, however inexpensive growth stocks will score high
Based on momentum of historical and projected earnings. We look at some historical and forecast earnings data to estimate a growth rate and acceleration of earnings. A momentum measure.
Changes in consensus earnings estimates over a 2-month horizon. Here we use a combination of the earnings estimates for fiscal year 1 and fiscal year 2 to put each company on the same basis. The change over a 2-month horizon is measured to minimize the impact of very small positive or negative earnings forecasts.
Looks at the changes in earnings estimates as well as the price appreciation over the recent past to identify assets likely to outperform over the next few months. Given two stocks with the same earnings estimate changes, this model gives a higher score to the one whose price has appreciated less. Based upon analysts expectations
Combination of two factors relating to the attention the company receives from institutional managers: market capitalization and analyst coverage. High Neglect indicates low institutional coverage and small size, relative to other firms in the company's economic sector (industry group).
Earnings-to-price ratio using 12-month median projected earnings. We weight the company's projected 1- and 2-year median earnings forecasts depending on where in the fiscal year the company is. A value measure.
Sixty-month realized historical alpha. It is the portion of each companyís five-year return that is not related to the movement of the overall market. This is also the alpha estimated in the regression of 60 monthly returns that is used to find the stockís historical beta. A momentum measure.
Rankings based on price momentum over a 1-year horizon, adjusted for various common factors. Relative strength capitalizes on the positive relationship between a stock's return over the last year and its return in the future. Because this is contrary to the stronger short-term reversal effect, we adjust for very recent performance that goes into the reversal model below. This adjustment for recent performance means that the scores will not change over the month, so they are updated monthly. A momentum measure.
Stocks that have performed poorly in the previous month tend to rebound (slightly). A score representing predicted short-term performance based on the well-documented reversal effect. Our model uses returns neutralized for the market, for our 55 industry groups, and for some style factors. Assets with relatively low residual returns in the recent past have higher scores and are likely to outperform the market in the near future. A momentum measure.
Based on the recent performance of the stock's sector. Historically there has been momentum here, as sectors that have performed above average recently tend to continue to beat the market. If an asset has exposure to multiple sectors, then the score will be a weighted mixture of the momentum of the various sectors. A momentum measure
Cash Flow to Price
A typical cash-flow-to-price ratio over the past five years. It is calculated as the average cash flow per share divided by the average year-end closing price. Again, it is a good measure of relative value. A value measure
Ratio of most recently reported earnings (annualized) to most recent price. This is the inverse of the traditional Price - Earnings Ratio. High Earnings-to-Price indicates a value company, in the sense that the investor is buying past earnings at a low price. This model can be compared with Predicted Earnings to Price Ratio, a foreward looking measure of value. A value measure.
Size is used as an initial screen for most institutions as well as a factor in most models and styles. It includes market capitalization as well as total assets of the underlying firm.
Trading Activity
Like size, trading activity is often an initial screen as well as a factor in models and styles. Trading Activity represents several measures of the stock's liquidity, including turnover, institutional coverage, and level of stock price.
Value/Growth reveals whether the market views a company as a value company, or as a growth company. Value companies, those with a high Value/Growth, have a high ratio of book value to market price, allowing the investor to purchase net assets cheaply. Low Value/Growth companies are priced to reflect the expectation of above average future growth. A value measure
Another primary screening factor is yield. Yield is calculated using the current annualized dividend and most recent stock price. A value measure.
Valuation Technologies LLC - All Rights Reserved 1998