Risk Associated With Individual Stocks

Thomas E. Berghage

Risk Associated With Investing

     Any investment in securities involves an inherent risk.  Even the most stellar of companies have risk.  History is filled with stories of big name companies taken down by rogue employees, dishonest management, product liability suites, or natural disasters.  Enron, Global Crossing, and Arthur Andersen are the most recent examples. There are in addition other forms of risk in the market, risks that are subtler and less apparent, but equally devastating.  Poor management decisions, untimely corporate strategic moves and inept business plan execution can all add up to disaster for the unsuspecting investor.   

Risk for professional money managers is expressed in terms of the standard deviation (statistical measure of dispersion) of investment returns.  Individual investors, however, view risk in terms of the probability of losing their investment principal or the probability of actually making money on an investment.  The professionals could derive such probabilities by using their statistical measures and normal distribution statistics, but they traditionally do not do so because they use their measures in other statistical operations associated with portfolio structure.  This failure to convert the statistical measure is, however, unfortunate because it would communicate the level of risk much better than current techniques.  I think investors, and even some professional money managers would be surprised by the magnitude of the probability of loss if such a conversion were routinely performed.   

The risk that most people focus on is the loss of part or all of their principal, but of equal importance is the loss of purchasing power by not having your investment keep up with inflation.  These risks are not equally distributed across the market.  Some sectors of the market involve greater risk than others, and the degree of risk associated with different sectors varies over time. To put investment risk and the analysis problem into some perspective lets look at some a priori probabilities of loss along with the potential for return.

     A Priori probabilities are the inherent probabilities associated with an investment asset class based on long historical experience.  There are lots of ways to slice and dice the market, but each of the resulting groups of assets has associated with it an inherent risk of loss.  If you remember your statistics 101 class the professor talked about having an urn filled with red and black balls and any sample of balls randomly drawn from the urn would have the same general statistical properties as the total population from which they were drawn.  In other words, if 75% of the balls in the urn were red, then any random sample would usually have more red balls than black.  The same principle holds true for randomly selected stocks taken from a given sector of the market.  If the sector is dominated by losing stocks your chances of having wining investments is inherently low.

The A Priori probabilities are basically the inherent risk that the investor is confronted with, by investing in a given area.  The probabilities are not fixed.  They change over time and are themselves a function of the economic conditions (Figure2-1).  If we look at them over an extended period of time however, they do tend to stabilize and change very little from one sample to another.  The probability of loss is not the whole story, however. If a long-term higher probability of loss is accompanied by a long-term higher mean return, the investment might make sense, depending on the individual’s risk tolerance.

Lets start by looking at the a priori probability of loss or inherent risk associated with some well known stock populations: The Ford Investor Service data base of 4,000 plus stocks, the Value Line data universe of 1,800 stocks, the stocks of the S&P500 index, and the thirty stocks of the Dow Jones Industrial Average.  The statistics shown in

Table 2-1 are based upon the stocks’ performance during the five-year period from 1996 through 2000.  This period included three very good years for the market and two years of weak market performance.   

Figure 2-1

Probability Of Loss On Random Selections

Held For One Year

Table 2-1

A Priori Probabilities of Loss For Various Sectors Of The Market Along With Potential Returns

Total Market = Total Ford Investor Service Database for years 1996-2000

Value Line = The 1,800 stocks covered by Value Line

S&P500 = The approximate 500 stocks included in the S&P500 Index

Dow Jones Ind. = The thirty stocks included in the DJIA

Probability of Loss = Loss of principal after holding the stock for 12 months

Mean Return = The arithmetic average of annual total return (appreciation plus dividends)

Median Return = The mid point of the distribution of annual total return(50% above and below this point)

Stand Deviation = measure of dispersion

     The first thing to note in Table 2-1 is that a random selection from the market(the 4000 plus stocks covered by the Ford Investor Service, Inc. database) that was held for twelve months during the five years from 1996 through 2000 was about as likely to lose money (48.4%) as make money.  If these figures had included the last two years they would have been even worse.  They could have been as high as 60 to 70 percent chance of loss. These odds are almost as bad as those on the roulette table in Las Vegas.  Now granted, in Vegas if your color does not come up, you lose everything, and that is generally not the case in the market.  On the roulette table you either double your money or lose it all, while in the stock market, companies infrequently go bankrupt and most losses are considerably less than 100% (Figure 2-2).  So the consequences of winning or losing in Los Vegas are considerably different, but frankly I was quite surprised by these odds of losing money on a randomly selected equity investment.

Figure 2-2

One-Year Total Return (Percentage) For Stocks

In The Ford Investor Service Database

During The Period 1996-2000 

Now, you can improve your odds of winning by confining your investment selections to some of the subpopulations of the market like the Value Line stock population, or the S&P500 stocks, or even the 30 stocks of the Dow Jones Industrial Average.  By using one of the subpopulations shown in Table 2-1 you have restricted your selections to companies with larger market capitalizations (market values, shares outstanding X share price, greater than $5 billion) and this is, as we will see here shortly, a significant factor associated with the probability of loss and potential gain.  

Risk Associated With Company Size 

 Over the last five years the larger market capitalization companies have done better than the smaller companies.  This is not always the case, but for any extended period of time the relationships shown in Table 2-2 are very likely to hold true.  Large, well-capitalized companies are less likely to lose you money, Enron being an exception. 

Table 2-2

A Priori Probabilities of Loss For Various Market Capitalizations

Total Market = Total Ford Investor Service Database for years 1996-2000

Large Cap = Market capitalization equal to or greater than $5 Billion

Mid Cap = Market capitalization between $1 and $5 Billion

Small Cap = Market capitalization between $500 Million and $1 Billion

Micro Cap = Market capitalization less than $500 Million

Probability of Loss = Loss of principal after holding the stock for 12 months

Mean Return = The arithmetic average of the annual total return (appreciation plus dividends)

Median Return = The mid point (50% above and below this point)

Standard Deviation = measure of dispersion

     One thing to take away from Table 2-2 is the fact that the analysts of large capitalization companies have a much easier job of producing positive returns.  The probabilities are stacked in their favor.  If you can’t produce a positive return when there is less than a 37.7% chance of losing money by randomly selecting stocks from the sector, you have a problem. Producing positive returns in micro cap stocks, however, is a much more difficult job.  Selecting stocks from a population where almost 65% lose money is a challenging job.  To invest in this sector of the market you better have some good analysis techniques, because relying on chance is not going to help you.   

If you don’t like the odds associated with the random stock selections you have two choices: (a) develop a security analysis system that works, and reduces the odds of loss associated with random selections, or (b) put your money in, supposedly risk free 90-Day Treasury Bills.  Of course the returns you will receive on 90-Day T-Bills do not come close to the average returns for small, medium, or large cap stocks and may not even keep up with inflation thus exposing you to the risk of losing future purchasing power. 

Probability Of Exceeding The Return Of 5%

     Now assuming you can get a 5% return, what are the probabilities that you can exceed this return in the various sectors of the market (Table 2-3)?  This idea of exceeding a 5% return raises the performance bar for the analysis process and makes it a little more difficult.  If your security analysis process cannot produce better than a 5% return, then you need to reexamine the process or find a different avocation.

Table 2-3

Probability That A Random Selection From

Various Market Sectors Will Exceed A 5% Annual Return

* Definitions are the same as Table 2-2

Risk Associated With Investment Style  

     Another way of looking at inherent risk or a priori probability of loss is by viewing the probabilities in terms of Investment Style or Industry Sector.  Both of these categories are used by money managers to justify their investment strategy so it is worth looking at the probability of loss associated with each category.  For Investment Style I am using a structure similar to that used by Moringstar.  They divide the market into nine styles boxes.  The matrix is made up of large, mid, and small cap stocks that are considered to be value, blend, or growth oriented.  The probability of loss and the mean returns associated with each of these styles is shown in Table 2-4.

Table 2-4

Risk Associated With Various Investment Styles

      Large Cap = Companies with a market capitalization greater

than $5 Billion

      Mid Cap = Companies with a market capitalization between

$1 and $5 Billion

Small Cap = Companies with a market capitalization between

$500 Million and $1 Billion

Value = A Ford Investor Service (FIS) Price/Value Ratio Relative To The Market less

than .565

Blend = A FIS Price/Value Ratio Relative To The Market

between .565 and 1.435

Growth = A FIS Price/Value Ratio Relative To The Market

greater than 1.435

 You must understand that these figures are for the five-year period from 1996 through 2000 and during most of that period value stocks were out of vogue and the dot com growth companies dominated.  If you were to take a longer period of time or a different period in history the figures could be quite different.  The figures shown here are probably more indicative of the environment in which we have lived during the past seven years.  A more representative figure could probably be calculated by taking the performance of stocks over the last ten years and weighting their representation by how old the data is. Information that is ten years old should carry less weight than the more recent data.  We will leave the exercise of calculating representative a priori probability weights to an MBA student that needs a class project.  The figures presented here are just an indication of what needs to be accomplished. 

Risk Associated With Industry Sector 

The final category of inherent investment risk that we want to look at here is that associated with the various industry sectors of the market.  For this we will review the probability of loss and mean return figures for the fifteen major industry sectors used by Ford Investor Services, Inc.  Again the figures are for the five-year period, 1996 through the year 2000 and could probably benefit from an evaluation over a longer period of time.  Despite the shortcomings of only using five years, it is interesting to see the inherent risk facing investors in various market sectors over that period. 

Table 2-5

Risk Associated With Various Industry Sectors*

          * Market Sectors are defined in Appendix A

Risk Associated With Various Holding Periods 

     Lets look at a one-year time horizon for these a priori probabilities.  We have selected a one year time horizon because we consider shorter time frames to be more speculation than investing, but we understand the interest in shorter time horizons so we have provided the a prior probabilities of loss for one, three, and six month holding periods to satisfy the curiosity of those more inclined to trade stocks than hold them.

Table 2-6

Risk Associated With Market Capitalization And

Various Hold Periods

Total Market = Total Ford Investor Service Database for years 1994-2001

Large Cap = Market capitalization equal to or greater than $5 Billion

Mid Cap = Market capitalization between $1 and $5 Billion

Small Cap = Market capitalization between $500 Million and $1 Billion

Micro Cap = Market capitalization less than $500 Million

P_loss = Probability of Loss, Loss of principal after holding the stock for 12 months

Mean = Mean Return, The arithmetic average return

     This table is a little busy so to put this information in perspective we have plotted it in a two dimensional graph. The lowest probability of loss is for Large Cap Stocks held for 12 months or longer.  The trends are pretty uniform except for the Micro Cap Stocks (stocks with market capitalization less than $500 million), which run contrary to the general trend of increased probability of loss associate with shorter time frames.  For Micro Cap Stocks the probability of loss increases with the hold-time. 

     The other important trend to take away from this Figure is that the risk of loss decreases across all time frames as market capitalization increases. Large Market Capitalization Stocks have produced better returns at lower risk over the past eight years and will probably continue to do so in the future.  These large companies may not be as sexy as some of the small new-starts, but they produce better than average returns.

Figure 2-3

Risk Associated With Market Capitalization And

Various Hold Periods

     Now that we have some inherent risk benchmarks to evaluate our performance against lets look at what analysts have used in an attempt to improve upon these a priori probabilities and produce positive returns for their clients. 

Return To Home Page

Return To Article Index