How to Pick Winning Stocks

How to Pick Winning Stocks

This article explains how you can implement a successful analytical method for more consistently picking out winners in the stock market. If you currently pick your own individual stocks for investment, then I believe that my suggested methodology will likely increase your investment returns while reducing risks. You will also become much more comfortable with your investment portfolio since you will understand exactly why you own each company and will know how its stock price relates to your estimate of its actual value per share.

Does this system work?

Absolutely, it has worked for me, here is a summary of my Performance since I first began applying this method in June 1999 through December 31, 2006.

The cumulative result of investing $10,000 in the strong buys at the start of 2000, and switching to the new strong buys at the start of 2001 and again at the start of 2002 etc is that the $10,000 would have grown to $52,033 for over a 400% cumulative gain! This compares to a gain of about only 53% on the TSX Composite over that period. My own money was not invested strictly in the Strong Buys and each $10,000 that I invested grow to $30,679 for a gain of 207% or almost four times the 53% gain on the market index.


The system that I use and recommend relies on many of the same fundamental principals that are used by legendary investor Warren Buffett. The Canadian Shareowner Association and the National Association of Investment Clubs (U.S.) also use somewhat similar systems. The system is also grounded in fundamental finance theory.

What is the system?

The system is to buy Great companies at bargain prices. That is, buy shares that are selling at a price significantly below their estimated “true” value AND which have certain vital characteristics including, most importantly, ethical rational management and a competitive advantage in the marketplace.
The “trick” of course is to figure out how to confidently estimate the “true” value of the shares as opposed to the market price. It turns out that the instructions for doing this have been laid out by Benjamin Graham, Warren Buffett, certain advanced finance texts and a few other sources. Although proven to work, this method remains amazingly under utilized for several reasons. It contradicts the efficient market hypothesis, it requires a fair amount of diligence, it requires patience and very few companies truly pass muster with this method which (to my understanding) leaves most stock analysts unable to use the method since it would reject buying the shares of most of the clients of investment banks. Therefore this method can be the little used secret weapon for independent investors.

The “true” value of a share can be easily calculated given the following inputs:

  • the starting “normalized” or sustainable earnings per share
  • the assumed growth in earnings per share and in the dividend
  • the initial dividend pay-out ratio
  • the assumed steady state P/E ratio at which the shares will sell in (say) ten years
  • an assumed discount rate (minimum acceptable rate of return)

The first box below describes in detail how this calculation of the intrinsic value per share can be made.
The process of finding an initial adjusted earnings figure is described in detail in the second box below.
After completing the steps in the boxes we have a conservative and a more optimistic estimate of the value per share. Note that the optimistic estimate should not be wildly optimistic, we want to leave some room for the company to out-perform our projection. This intrinsic value per share provides an estimate of the value that we might be willing to pay per share. We would only buy the shares if the current price as below our estimate of intrinsic value. Ideally, it would be below even our more conservative estimate of value.

We also ideally want to buy companies that are not only bargains, but that are also great quality companies in other ways.

The following are Vital Qualitative Characteristics to analyse:

  • Industry attractiveness, intensity of competition
  • future outlook
  • quality, integrity and rationality of the management.

The industry attractiveness and intensity of competition can be judged by looking at the five forces of competition identified by Michael Porter of Harvard University. He believes that an attractive industry is one that does not suffer from intensive price competition. The five factors that determine the level of price competition are:
Barriers to entry – if new competitors can easily come in, there will be more intense competition. Such things as patents, special knowledge, product differentiation and brand loyalty create barriers to entry. For commodity products there are no barriers to entry.
Powerful Suppliers – Such as strong unions or other strong suppliers can effectively compete with a company and usurp all the available profit in the industry.
Powerful Customers – Sometimes there are only a few large customers and they may control access to the ultimate consumer and can usurp most of the profit for themselves. Wal-Mart could be an example.

Substitute Products – Some products have other substitutes which limits a company’s ability to extract high profits.
Competitor Response – Some industries simply seem to fall into a habit of being more interested in getting market share than in making profits. This occurs most often where fixed costs are high and where competitors for one reason or another fail to act together to keep prices higher.
The quality of management can be judged to a degree by the past history of the company and to an extent by the level of candor and disclosure in the annual report. I consider grossly excessive executive competition to be a strong negative indicator in this regard.

A number of other factors can also be looked at to provide further comfort.
The book value per share should be calculated and compared to the share price. Usually shares do sell for well above book value but an extreme value is cause for caution. For example a share selling at ten times its book value may signal a stock that is possibly significantly over-valued. In looking at book value it is important to analyse the nature of the assets in which the equity is invested. If the assets are largely financial in nature, then the book value is more reliable. If the assets are represented by specialized industrial equipment then you should place less reliance on book value. If the assets consist primarily of purchased goodwill or of capitalized development or exploration costs then I recommend placing little or no reliance at all on the book value.
The Balance Sheet should be examined to see if the company has enough equity and cash to continue in operation. This is very important for emerging technology, bio-technology or other companies that have not yet reached a profitability stage. If they are about to run out of cash then bankruptcy becomes a concern notwithstanding that their technology may have great potential.
In attempting to identify Great companies available at bargain prices, Warren Buffett reportedly uses a screen consisting of a number of tenets. My understanding is that he requires the company to pass each and every one of his tenets or he does not invest. In brief his tenets are:
Simplicity – It must be easy to understand the product or service of the company and how it makes money.
History of profit – Profits must be historic and proven not just potential
Strong outlook – There must be reason to believe that the company can earn above average rate of return. This requires a competitive advantage, barriers to entry, limited price competition. This effectively rules out commodity producers.
Ethical Rational Management – Management must be acting rationaly and in shareholder interest and be trustworthy.
Strong Return on Equity – He will not assume, much less pay for, high future returns unless the company is already achieving this. He believes that a high ROE is the key to retaining earnings and growing shareholder wealth.
High Profit on Sales – It is possible to achieve high ROEs with low profits on sales (given enough volume) but Warren requires high profits on sales.
Low Debt Levels – Higher debt levels create a leverage that can lead to higher profits but Warren requires high profits without excessive debt levels.
Selling at a discount – Warren will not buy a share that is selling at a price that is above his calculation of its intrinsic value. (But interestingly, he will continue to hold such shares even though that seems inconsistent.
As discussed in the opening paragraphs of this Article, after over five years years of using and refining this method, I have found that it is remarkably adept at predicting winners. In the past, I have not required my Strong Buys to pass all of Buffett’s tenet’s but more recently I am moving more in that direction, in most cases.
A consistent application of this methodology requires an investor to set up a spreadsheet to enter income and balance sheet data from each company and to calculate key ratios and the intrinsic value per share after first adjusting the net income as discussed in the box. It also requires consideration of the outlook for the company based on the quality of its management, its strategies and its environment.

In conclusion I urge those investors who are interested in picking individual stocks for investment to use this method, which will allow you to invest with much more confidence and will likely lead to much more consistent results. Alternatively, investors can seek out those analysts such as myself who apply this type of fundamental analysis.

Shawn Allen, Editor, June 2, 2001, with updates to December 31, 2006

Calculating the Present Value per Share of all Future Earnings

The prospect of calculating the present value per share of all future earnings sounds daunting or even impossible. The calculation is actually quite easy (that is, for those who are familiar with time value of money calculations) and while it is subject to error useful approximations can be made, at least for some stocks. Essentially, the process consists of forecasting the earnings level and growth for the next ten years. After ten years it is assumed that the stock will be sold at a conservative Price / Earnings (“P/E”) ratio such as 12 or 15. Cutting the analysis off at ten years nicely eliminates the impossible task of predicting earnings into the infinite future. Using this method, the present value of the stock is then the present value of the calculated amount that that the stock will be sold for in ten years plus the present value of any dividends that will be received.

For example, if a stock presently earns $2.00 per share and earnings are expected to grow at an average of 15% per year then in ten years time the earnings are expected to be $2.00 times (1.15)10 . This equals $2.00 x 4.0456 = $8.09. So the company is expected to be earning $8.09 per share in ten years time. If we assume that we can sell the stock at a conservative P/E ratio of say 15 in ten years, then we will receive 15 x $8.09 = $121.35 for the share at that time.

Using a 10% required rate of return, $121.35 to be received in 10 years is worth $121.35 / (1.10)10= $121.35 / 2.594 = $46.78. That is, if you pay $46.78 for the share today and sell it for $121.35 in ten years, you will have earned 10% compounded annually on your money.

So, if you expect that the earnings will grow at 15% then you now know the stock is worth about $47, If the stock is selling in the market at substantially less than $47 then you would expect to earn more than 10% by investing at that lower price. Conversely, if the stock were trading in the market at more than $47 than you would conclude that the stock was not a good investment since you would expect to earn less than a 10% annual return.

If the above stock pays a dividend then you add the present value of the dividend stream to the $46.78. For example, if the prior year’s dividend was $1.00 per share and the dividend is expected to grow at 15% per year then the present value of that amount can be calculated as $1.15/1.10 + ($1.00×1.152)/(1.10)2 + ($1.00×1.153)/(1.10)3 …  + (1.00×1.1510)/(1.10)10 = $12.87. In that case the calculated present or intrinsic value of the share is $12.87 + $46.78 = $59.65.

To account for the fact that some companies are riskier than others, investors can assume a lower growth rate (to be on the safe side) or can use a higher discount (required return) rate in the present value calculation. I believe that investors rarely make these kind of calculations. Furthermore, most advisors and brokers do not make these kinds of calculations. Therefore those investors who are capable of and willing to make such calculations should gain an advantage over the majority of investors who are not making or relying on those calculations.

The intrinsic value calculation requires just 4 basic inputs. 1. The initial earnings and dividend per share; 2.The expected earnings per share growth rate; 3. The required return or “discount” interest rate; 4. The P/E ratio at which the stock is assumed to be sold after ten years. More information on setting each of the four inputs is provided below.

The initial earnings per share is simply the current adjusted earnings divided by the diluted weighted average number of shares outstanding. Average diluted shares are higher than the actual average number of shares outstanding to account for the net dilution from stock options and convertible bonds. The diluted average number of shares can be determined from information provided on the income statement. The calculation should be based on adjusted earnings, rather than actual earnings, as discussed in the other box.

Calculating an expected growth rate for earnings is very challenging. But, an investor can consider the nature of the business and the competitive environment of each industry to make a more informed judgment regarding earnings growth.

In predicting the earnings growth it is very useful to calculate and graph past earnings growth rates. Since earnings can often be impacted by a variety of one-time gains and losses you should also graph the adjusted or normalized earnings per share to reveal the under-lying growth trend.

If the historic earnings and sales graph shows a very erratic pattern, as is often the case, then it may not be possible to forecast future earnings. (In which case you really can’t apply this intrinsic value method to judge the true value of that stock.) If however, the company has shown a reasonably consistent pattern of earnings then an initial assumption might be that the historic growth rate will continue. However, it is advisable to also consider the environment in which the company operates in order to attempt to discover if it is likely that the company’s past trend is no longer applicable. The current state of the economy is also a factor, although we should remember that this is a ten year projection and we should not be overly influenced by the current state of the economy, since that will change.

The return on equity can provide an estimate of the sustainable growth rate. In theory, the sustainable growth rate is equal the return on equity rate multiplied by the percentage of earnings that are retained (not paid out as dividends). In this case, it is important to use the adjusted or normalized earnings to calculate the return on equity and to consider whether or not the return on equity level is sustainable for the next ten years.

In some cases management provides an outlook for the longer term growth in earnings per share.

Since the growth rate over the next ten years is inherently uncertain, the calculation should be done twice, once with a relatively conservative growth rate and once with a more optimistic growth rate. It is important to avoid the use of a wildly optimistic growth rate since doing so would result in a calculated intrinsic value per share that assumes that the wildly optimistic growth will occur. If you paid a price that assumed all that wildly optimistic growth would occur, you would essentially have a stock that had little or no up-side left and plenty of down-side risk.

The required rate of return can be set at a standard level. It consists of the return available from risk free bonds plus some risk premium. In theory the risk premium could change with every company. In practice I estimate the return on bonds as about 6% and add 4% for a risk premium to arrive at a 10% required return or discount rate. I then adjust for risk by being more or less conservative in my growth assumptions in step 2 above.

Finally, estimate the P/E ratio at which the stock can be sold after the ten year holding period. Today’s high growth, high P/E stock cannot be assumed to grow rapidly forever. To be conservative, assume that ten years out the growth rate will have slowed to a more normal level and that therefore the stock will be selling at a relatively normal P/E between 12 and 25. In most cases you should assume a P/E of about 12 for a conservative calculation. Repeat the calculation with a P/E of about 15 for a more optimistic scenario. In unusual cases you could assume a P/E as low as about 8 or as high as 20. If you assume a P/E that is too high, you leave yourself with little up-side risk and plenty of down-side risk.

So after all of that, for the more predictable companies, you will have arrived at a conservative and a more optimistic calculation of the present or intrinsic value per share. If the earnings do not seem sufficiently predictable then it is appropriate to decide not to make the calculation or alternatively to assume a very conservative growth rate.

A variation of the above method is to use an assumed 5 year holding period rather than 10 years.

Analysis of Adjusted Net Income

Net income is the most important consideration in valuing a stock. There is often a tendency to trust and use the reported net income as is. In reality you usually have to correct for any unusual gains and losses that distorted the net income. The following are items that often may distort net income.bulletOne-time gains and losses – A company may have a substantial gain or loss on the sale of land or equipment. Or a company may take a one-time charge for restructuring or some very unusual item such as a strike or other disruption. These items should be reversed in attempting to calculate the future income since they are not expected to recur.

bulletUnusual income tax rate – Some companies pay very little in income tax expenses due to past tax losses. This is a temporary phenomenon and should be adjusted for. Calculate the effective tax rate of each company and then investigate the reasons if the tax rate is much different than the expected statutory rate (about 45%) for large companies.

bulletCapitalized exploration expenses – Mining and oil exploration companies are allowed to capitalize exploration costs even for “dry holes”. For this reason, it may be appropriate to be very skeptical of the net income of such companies. In fact such companies typically do not lend themselves to this type of analysis since their earnings are often inherently unpredictable.

bulletR & D expense – Under U.S. GAAP all R & D costs are expensed while in Canada all research costs are expensed while development costs are capitalized. In either case the result of this conservative accounting can be an “artificial” lowering of net income. If it is believed that the research will create future value then it seems appropriate to add back at least a portion of the expensed R&D or recalculate income as if R&D were amortized over say 5 years.

bulletDepreciation – Consideration should be given to the level of the depreciation expense versus the likely true deterioration (or possibly appreciation) in value of the fixed assets or the cost to eventually replace an asset due to its usage or the passage of time. In rare cases such as with a portfolio of newer buildings, it may be appropriate to add back a portion of the depreciation expense on the basis that the actual replacement of the building will occur many many years in the future and the cash outlay at that time (even with inflation) has a present value that is less than the depreciation expense. And it is appropriate to add back amortization of goodwill since goodwill is not usually an asset that is used up over time and it often appreciates in value.

bulletDeferred Income Taxes – These have to be paid eventually only after some years. The present value of this future outlay is less than the non-cash expense deducted under GAAP. Therefore, it is appropriate to add back some portion of this expense. I would be conservative with this and add back say 20% of the deferred tax expense.

bulletPreferred Share dividends – Technically net income is properly reported prior to any preferred share dividends. But most companies also show the deduction to arrive at net income applicable to common shares. For those companies that do not show the deduction, you must make that deduction.

In summary every income statement has to be taken with a grain or two of salt. With some digging, you can make an informed adjustment to the published net income to calculate a normalized or adjusted net income.

Conveniently, some managements provide an adjusted earnings figure. This goes beyond a disclosure of extraordinary items, which is required by GAAP, and includes a variety of unusual items. You should focus on that number when it is provided. But also check for any unusual items that management did not highlight.