Valuation Metrics

Valuation metrics are the financial ratios that compare a stock (or market total) price to earnings, book-value, sales, dividends, cash flow, or any other metric. They all measure the cheapness of the stock or market. They are used to answer the questions … “What should this stock’s price be?” or “Is the market over (under) priced?” or “Which stocks are more cheap (expensive) than the others?” or “Will valuation multiples expand or contract going forward?”.

Novice investors start and end their analysis with these metrics. It is unlikely they go beyond the P/E and Dividend Yield, coupled with someone’s assurance that “the dividend is safe”. Possibly they require the dividend to have increased over time. These metrics are on the first summary page of all websites offering company data. Little effort is required to find them. Analysts love to use cheapness to promote their ‘buy’ recommendations. Advisors repeat the refrain: “historical returns from cheap stocks have outperformed the market”. These metrics are the bedrock of Value Investing. These are the metrics most used when screening stocks.

But the logic of the markets warns against joining the bandwagon. How can you earn excess returns from a strategy everyone knows about, uses and promotes? Times change. The market’s attention switches from growth to value and back again. Popularity guarantees its own demise. Popularity is cyclical. While no one disputes that lower present prices lead to higher future returns, all else equal, and that higher present prices lead to lower future returns, all else equal — all else is never equal, and ‘lower’ and ‘higher’ that what?

………………………….

Backtesting results from screening for value metrics, or from portfolio weightings determined by value metrics usually come up with results similar to these from a MSCI April 2011 paper “Capturing the Value Premium”. It presents a 15 year comparison of the normal MSCI Index with its Value-weighted counterpart (weighted by book value, earnings, cash earnings and sales, not dividends), and also with portfolios screened for single metrics.

………………………….

None of the metrics tell you ANYTHING about the company itself. They do not distinguish between companies that are cheap for a reason, and stocks that are mispriced by the market. Do not think you can assume the label “Value Investor” just because you use these value metrics. Value Investors analyze the company itself to determine its worth. It is the analysis of the COMPANY that determines what valuation metric is appropriate for a specific stock – the company’s growth prospects, its stability, its leverage, etc.The implicit presumption of cheapness-advocates is that the reason for a stock’s disfavour is never a good reason; that companies always recover from bad times; that market sentiment always overshoots on the downside; that the economy always cycles back up; etc. But if you believe the markets are mostly efficient, then you must admit that few stocks are cheap because of market mispricing, or because no analyst covers the company, or because no one else sees the growth potential that you see. Most stocks are cheap because bad things are happening to the company. A paper by Kok, et al (2017) found that stocks with extreme pricing metrics are not because the price is wrong. It is the other factor in the ratio that is off. High P/Bk ratios are because the market knows the book value will be written down. High P/E ratios are because earnings are temporarily high.

Only hind-sight will tell whether the stock was a value-trap or a buying-opportunity – whether the company stagnates and dies, or recovers. The research showing excess returns from cheap stocks always measures large portfolios. It is usual that less than half the stocks making up the cheap-decile portfolio have returns that beat the market. It is the very high returns of just a few that impact the results. Your personal stock-picks are very unlikely to include those particular securities.

………………………….

Valuation matters most when all company profits are paid as dividends. This keeps the secondary market in line with the primary market. If your required rate of return is 8% then you cannot pay more than P/E 12.5 (= 1 divided by 8%) for the stock. E.g. a $12.50 stock with P/E 12.5 has earnings of $1 and a dividend yield of 8% (= 1/12.50). At the other extreme, valuation metrics need not have any effect on equity returns if those returns all come from price appreciation (capital gains). No matter what valuation you pay at the time of purchase, if that valuation is the same at the time of sale, then your return will always equal the company’s growth (which equals its incremental ROE).

But valuations do affect equity returns even when all the return comes from capital gains because valuation metrics mean-revert to long-running averages. Your rate of return is predicted to be the sum of the dividend yield, plus the company’s growth per share, plus the mean-reversion of the valuation metric. Your realized return differs because growth will differ and because the valuation metric will almost never equal its long-run average at the time you sell. In the end, valuation metrics are only capable of giving a crude idea of what future returns will be. They can’t arbitrate the debate between bulls and bears because valuation metrics don’t reliably mean-revert.

………………………….

A very interesting model is presented in three webpages from Philosophical Economics (in the Readings below). His idea is that any market or stock valuation is the byproduct of the same supply / demand forces that price any other product. He disbelieves there is any intrinsic value at which the asset should be priced. Rather, it is investors’ desires to hold certain asset allocations between stocks, bonds and cash that determine stock prices. He gives a compelling example how prices can change even without any actual trades.

Another of his ideas is that the seemingly-permanent increase in valuations for risky asset classes is a valid response to “improvements in the way a market functions … which lead to reductions in the costs of those who use it.” If, as most believe, the higher historical returns (lower valuations) from equity (vs debt) were required compensation for assuming the risks of price volatility, and since diversification that reduces idiosyncratic risk is now essentially free (passive ETFs), markets should no longer require all that compensation.

………………………….

Interesting Reading
Fundamental Analysis Redux – 2018 Richard Sloan
Market cycles determined by valuation – fact or illusion?
The 1st Philosophical Economics page, and the 2nd and 3rd.
Lecture material from W. Trainor
Thoughts on Valuation.
Stock returns from value investing
Ian Giddy “Methods of Corporate Valuation”

You should never consider a stock ‘cheap’ because it’s current price is lower than recent highs. The stock moves on sentiment AND changing facts. You cannot attribute its decline only to sentiment that will be proven wrong in the long run. A stock will not regain old highs just because it was there once before. History does not repeat itself. This warning applies to fast rising stocks as well. They are not ‘expensive’ just because they have doubled in price. Keep your focus forward.

There is a second common misconception of ‘cheapness’. Many people think that a $2 stock is cheaper than a $10 stock, simply because it is priced lower. It is best to show the error with an analogy:

You are walking down the street and see two signs advertising watermelon slices. One cost $1.00 the other $1.50. Which is cheaper? According to you, the $1.00 is cheaper because the dollar value of the price is lower. Does that make sense? Of course not, because you don’t know how big each slice is.

Lets say the signs say “Half watermelon $1.00” and “Quarter watermelon $0.75”. Which is cheaper: the half melon because it costs only $0.50 per quarter? Of course not, because you don’t now the size of the melon.

The same thing happens with stocks. One stock represents a portion of the whole company. But you cannot tell from the price how big the whole is. And you cannot tell from the price what %portion of the whole you are getting. So when a stock is called “cheap” no one is talking about the dollar of the share price

Holding a low PE stock is less risky because “it’s hard to fall out of a ditch”. When a company’s earnings fall, it is likely that both the earnings AND the earnings multiple (measuring sentiment) both fall. The effect is magnified. If you start from a lower position your downside is limited. The reverse argument is that low P/E metrics reflect the market’s low pricing because the company is in trouble. It is cheap ‘for a reason’. Its cheapness proves it is more risky by definition.

………………………….

Predictive Value: Sorting US stocks by P/E percentile shows that the lowest P/E class earned 16% on average, while the highest P/E class earned less than 9%. This is the kind of back-tested data mining that prompts the popularity of stock screening, with all its problems. But don’t let bar-charts fool you into thinking they promise any surety. See the discussion at Screening Stocks. That said this metric’s results are very consistant through time and between deciles. And it is common sense that earnings determine the value of a company and stock.

Does the P/E metric also work to predict index level returns over time? Charting the beginning of the year P/E by the following one-year’s total return gives you the next two charts. Very little of the resulting returns are explained by the P/E. The American R-squared is only 9 percent. The Canadian R-squared disappears to 0.3 percent.

P/E is not a good predictor of index-level returns because overall valuations remain essentially correct through market cycles while the metric is pro-cyclical. E.g. in the depths of a depression earnings fall much further than stock prices because investors understand the depression to be temporary. The P/E metric skyrockets as a result, wrongly indicating that the market is still overpriced. Notice the outliers at very high P/Es in the charts above when markets recover. The Canadian data shows less statistical relevance than the US data. Maybe this is because the index is heavily weighted in highly-cyclical resource stocks. There is a saying “Sell resource stocks when their P/E is low and buy when high”. Resource investors are very mindful to be counter-cyclical.

The following chart shows how big changes in earnings are always offset by reverse changes to P/E valuations because prices stay more stable. Trends are multi-year ups and downs.

………………………….

Add Cash : It is common to hear a stock price justified with the phrase “after you back out the cash.” This assumes that cash is not a productive asset – that earnings are a product of the other assets only. Any valuation of a stock would therefor use P/E to value the productive assets and then add the cash balance per share on top. But this is nonsense. For one thing, month end cash balances are transient. They can disappear with the next week’s cheque run. As well, if the business borrows $Million in debt and holds the proceeds as cash, does that cash make the company worth more? No.

………………………….

The ‘correct’ P/E is personal to the company itself. It is best judged by its historical value. The market seems to ‘assign’ a P/E to individual companies, that takes a long time to change. Deviations from the norm cause investors to ask “is this an opportunity or a value trap?” The following 10-year chart of General Electric shows how slowly and painfully the market changes its idea of the proper P/E. Even as the earnings have doubled, the dropping P/E has resulted in the stock’s decline.

………………………….

The earnings yield is the P/E flipped upside down. This yield is not really comparable to yields you earn on fixed income securities, even though you frequently see them compared. The graph below comes from this spreadsheet’s Sheet 2. There is of course the difference in risk, but also growth. Because they are compared some investors are led to believe that the equity owners’ expected return can be estimated as the sum of the earnings yield plus earnings growth. This is not so. The basic formula for the investors’ expected returns is the sum of the dividend yield plus earnings growth.

………………………….

P/E (ttm) : The (ttm) in the heading refers to ‘trailing twelve months’. The EPS in this metric are the historical reported accounting earnings from the most recent 12 months. This is the measurement used to develop all the historical averages. Be aware that what is quoted in the media today is probably NOT this metric. More probably the metric used today anticipates future earnings adjusted for many items.

The Diluted EPS metric may be used in the trailing P/E. The accountants have tried to factor into the calculation the dilution events just waiting to happen, but do not assume they covers off all possibilities. The wrong assumption that it correctly measures the effects of outstanding stock options is discussed in depth on the Understand Equity page. Another problem not incorporated in the dilution results from debt that management can pay off with common shares instead of cash.

Most often this type of clause is found in preferred shares or traded debentures, and the common share owner rarely reads the prospectus of this debt. Management is usually forced into this dilution exactly when it is most harmful – when the company is in trouble and the share price is low. The number of new shares to be issued is usually the face value of the debt divided by roughly the stock’s price, so the lower the price, the larger the number of shares issued.

The forward P/E uses estimates of future earnings (see chart of estimate errors). The pitchmen will always be optimistic with their estimates. And the future they are talking about can be fully two years forward. Beware of double counting growth by using both higher future earnings PLUS a high multiple for growth. Forward earnings have already incorporated the future growth, so using a high multiple as well will double count. Don’t pay now, for the future value of the company. Pay only the value of the company today, so that its growth will be your gain.

In early 2008 all media stopped any references to the trailing P/E of major indexes. Complicit in promoting stock investing, they all covered up the fact that the P/E (ttm) was ballooning above P/E = 22 at the beginning of the year, to P/E = 99. They replaced their quotes with forward P/E’s and claimed “the markets have lots of upside because the index PE is only 12 and historically PE=15 was the average.” This was a bald-face misrepresentation. The historical metric uses trailing EPS (ttm), while their metric used forward earnings estimates. That situation has continued until into the 2011s. By that time the earnings had recovered enough to risk disclosure. Beware. The media often won’t tell you when they use forward earnings, or a normalized historical earnings (that leaves out all the bad expenses).

………………………….

PEG ratio. Morgan Stanley found that between 1986 and 1998 the lowest PEG stocks earned 2% over the market return. This was too short a time span to jump to conclusions, but there is evidence of continuing excess returns from a hedge portfolio of Low-minus-High PEG stocks.

Companies’ earnings by themselves have no value to investors. Their value lies in the presumption that they either get paid out as dividends or reinvested for growth. A company with higher growth is worth more ‘per $ of today’s earnings’ than a company with low growth. The PEG ratio is supposed to put a differential value on the EPS based on projected growth rates. Its calculation is the P/E multiple divided by the growth rate … Price divided by EPS divided by Growth … PEG. Supposedly the shares are correctly valued when PEG = 1, or when the P/E equals the growth rate.

But does this valuation actually work? If you had three companies, exactly the same in all ways except the expected rate of growth, would your investment returns be equal if you valued your purchase by PEG = 1?

The following table shows your annualized rate of return for each growth rate over a 5 year time span. The differences are huge. Paying PEG=1 with a presumption of high growth reduced your returns from 19 percent to 13 percent. PEG did NOT normalize the returns between different rates of growth.

Nor does it make any difference to change the assumed ending P/E multiple to 18 from 15. The differences between the returns is of the same nature. Historically P/E = 15 is about average.

It is only when you presume the growth rates continue for 10 years, that the returns get normalized by the PEG valuation. But should any investor expect growth rates higher than normal to continue for more than 5 years? Even 5 years is stretching it.

There are other issues with the PEG ratio.

• How is the expected growth calculated? Is it growth of revenue, or growth of Net Income, or growth of EPS? Only EPS growth is relevant, but no analysts use that metric. It is reasonable to use the same period (eg 5 years) of growth for all comparisons, but doing so will penalize companies with growth expected to last longer.
• What initial Earnings value is used, the (ttm) earning per share, or the estimated future earnings, or the normalized operating earnings? Using forward earnings will double count growth, but that is the normal metric used.
• Where does the growth rate estimate come from? Every website posts consensus estimates, but those have been shown to be overly optimistic.
• This metric generates more underpriced stocks when interest rates are high, and also more stocks from emerging markets (because their interest rates are higher). A high interest rate creates a high opportunity cost for stocks and lowers their price. The lower price (even while growth rates are the same) creates a lower PEG. E.g. “In 1981, when treasury bond rates hit 12%, more than 65% of firms traded below PEG = 1. In 1991, when rates had dropped to about 8%, the percent of stock trading below PEG = 1 also dropped to about 45%. By the end of the nineties, with the treasury bond rate dropping to 5%, the percent of stocks trading below PEG = 1 had dropped to about 25%.” (Reference pg 35). The metric is most relevant when used as a comparative metric between stocks. There is no ‘correct’ value, especially not ‘= 1’. Economic realities will impact what PEG is acceptable, same as they impact PE – the risk in the economy, interest rates, inflation, etc.
• PEG ignores differences in stocks’ price volatility. If you sort stocks by price volatility there is a direct correlation between high risk and low PEG.
• The ratio implies a linear relationship between growth rates and P/E that does not exist. E.g.The historical average Canadian P/E = 15.4 and earnings growth = 6.5% (see the spreadsheet link above). That gives a long-run average PEG = 2.4 (not 1). In the example above, the company with 10 percent growth was initially priced at P/E =10, if PEG = 1. Very few companies trade have traded at that low multiple for many decades.
• Growth is only part of the return to investors. Dividends are the other part. For any given earnings$$, the higher the dividend yield, the lower the necessary growth for a valid investment, so the higher the acceptable PEG.

………………………….

Shiller’s P/E : Shiller has gone to great lengths of data mining. He eventually found an explanatory pattern between a) the ratio of the index’s price to its 10-year-historical-average earnings (PE10), and b) the future-10-year’s real returns. The explanatory powers of PE10 increase as you lengthen the subsequent period to 20 years. But PE10 has little explanatory power for shorter 5-year returns and no one in real life makes investing decisions based on a promise that take 10 years to materialize.

It seems clear from his results that using averaged historical earnings is useful in valuations. And that buying at really high P/E’s is a bad idea, and really low P/Es is good. Most of the time (P/E 12 to 22) returns were widely variable. Pretty much what common sense would tell you. Pretty much what people do anyway.

The power of the PE10 to explain real-returns varies over time, and between markets. Klement (2012) studies 35 countries’ PE10 over the previous recent history for which data is available and found widely varying explanatory powers. In Canada since 1965 PE10 has explained only 11 percent of returns, while Ireland’s PE10 since 2000 has explained 90 percent.

This may be because different markets show different cyclicality of returns – US stock returns are very cyclical, while Canadian stock returns are not. The metric predicts the next phase of the cycle – so there must be a cycle for it to work. The problem is … how can anyone know in advance whether the market they are looking at will be cyclical in the future, or not? Just because it was cyclical in the past?

But PE10 may have a place in retirement planning. The portfolio returns in the first 10 years after retirement have a huge impact on your probability of eventually running out of money. If the returns for this period are predictable then the appropriate withdrawal rate can be set to prevent failure. Also consider the inability of anyone to know whether they are saving enough money 10 or 20 years before retirement. The PE10 gives some estimate of the very-long-term returns that can be expected.

………………………….

Ben Graham’s P/E. Graham created a value metric for stocks that gets multiplied by EPS just like a P/E metric. His P/E equation was

(8.5 + (2 * growth rate)) * (4.4 / AAA corp bond yield)

It considers both the company and the business cycle. By including the trailing ratio he allows for higher valuations when interest rates are low and vice versa. Since most all business must grow at least as fast as inflation, the first factor sets a floor for the P/E equal to about 12.5 when inflation = 2% and interest rates are 4.4%.

Like all other P/E metrics this valuation ignores dividends. It looks only at earnings and growth. It assumes that earnings NOT generating growth will be paid out as dividends.

A low P/Bk has been shown to be a very reliable indicator of value, correlating with above average investment returns. The reason is debated. It may be because when assets are not being used to their full productive capacity, new management is brought in to sweat the assets. It may be be that temporary economic factors disproportionately affect some industries more than others. It may be that investors are mislead by the different results from traditional accounting rules …. expanding balance sheets for companies that grow by acquisition, compared to those that expense the costs of salesmen. It may be because investors over-react to changes in the balance sheet and place less emphasis on its size.

It may be that insufficient assets to backstop debts make the company too risky, unless investors force down values enough to produce higher expected returns. When a business buys additional productive assets using debt financing, the net book value does not increase. When life is good that leverage pays off with excess earnings. Stock valuations based on earnings will increase due to the increase in earnings. The resulting P/Bk will rise, even while the P/E is stable. This was the situation with banks from 2003-2008. Investors can evaluate what is causing the rise in P/Bk by looking at the Return On Invested Capital metric.

Interesting Reading
Book Values Greater Than Market Values – 2019 Mary Barth, et al.

………………………….

The metric is intrinsically forward looking, based on the most recent valuation of assets. Compare it to the P/E (ttm) ratio which measures performance during the past year (or only a guess about the future). E.g if a company issued additional shares (at a price greater than book value) just prior to the last quarter’s end, none of the premium’s benefit would be showing in earnings. But the P/Bk metric would have fallen indicating a purchase opportunity.

Investors choose between using P/E and P/Bk depending on the industry. Tech companies’ major assets are their employees – who don’t show up on any Balance Sheet. No one looks at their P/Bk ratios. But banks and life insurance companies hold financial assets which are marked-to-market each Balance Sheet date. Their income is mostly portfolio returns – dependent on their assets. So most investors prefer to value banks with P/Bk.

P/Bk generates a metric that is a lot smoother than P/E over time. This is because the book value is much larger than earnings. A dismal year may cut earnings in half and double the P/E, but the effect of that lower growth in assets is only a small percent change in book value. P/Bk is less subjective than the alternative ‘forward P/E’ or ‘P/E based on normalized earnings’, when the company hits turbulence.

There is a mathematical relationship between P/E and P/Bk, moderated by Return on Equity (ROE). If you take out the Price from each ratio you are left with EPS and Book value. On a ‘per share’ basis ROE is just the EPS divided by the Book value per share. If you know any two of the three metrics (P/E, P/Bk, ROE) you can calculate the other. E.g. If a company is priced at P/E = 16 and its ROE = 12.5 percent, then its P/Bk = 2. ROE divided by the P/Bk equals the earnings yield (flip of P/E).

………………………….

Theoretically P/Bk can be used to judge the risk of not recovering your investment if the company liquidates. In reality the liquidation value of assets (even financial assets) is very different from their historical cost, or their market value, or the present value of their productive use. Common stock investors should simply ‘not buy’ companies risking liquidation. The probabilities are high there will be zero value leftover for them.

Since equity owners are only paid from the bottom line (profits), they will pay more for sales (Price/Sales) when the company has a high net margin (net income as percent of sales). But using net margin as a metric for stock screening has been shown to not be useful, while using price/sales HAS predictive value. That may be because competition within industries enforces a certain net profitability – or the boss will be fired. When the company’s growth shows increasing sales$$, it is only time before the net profits will also appear. This reasoning would explain why gross margin (sales less cost of sales, as percent of sales) is usefull for screening

A major problem with Price/Sales (and also Price/Cashflow) is that it is never adjusted for any minority ownership of subsidiaries. When the results of a subsidiaries are consolidated in full, but owned by their own set of shareholders, not all the consolidated sales will be ‘your’ sales, just like not all the net income is ‘your’ net income. While the Income Statement has a specific line item to remove the net income owned by the subsidiary’s shareholders, there is never any adjustment to the revenue line.

Net Income as a % of Sales (Net Margin) is a good way to compare the efficiencies of two companies in the same industry. Some industries are low margin (e.g. food retailers) but compensate with high volumes of transactions. Only a small misstep will drag them into loss, so they are risky. But it is probable that high infrastructure capital costs prevents competitors. Other industries are high margin (e.g. consulting firms). Their extra padding in margin, though, is frequently offset by a greater volatility in revenues. That same profit margin encourages competition.

The dividend yield measures the future year’s dividend payments as a percent of the current share price. The payments’ estimate comes from management’s stated policy. It does not include special dividends or capital gains distributions. In some companies the distributions are so variable that the computer systems tracking the metric have a hard time distinguishing between normal and ‘special’ dividends. If the metric seem weird, investigate further.

The dividend yield, by itself, is not a basis for stock valuation. Most everyone understands that –

• Returns come from a combination of income and capital gains.
• There is a trade off between the two. A $1 dividend creates a $1 capital loss.
• The markets, for the most part, equate (eg) 9% total returns, whether from 3% dividends plus 6% capital gains, or from 6% dividends plus 3% capital gains.

One of the earliest stock valuation models was the Gordon Discounted Dividend Model. It uses $dividends, not dividend yields. Users believe that the only value to investors from shares is the actual cash distributions they receive from the company. The model ignores cashflow and profits that are reinvested. In their place it integrates a projected growth rate for the dividends.

This model effectively sets a price for the stock so that the sum of its dividend yield plus its growth rate equals the investor’s required total return. The math of the equation may confuse you, but see how this works with different input variables in this Discounted Dividend spreadsheet. The sum of the yield plus growth rate always equals the total return.

The growth rate in the equation is that for the dividends paid, but when talking about an unknown future, the dividend growth is just a proxy for capital gains. Both derive from the growth of earnings. The stock price won’t move in lock-step with earnings because of changing market P/E valuations. And the dividend won’t move in lock-step with earnings because of changing payout ratios. But at time=0 you cannot predict those variances. The only logical assumption to make is that the growth in dividends, earnings and stock price will be equal.

No one really uses models like this. At most, people will use them to reverse engineer a stock price to see whether the input assumptions are reasonable. Try changing the projected growth rate from 7% to 6% in the Discounted Dividend spreadsheet leaving the other inputs as shown above. No investor can predict to the exactitude of 1% what the long-term growth rate of a stock will be. The stock price drops 25% from $83.33 to $62.50. So the calculated value is really “… give or take 25%”. You can compare the Gordon Model to Graham’s Formula, and to valuations using P/E and PEG ratios. The reconciling metric is the ROE of the reinvested earnings. In the diagram above you see how ROE assumptions vary widely according to the model.

………………………….

Share Buybacks theoretically impact investors the same as dividends. It is appropriate to include buybacks along with the dividends in valuation models. But there are issues.

• A large portion of stock buybacks only sop up new shares issued by the company to employees exercising stock options. Net/net there was no reduction in shares outstanding. So only the cash cost of net reductions should be capitalized as quasi-dividends.
• There is no market expectation, or management commitment, that the buy backs will continue. The market only finds out in hindsight that they have occurred. They should only be capitalized when expected to continue forever.
• The shares repurchased may be kept in treasury and reissued at a later date. These purchases would be better considered a financing maneuver, than a return to shareholders.
• The buybacks may NOT result in the theoretical increase in share price. Investors have no ability to monatize their ‘return’ by selling the shares.
• The growth rate used should not include the rate of growth produced mathematically by the reduced number of shares.

………………………….

High dividend yields are very supportive of stock prices during recessions and turmoil. Investors are more inclined to stick around if they are ‘paid to wait’ with dividends that telegraph management’s opinion of long-term profitability. During bull markets high-dividend stocks are likely to underperform, as re-investment opportunities abound. During high inflation high-dividend utilties may underperform because their pricing powers are limited.

………………………….

A clear distinction should be made between dividend payments that are a portion of profits, and distributions that exceed profits. This problem is most noticeable with income trusts that payout profits … as well as the non-cash expenses like depreciation. Any distribution exceeding profits is a return of capital (ROC). The tax designation of a payment does not prove its economic reality, but certainly if the company discloses their distributions as ROC, then chances are it is so.

Investors cannot make ANY valuation decisions based on these distribution yields. By definition they are NOT sustainable. In order to survive the business will eventually need financing to replace the cash distributed. They may do this by issuing shares, or increasing debt, or by not buying replacement fixed assets until the first two options have been effected.

The dividend true-believers promoting a dividend-growth strategy are firmly attached to the Yield On Cost (YOC) metric used by no one else. It is a calculation of dividend yield that leads long-term holders of a stock to believe they earn a higher yield than new purchasers. Its users are very resistant to learning that their calculations are a delusion. They claim their metric is meaningful because they disclose the calculation and their math is correct. An analogy would be the person who re-calibrates his bathroom scales below zero, then brags about his weight loss, and truly believes he has lost weight. His conviction holds even as he discloses his re-calibration.

………………………….

There are different ways to calculate the metric. The numerator is the same used in the normal ‘current dividend yield’ – the next 12 months of regular $$ dividend payments. It is the measurement of cost in the denominator that varies. In its basic form ‘cost’ is the historical price paid for the shares.

Example:

• You buy the stock for $20 with a 4% current yield paying $0.80 dividends.
• Years later the stock you still own now pays $1.20 in dividends.
• Your YOC is now $1.20 / $20 = 6%.

………………………….

A large YOC can be the result of four factors. There is no way to tell which factor has caused the resulting value.

1. The yield at purchase. A higher yield at purchase will create a higher YOC at a later date. Two stocks are purchased for the same $100, and held for the same 1 year, and grow their dividend at the same 5%. Stock A with an initial yield of 2% will have a YOC = 2.1% (=(2 * 1.05)/100). Stock B with an initial yield of 8% will have a YOC = 8.4% (=(8 * 1.05)/100).
2. The holding period of the security. The longer the time frame the higher the cumulative growth of the dividend. Two stocks are bought for the same $100, each with a yield of 2%, each growing their dividend at 5%. Stock A will have a YOC = 2.6% because bought 5 years ago. Stock B will have a YOC = 3.3% because bought 10 years ago.
3. The dividend’s growth rate. A faster growing numerator creates the larger YOC. There are two factors affecting the dividend’s growth rate.
   (a)   Management can simply declare a higher $dividend, creating a larger payout ratio. This does not reflect any performance of the business.
   (b)   Dividends that are set as a stable percentage of earnings will grow as earnings grow. In the long run earnings growth is necessary to sustain dividend growth.
4. Adjustments to $Cost in the denominator. The smaller the cost the larger the YOC. There are various excuses for reducing the $cost.(a.1)   When dividends are DRIP’d people reduce the $Cost by ignoring the cost of the additional shares purchased using the dividends received. Only the cost of the very first purchase of shares is used. Since dividends paid for the additional shares, they are considered to have been ‘free’.(a.2)   When dividends are not DRIP’d people reduce the $Cost by the value of all $dividends received subsequently over the years. In both these cases the value of dividends is being double counted – once as ‘income’ in the year received, and again as a ‘return of capital’. By analogy, that is the same as forgiving someone’s debt to you equally with every dollar of interest they pay you.(b)   A second variation on the calculation of YOC allows the dividend true-believer to escape the nasty effect of selling a stock and having start all over. The replacement stock now has a larger historical cost and the investor’s YOC has shrunk back to everyone else’s current yield. What to do? Pretend you never swapped stocks. Just continue using the historical cost of the stock now sold. Who cares that the basic premise of the metric is destroyed? The dividend true-believer does NOT want to celebrate capital gains.When YOC is applied to a portfolio in total it uses this same fudge. The portfolio’s value at some point in time is taken for the ‘cost’ and continues to be the ‘cost’ no matter that stocks are bought and sold within it. And all those savings added along the way? Well of course any additional dividends generated are included in the metric, but there is no need to adjust ‘cost’. Surely?

………………………….

So, what does YOC measure? Good question. The answer is ‘nothing’. Just defining the math does not answer the question. What is the economic reality it measures? You can ask this question until you are blue in the face without getting an answer. The metric cannot be compared to any other financial metric measuring past returns or future growth. All normal metrics measure the change in some attribute over a period of time, relative to the principle dollars invested during that same period. When calculating a yearly return, the principal invested during that year is used, not some arbitrary value from the past.

Here are the situations in which you hear YOC used.

1. Most often someone throws their own very high YOC into a conversation where it’s obvious intent is to impress you with the high number. Whether or not they disclose that it is YOC (not the current yield) you are clearly meant to compare their metric’s value to your own smaller ‘current yield’, and conclude that they are investing gurus. E.g. “The market is down 10% this year and I am up 9%. My stocks pay me between 5% and 10% based on what I paid for them.”But the investor who has held a stock for years is never any better off than today’s purchaser. This point is proved with a thought experiment. Pretend your broker made a mistake and sold your shares today, only to catch the mistake and reverse the transaction. You will not be aware of the transaction because your economic position has not changed. You would
   • own the same value of the stock
   • hold the same number of shares
   • receive the same value of dividend.
   But now your cost is the same as the current market price – and your YOC is the same as the ‘current yield’. If you never learn of the transaction you may continue to brag about that outsized YOC – your own personal delusion.Too often you hear investors making very bad decisions based on this use of YOC. E.g when a company’s prospects change and the stock should be sold, the long-term owner says “Oh, I can’t sell my stock NOW. I’m earning a huge yield and I couldn’t match that with anything else”. He decides to not sell because his YOC will fall.The way to convince him this is wrong is to change the discussion from ‘percent yield’ to ‘dollars of dividends’. Show him that the proceeds from the sale would buy new shares that generate the same $$ of dividends.
2. Another users of YOC claim it is a psychological aid helping them NOT sell their securities. This can only work if the users believe their large metric shows they are better off than other investors. It is ludicrous for them to claim in the same breath that they don’t really think the metric measures anything real. Either they believe it or not. If they don’t then there would be no psychological effect.It should not be an objective of any investor ‘to hold securities for the longest period of time’. You invest to increase your wealth and fund future needs. You are no closer to those goals when you hold the same security for a longer period of time. There is no payoff. With few exceptions the fortunes of companies ebb and flow. What is blue-chip today will not be tomorrow. You might even benefit from tax-loss selling good stocks.
3. Other users claiming psychological support from this metric to ‘feel good’ cannot articulate what behavior is changed by its use. You can only assume that their use of YOC has prevented their use of proper metrics to correctly evaluate their investment returns. But without the truth they cannot make changes to their strategy or learn better methods. It is never better to shield yourself from the truth in order ‘to feel good’.
4. Sometimes YOC is defended by claiming it allows for comparison of returns from equity vs. debt. They say “The 4% debt I bought ten years ago is still generating the same $40 yearly. But the equity I bought with the same $1,000 has increased its dividend from $20 (2% yield) to $60 giving me a 6% YOC. See how superior dividends are.”But YOC ignores the actual payments over the period in question. It implies that the 6% was paid every year. But if the increase to $60 happened only in the last year, you would come to the wrong conclusion. The debt would have been the better investment. 10 years of 4% is greater than 9 years of 2% plus 1 year of 6%.
5. There are some users of YOC who defend its use (instead of the normal ‘current yield’) by redefining it to BE the ‘current yield’. They include qualifying phrases like “YOC using the yield when the stock was purchased” or “YOC using the stock’s current value”. I guess this is a case of “If you cannot beat them, join them – just don’t admit it.”
6. There are user of YOC who are doing nothing but running in circles. They use normal financial tools to evaluate situations, then translate the situations into YOC, and then translate the YOC back into normal metrics that are meaningful.E.g. some claim that YOC allows them “to see what kind of growth I’m getting on my portfolio.” Imagine a situation where
   * Last year’s portfolio received dividends = $300.
   * This year’s portfolio received dividends = $350.
   Normal finance would calculate the growth as (350 – 300) / 300 = 16.7%.In order to use YOC you would need three three additional steps.
   * Keep track of the cost of the portfolio at some point in history (say) $5,000.
   * Calculate the YOC for last year = 300 / 5,000 = 6%
   * Calculate the YOC for this year = 350 / 5,000 = 7%.
   Then calculate the growth as (7 – 6) / 6 = 16.7%. Nothing was accomplished by adding the three additional steps. The answer could have been calculated directly without any recourse to YOC.
7. There seem to be endless defenses of YOC that involve recounting the historical stock prices and dividends paid of selected companies. At the end of which the great results are attributed to the YOC metric. As if a metric that measures something actually causes the historical performance of a stock. E.g. “YOC allows me to spend $1,000,000 today rather than $3,333,333 in 10 years, to end up with the same annual income.” As if growing the portfolio’s value at 12.8% yearly is questionable (it is) but growing dividends at 12.8% yearly is a perfectly safe assumption because of the wonders of the YOC metric.
8. Then there are the users who simply state “I find YOC on cost relevant. Period. And my opinion is as valid as the next guy’s.”
9. You can work your way though the list of defenses offered by the commentors on this bulletin board thread. Each one is simple to prove wrong, if not ludicrous.

This metric is discussed on the page called Discounted Cash Flow.

Enterprise Value (EV) is a joke. If any advisor or cocktail buddy quotes this metric, smile and walk away. It may (hopefully) be used appropriately by private equity when evaluating companies to buy, but even when correctly interpreted, it has no relevance for retail investors in publicly traded stocks.

The wrong interpretation is that EV puts a market-based value on the business in total – its debt plus its equity. That interpretation is wrong because of the role of CASH in its equation. The EV definition is – the sum of the debt, preferred shares and common shares, all measured at their market values, LESS any excess cash on the Balance Sheet.

Given two companies exactly the same in every other way, the EV of the company with more excess cash will be LOWER. That conclusion does not make sense if you think EV measures the company’s current market value. Why would someone pay LESS for the company that could pay a special dividend tomorrow? A controlling interest that could force a special dividend would be willing to pay MORE for the company with excess cash.

EV does NOT measure an enterprise’s value ‘now’. It measures the total ongoing financing that WILL BE necessary after all the existing long-term sources are bought out, AND AFTER excess cash is withdrawn by the new owners.

There is no relevance for retail investors with minority interests. We do not even have the power to demand a special dividend of excess cash. That is why excess cash is not included in the market’s value of a stock until the day it is declared. We certainly don’t have the power to refinance debt at different rates and terms, or to change tax jurisdictions.

Interesting Reading: Survey of private equity firms asking how they value targets, source deals, change the business to add value, and when to exit.

When a company is bought out from the public market (going private), it is common to hear that the buy-out price justifies an increase in the market value of similar companies still in the market. For example REITs were promoted in 2007 as worth their Net Asset Value, which was defined at that time by the property’s resale value. But buy-out values do not prove values in the public market. Value is subjective to the owner, and different owners face different realities that will effect the subjective worth of the business.

1. The economic forces of supply and demand can work differently for different players.
   E.g. When money floods into private equity (PE) from the public markets, it must be deployed. Demand from PE will bid up the prices of very specific types of companies, while reduced demand (because money exited) from public markets will decrease the value of the remaining companies.
2. Different buyers have different costs of capital – the hurdle rate imposed by debt leverage and the return expected by their owners. The higher the CofC the lower the price a buyer is willing to pay. Companies with lower costs of capital (think GE) get higher returns from the business they purchase so they can pay more for them. Retail investors will most often have CofC higher than most businesses or pension plans.
3. Business buy-outs transfer control. Control is worth more than just a minority interest. The reasons for this include all the following issues. An example of this is in the oilsands, where many projects have multiple partners. The market values the interests of the operating partners higher than those of the limited partners.
   • Your market share increases when you buy-out your competitors. This gives you increased ability to control prices and even restrict supply
   • Every business buys and sells with related parties. Control allows you to assume this business for yourself. It gives you a captive market.
   • Increasing operating efficiency can increase the profits accruing to the owner. Control allows you to impose the necessary changes. Since public company management is shielded from their shareholders, management change is effectively impossible for the public owner. Buy-outs by owners of a similar business, allow them to centralize and reduce overhead costs.
   • Management compensation with options is frequently equal in value to each year’s reported Net Income. Control allows the new owner to collect this value for himself.
   • Taxes play a big role in creating value for only the buy-out owner.
     • Jurisdiction: Control allows you to move the business’s taxable earnings into lower tax jurisdictions.
     • Leverage: Control allows you to load the Balance Sheet with debt whose interest shields the operating income from tax.
     • Tax Status: When you are a tax exempt organization you receive more of the operating profits than if not. E.g. pensions will bid more for a business, because they receive the pre-tax income, while the retail investors gets only the after-tax income.