"Expected" is Not the Same as "Most Likely" -- the Case of Bond Valuation

A common valuation approach is the discounted cash flow method.  Future cash flows are discounted to present value using a risk adjusted rate. The mechanics behind this valuation approach are important in establishing an appropriate valuation mindset.  In the case of standard bond valuation, I believe the widely accepted discounted cash flow valuation mechanics are incorrect, and fail to incorporate an important valuation consideration.

The standard bond valuation equation is:

V = SUM{Ct / [(1+d)^(t)]} + Px / [(1+d)^(x)], where 

V = bond value

C = bond coupon

t = time period, beginning at t=1 and ending at t=x

d = risk adjusted discount rate

P = bond principal

x = remaining life of bond, expressed in number of periods

Here and here are two websites that provide bond valuation equations like the one shown above.

The problem with such a valuation lies in the cash flows used.  The appropriate cash flows to use are expected cash flows, a standard valuation concept.  Coupon and principal payments represent the maximum cash flow possible, not the expected cash flow.   Therefore, the cash flows used to value bonds should be lower than the coupon and principal amount if there is any risk of default.  Even if the bond will probably pay off in full, expected cash flows should incorporate probabilities of possible scenarios.  Thus, I draw an important distinction between "expected" and "most likely."  Coupon and principal payments can only represent expected cash flow if the investor (naively) believes default probability is zero.

For highly rated, low risk bonds, I concede the difference between the standard approach and the probability adjusted approach is likely negligible.  Nevertheless, the importance of illuminating the theoretical shortfall in the standard valuation approach remains.  Thinking in probabilistic terms is challenging. Substituting a shortcut approach for a theoretically sound approach encourages intellectual laziness.  I think students and practitioners would benefit from reinforcing the need to assess alternative scenarios and the probability of occurrence for each.

 An approach incorporating scenario probabilities is useful for more than determining the expected cash flows of an investment.  It also is highly relevant in evaluating the appropriate risk adjusted discount rate.*  After all, in the standard bond equation noted above, the expected cash flows are the promised, or maximum, cash flows.  Implicitly, using them as expected cash flows implies no chance of default.  Conversely, the discount rate used in the equation is a risk adjusted discount rate, with credit risk a component of the risk premium.  This creates a schizophrenic proposition--that bond's cash flows have no risk, yet are discounted using a risk adjusted rate.


Even if it does not always result in a meaningful valuation difference, better to use probability adjusted cash flows with a risk adjusted discount rate.  When done properly, the cash flow scenarios will be consistent with the expected cash flows (numerator) and the risk adjusted discount rate (denominator).used in the valuation equation.*

* The risk adjusted discount rate depends on more than simply the variability of the cash flows.  The correlation of the cash flows  with those of other assets is another factor in determining the appropriate risk adjusted discount rate.  

Opposing Cognitive Biases and Rational Behavior

A friend and I were discussing the logical fallacies that often occur when individuals are making decisions or constructing an argument (hat tip to Ryan Miller).  We identified a particular fallacy--the gambler's fallacy--that exists when an a sequence of identical outcomes is mistakenly construed as implying the probability of the opposite outcome increases for the next iteration.  For example, if the roulette wheel results in black outcomes for 5 spins in a row, there exists a common misperception that the next spin is more likely to land on a red outcome, because overall, red and black outcomes should be approximately equal.  According to the gambler's fallacy, red outcomes would need to "catch up" to make equal outcome possible, and therefore red outcomes become "due" when preceded by a series of black outcomes.


We then considered an alternative and opposite bias that can occurs in the exact same situation.  Many other gamblers believe that a series of black outcomes implies that black is more likely to occur going forward.  From this perspective, black is on a streak, or is the "hot" play.  Its common to hear in casinos that certain craps tables are hot, an implicit belief that a series of outcomes is likely to persist into the future.


I find it very interesting that, in certain circumstances like the one described above, we are prone to any erring on either side of the rational conclusion.  Its unclear whether a sample of 100 gamblers (each without a favorite roulette color) would be more likely to bet on black or red if they observed five black outcomes in a row.


Behavior finance teaches that humans have natural heuristics in evaluating issues and making decisions that make us prone to certain biases.  These biases are generally described as unidirectional.  For example, behavioral finance teaches that individuals generally sell winning investments too early, and hold on to losing investments too long.  This may certainly be true, in general.  However, as described above with the gambler's fallacy, the extent of biases can be more complex, with other possible biases working in the opposite direction.  When competing biases exist, the impact of such biases is not clear then. It depends on which bias, among those that exist, dominates in the circumstance.  Moreover, it seems odd to me that many finance and economics analyses assume investors act rationally, but for certain biases that can be accounted for.  Since we are subject to biases--biases that allow us to err on either side of the rational conclusion--it appears more likely to me to simply accept that humans don't always act rationally.  

Stock Market Attractiveness - A Theoretical Framework

With the choppy stock market of 2011, I thought it would be interesting to evaluate a step-by-step approach to assessing the attractiveness of the stock market.  DISCLAIMER:  I do not want to offer investment advice and this is not a recommendation to buy or sell stocks or any other investments.  Instead, I'd like to share a fairly common approach based on some ideas from CAPM and general macroeconomic evaluation, with some personal tweaks built in.


Because we are looking at the attractiveness of a diversified portfolio of equity securities, we can measure that attractiveness by the risk premium earned relative to other risky assets.  For this post, I'll simply evaluate the risk premium on S&P 500, and not other asset classes. A full analysis would be to evaluate the risk premium's across all investable asset classes (bonds, real estate, private equity, art, coins, etc.), and see which premium-to-risk ratio appears most appealing.  Opinions certainly can differ, even if two investors share the same investment analysis data.


First, lets look at what common stock investments represent.  Each share is entitled to receive 1/X of the firm's dividends, where X is total shares outstanding at the time of the dividend.  Although dividends are paid in cash ("earnings"  can not be distributed), I am going to assume that earnings = free cash flow to equity.  And although dividends are traditionally less than 50% of corporate earnings, to simplify our analysis I'm going to assume that all earnings are distributed.


We can evaluate the attractiveness of the market by first evaluating the price of earnings implied by the market, and then evaluating the expected trends for those earnings.  Thats it.


The price of earnings


Our first step evaluates the cost of "purchasing" the right to annual earnings.  This is simply the P/E ratio. The S&P 500 closed at 1,277.81 at the time of this analysis.  Earnings for 2011 are expected (actuals through 6/30 and estimates for Q3 and Q4) at 97, and expected earnings for 2012 are 106.  In a perfect world, forward earnings should be used.  I prefer current P/E unless its an exceptional year--our world certainly isn't a perfect one.  For this analysis, I'll use an earnings number of 100, resulting in a P/E of 12.8.


What does this number tell us?  On a first approximation,  it tells us that, assuming earnings remain constant into eternity, we'll earn a return on investment of 7.8% (7.8% = 1/12.8).  Again, we're assuming all earnings are paid out in dividends.


Is this a good return for an index of stocks?  I depends on our alternative investments available.  A good benchmark for a safe investment is  US treasury bills, notes, and bonds.  As stocks have no maturity date, its generally considered appropriate to compare yields on stocks to yields on long term notes and bonds.  Currently, the 10 year note is yielding 2.0% and 30 year bonds are yielding 3.1%.  Therefore, our stock yield is between 4.7% and 5.8% more than "safe" government investments.  Is that enough of an additional expected return?  It depends on the investor.  For the past 100+ years, stocks have returned over 7% more than US treasury bonds.  In more recent times, though, many experts believe any excess returns of stocks over treasuries greater than 3% is quite attractive.  Ultimately, attractiveness ex ante is the eye of the beholder.


Adjustments and additional considerations


Our analysis above has been more "back of the envelope" than comprehensive.  Therefore, a number of additional factors may considerably alter the preliminary calculation.  These factors are all considered in the context of the following widely known valuation formula:


V = D / (k - g)


Where,V = Value, in this case the value of the S&P 500 index
D = Dividend, in this case equal to current earnings of the S&P 500 of 100
k = Discount rate.  Discount rate represents the required return for the investment.
g = Annual growth rate in dividends, or equivalently in our case, earnings


The variables V and D in the above equation are known already.  All of our adjustments  are going to affect the value of the index by affecting the future growth rate for earnings.  Once we know g, we'll impute k, which is the market's required return at the moment.   If the market's required return is attractive to us, we invest.  If its insufficient, we invest in other assets, or hold cash.  Note if g = 0, then market's required return is 7.8%, the inverse of the P/E ratio.


Below are factors that should be considered when determining the appropriate value of g in the above valuation formula.


1) Inflation


Common stocks represent ownership of companies.  Companies provide real goods and services.   As prices change from inflation, revenues, expenses, and profits should also adjust.  This makes stocks quite different than bonds, which generally provide only nominal returns.  All else equal, earnings should grow with the rate of inflation.  An a good estimate of inflation is 2% annually going forward.


 Therefore, we should adjust variable g by +2% for inflation.


2) Growth in GDP


As the economy grows, so should corporate earnings.  A reasonable estimate for annual real GDP growth is approximately 2.5% per year.  Note that we should use real GDP growth as opposed to nominal GDP growth. Our adjustment for inflation has already been factored in as part of 1), above.


Therefore, we should adjust variable g by +2.5% for growth in GDP.


3) Earnings % of GDP 


GDP measures of the total value of goods and services produced in a given year.  The cash received for all of these goods and ultimately gets distributed to a variety of "factors of production," or resources used in the creation of those goods. Equity capital is only one factor of production.


If we believe that returns on equity capital (i.e. earnings) is going to increase or decrease as its percentage of GDP, our g variable must be adjusted for this change.  In making this evaluation, its helpful to make separate considerations on macro and micro levels.


a) Macro considerations


GDP reflects public goods and services (i.e. produced by the government) and private goods and services (produced by sole proprietorships, partnerships, and corporations).  If we believe there to be a shift in the sources from which goods and services are produced going forward, that should be considered in our evaluation.*  For our purposes, we will assume that government produced goods and services will remain constant, although I recognize this may be a highly debatable point.


b) Micro considerations


In the private sector, earnings share the value of goods produced with the following other factors of production:


Rents (land and property)
Labor
Debt capital (interest)
Government claims (taxes)


As these "claims" on value shift as a proportion of total value (i.e. total revenue), other sources must reflect the offsetting changes.  For our purposes, we will assume that our relative mix of claims is likely to remain consistent.


As a result of our macro and micro considerations, variable g is unaffected.


4) Creative destruction


Creative destruction is a hallmark of a capitalist society.  It was a phrased most recently offered by Austrian economists to describe the process of competition, innovation and improvement that occur from the entrepreneurial spirit in our economy.  Creative destruction results in the demise of some firms as new firms emerge.  Our creative destruction evaluation will be performed in two stages.


a) Corporate vs. noncorporate earnings


Not all economic activity occurs in publicly traded companies. Sole proprietorships, partnerships, and nonpublic corporations also are responsible for a large percentage of economic activity. Sarbanes Oxley made the costs of being a public company more expensive.  Access to the capital markets is quite value still, however.  We'll project in our analysis that the economic production from publicly traded companies remains the same percentage as currently exists.


b) New corporations vs. existing corporations in the S&P 500


In 10 years, the composition of the S&P 500 will be different than it is today.  Some companies will have  gone bankrupt, others been replaced in the index by new and growing companies, and others still acquired.  One thing that is predictable, however, is that the existing S&P constituents will be a smaller percentage of total public corporate earnings in 10 years than they are today.


This article here identifies that the average lifespan of a corporation is 40 years.  That equates to an average decay rate of 2.5% per year.  We'll use this rate as our adjustment to g to factor in creative destruction.


We'll assume this effect combines to erode existing corporate share of future profits by -2.5%, so we'll adjust variable g by -2.5%.


Conclusion


In total, we have a combined adjustment to our earnings growth of 2% (+2% inflation, +2.5% real GDP growth, -2.5% creative destruction).  We know from our equation above that D / V = 7.8% and  D / V = (k-g).  With g established at 2%, we can solve for our adjusted expected return on equity, which equals 9.8%.  Is 9.8% a good expected return?  As stated above, it depends.  Given the low level of interest rates, receiving an almost 10% expected return on stocks looks even more attractive.  We're still in turbulent times, however.  European debt crisis and American political dysfunction are only two of many concerns affecting economic progress.  It appears to me that our current environment remains one of above average reward requiring above average risk tolerance.  Thats my $0.02.


 *Government production of goods and services is not the same thing as government spending.  Government spending includes transfer payments, such as medicare/medicaid, welfare, and social security.  These payments do not produce goods or services, but transfer money from between sources.

Liquidity and Asymmetric Information

I've been reading a bit on the financial crises recently. One paper by John Taylor, here, describes an endeavor to decipher whether frozen credit markets were a result of liquidity risk or counterparty risk. The discussion of liquidity risk brought up an old nuisance of mine -- the meaning of liquidity across the various contexts in which it is used.

The term liquidity is thrown around in the financial press quite often. Some common examples heard often:

• The fed is adding significant liquidity to the market 

• There's a significant liquidity premium for those assets 

• Banks are facing a liquidity crisis 

I've thought quite a bit about what the term liquidity actually means, and wanted to share my conclusions to date.

First, a few preliminary matters:

Investopedia defines liquidity as "The degree to which an asset or security can be bought or sold in the market without affecting the asset's price. Liquidity is characterized by high levels or trading activity. Assets that can be easily bought or sold, are known as liquid assets."

Multiple types of liquidity exist. The definition above is used in the context of market liquidity. Another liquidity, funding liquidity, refers to a company's ability to meet obligations as they become due.

These separate meanings offer the first insights into why some confusion exists (at least on my part) when liquidity is used in economic discussions.

Both of the definitions above appear to be static characteristics of assets (market liquidity) and companies (funding liquidity), at least in the short term. Given these definitions, certain questions remain unresolved. How can central banks "create" liquidity? How could central bank purchases of low-risk assets, such as US treasury obligations, affect liquidity in markets for risky assets?

In contemplating these questions, I came to believe the existing definitions noted above are incomplete with respect to critical characteristics of liquidity. Below I propose revised definitions, and then explain why I believe these create a more useful understanding of the term.

Market liquidity -- the degree to which market participants believe asymmetric information between buyer and seller exists with respect to assets offered for sale.

Funding liquidity -- the degree to which asymmetric information costs prohibit companies from meeting debt obligations without incurring significant value impairment.

Both of these liquidity definitions are premised on the role of asymmetric information in the market. When an asset owner knows more about the asset than a potential buyer, asymmetric information may prevent the transaction from occurring at fair value. This can occur for two reasons:

1) The classic "lemons" problem: The potential buyer believes the seller would only part with an asset at a price above its fair value. Therefore, the buyer requires an extra discount to mitigate such risk.

2) Owner dependent values: Some assets become more valuable with more informed owners. Loans to customers with whom a bank has an extensive relationship can be one example--such loans may be substantially less valuable to an investor less informed and less effective at responding to any delinquency situations.

Given these considerations, assets require greater risk premia as the available buyers are less informed.

When shocks hit markets or certain market participants must withdraw from trading activities (say a hedge fund experiences significant investor redemptions) fewer informed buyers exist. Central banks can respond to this problem by buying assets, such as US treasuries. You might wonder how buying US treasuries for cash could assist in added liquidity to risky investments. The answer lies, in my opinion, in how such central bank activities affect asymmetric information in the marketplace.

When central banks inject "liquidity" in the market, some investors previously holding US treasury securities now have available cash. Also, the central bank's buying activities has pushed prices of low-risk assets higher, lowering their related yield.  At the margin, investors now have greater ability (i.e. cash available) and incentive (i.e. less attractive yields in low risk assets) to become more informed regarding higher risk assets. 


As an example, consider an economy with three types of assets--cash, low risk treasury bonds, and high risk stocks. For simplicity, assume cash earns zero interest. As the central bank purchases treasury bonds, the economy experiences an increase in cash and a decrease in low risk bonds.  High risk stocks remain unchanged.  The yields of the low risk bonds also decrease due to central bank purchases.  These shifts result in two important observations:


1) Cash represents a larger percentage of assets held by investors. Since cash does not earn any income in this economy, it is reasonable to assume these investors are "looking" for ways to effectively deploy their cash holdings into income-earning assets.


2) high risk stocks become more attractive relative to low risk bonds--simply due to the decreased yields from the bonds.  


As a result of these observations, investors have more incentive to research the high-risk assets in search of income.  Said differently, central bank asset purchases encourage investors to lower the asymmetric information existing for risky assets. As investors perform research and asymmetric information declines, market liquidity and funding liquidity improve.

Barriers to Entry and Fixed Costs - Part II

In my prior post, I described some thoughts on how fixed costs borne by customers (both monetary and temporal) affect industry barriers to entry.  Here I'd like to discuss barrier to entries resulting from fixed costs borne by the company.

My overall theme across these posts is that fixed costs drive barriers to entry.  Despite this theme, I also want to explain how the presence of fixed costs within a company's cost structure do not translate into barriers to entry.  That is, certain characteristics of a company's fixed costs enable those fixed costs to create barriers to entry, while other characteristics prevent fixed costs from becoming a barrier to entry.

Fixed costs enable economies of scale.  As more units are sold, the fixed costs are spread over more units.  Profitability per unit increases as the unit sales increase.  Economies of scale exist only to a point, however.  For example, flight crew and fuel costs are fixed once an airline schedules a flight.  As more travelers are added to the flight, profitability (loss) per passenger increases (decreases). A flight only half filled is unlikely to be profitable for the airline, as the costs per passenger are significant. This is why airlines adjust prices in attempt to fill their flights.  There are capacity constraints on economies of scale per flight, however.  Airlines must limit the passengers on a given flight to the seats available.

This brings me to an important term known as minimum efficient scale.  This is the smallest size of operations that will allow any firm to have a competitive cost structure.  Other firms may be larger, but the additional size beyond the minimum efficient scale does not translate to economies of scale.

When minimum efficient scale is large relative to the market size, economies of scale translate into barriers to entry.  Consider a market with a size of 100 units given existing costs of production (i.e. if companies set price so that they could just recover their costs of material, labor, rent and costs of capital, demand at this price would be 100 units.).  For convenience, we'll say the price at a market size of 100 is $10 per unit.  The minimum efficient scale to serve this market is 40.  Two companies operate at the minimum efficient scale, produce 80, and hence enables them to charge a higher price.  At a market size of 80, the market price is $12.  Hence, each firm earns "extra profits" of $80 (40 produced * $2 elevated price per unit)

A potential new entrant could spot these extra profits and consider entering this market.  If the potential new entrant thought carefully, he or she would realize that two options exist:

a) Enter at a scale of less than 20, so that prices elevated above $10 could persist.  Lets say that the new entrant chose to produce 10, and the market price at a total supply of 90 was $11. If the entrepreneur chose this option, his or her firm would operate at a competitive disadvantage to the incumbent firms, given the minimum efficient scale in the industry of 40 units.  This means the new entrant will have a higher cost structure, possibly even higher on a per unit level than the market clearing price.  In addition, incumbents would be able to lower prices to the $10 minimum price, and still cover all their costs.  The new entrant could not cover all costs at a $10 price, as we know that the minimum efficient scale is greater than scale chosen by the entrant for production.  As a result, a $10 price would result in losses for the new entrant, and persistent pricing at this level would drive the new entrant from the market.

b) Enter at a scale of 40 to achieve competitive parity with the incumbent firms.  Total production would be 120 units.  At 120 units, the market clearing price is $8 per unit.  We know that total costs per unit (including cost of capital) is $10 at the most efficient scale.  Therefore, all three companies would operate at a loss due to excess capacity.  None of the three companies would be willing to reduce capacity, because it puts them at a competitive disadvantage relative to the other two (see a, above, for discussion of this situation).  The most likely situation is that one of the companies exit the industry or go out of business, enabling the other two companies to begin profitable operations again.

In both scenarios above, the profit opportunities for a new entrant appear unlikely.  This prevents new entrants in the industry, despite incumbent profitability.  Here we have true barriers to entry.

Contrast the situation above with an alternative situation where the minimum efficient scale is small relative to the size of the market. Lets say another market had a total size of 10,000 units given existing costs of production, and the minimum efficient scale for operations was 40 units. Price at a total production of 10,000 is $10 per unit. In this case, there are likely to be 250 firms operating at minimum efficient scale*, and each earning only enough to cover total costs.  Lets put aside this likelihood and consider what would happen if there were only 200 firms, each operating at minimum efficient scale.  Production would be 8,000 units, the price charged exceeds $10, and incumbent firms all make extra profits.

How does a potential new entrant evaluate the industry above?  In this case, the entrant realizes he or she can enter at the efficient scale and take some of the extra profits.  Incumbent firms will not possess competitive cost advantages necessary to drive out new entrants.  In this market, new entrants should arrive quickly, until all excess profits are eliminated.

Examples:  I think its helpful to provide specific examples two industries that each have high fixed costs, but one with large minimum efficient scale relative to the market size and the other with small minimum efficient scale relative to the market size.

Large minimum efficient scale - utility (power) companies.  The cost of generation and creating the infrastructure to serve a market are largely fixed.  Power companies generally serve a limited market size as well.  The barriers to entry are so significant in power companies they are often labeled as "natural monopolies", which is another way to say that the minimum efficient scale is the entire size of the market.  Why aren't utility companies wildly profitable then?  Because of the huge barriers to entry, the government has identified the potential for monopoly pricing and regulates prices in these industries.  So although barriers to entry do not translate into high returns on capital for utility companies, this is a result of government intervention in response to existing barriers to entry in the industry.

Small minimum efficient scale - airline industries.  Airline costs exist at the flight and airport level.  Only minimal costs exist at the passenger level, which would be considered variable costs.  An extremely important aspect of the airplane industry is that the capital used in the industry (i.e. airplanes) are mobile, making the relevant market for assessing fixed costs as the global market for flights.  For example, because airplanes can be redeployed to other airports, any profitable routes that emerge will immediately see additional flights offered until profitability is eliminated.  Due to the fixed costs that exist for flights and airport service, airlines seek maximum capacity on flights.  Prices are lowered to gain additional passengers to help recover the costs for flight crew, ground crew, fuel, and airplane leases.  With the number of airlines in the market (remember the relevant market is global due to the ability to redeploy airplanes, even if individual airlines choose to operate only domestically or regionally), price competition erodes all extra profits.  The recent merger between Continental/United and the bankruptcy filing of American indicates just how difficult the airline industry is, despite cost structures for each company that largely consist of fixed costs.

*There of course could be fewer than 250 firms, because individual firms could operate at any scale that is a multiple of 40 and still be operating at an efficient scale.

Barriers to Entry and Fixed Costs - Part I

Much has been written about barriers to entry in assessing industry dynamics.  My favorite author on the subject is Bruce Greenwald - especially because he is both an economist and a value investor.  His book Competition Demystified offers insights to barriers to entry that can relatively easily be added to an investor's toolkit.  He discusses similar thoughts in other books, such as Curse of the Mogul.  All are informative and interesting.


I'd like to begin by address the term "barriers to entry."  When I first became interested in learning about competitive forces and competitive advantages, I took the phrase "barriers to entry" quite literally.  I thought the term meant that when barriers to entry existed, new firms were prevented from entering the market.  Sometimes this is true, such as in the case of when government licenses and patents prevent others from copying the incumbent's business model.  In other instances, the "barrier" in "barriers to entry" simply refers to the rational behavior from potential entrants.  By this, I mean that despite profits earned by industry incumbent(s), potential entrants realize there is no opportunity for abnormal profits on the margin.  With this in mind, the potential entrants rationally choose not to enter.  The assumption that firms will act rationally serves as the barrier to entry in this circumstance.


How can it be true that a firm in the industry can earn above average profits, but potential entrants all believe that they will be unable to do the same? From what I've learned, it largely occurs when fixed costs are present. Fixed costs are costs incurred that do not vary with the number of units produced or sold.  Fixed costs relevant to the evaluation of barriers to entry exist at the company, its suppliers, and its customers.   Here, I'd like to focus on evaluating barriers to entry resulting from analysis of the company's customers.


Multiple customer-related rational barriers to entry exist.  Commonly cited examples include search costs, switching costs, and network externalities.  A brief description of each of these follows below:


Search costs:  Costs involved in looking for a new supplier or service provider.  It can include monetary costs of the search, or time spent and the inconvenience of conducting the search.  For example, if one wanted to change financial advisors, the selection of a new advisor would likely involve an in person discussion of philosophies, investment objectives, and investment capabilities.  After undergoing these discussions, there is no guarantee one would find an investment advisor better than the one used previously.


Switching costs:  Costs involved in using a new product.  A common example includes time and effort learning how to successfully use the product.  I remember when I was an adolescent I stuck with Nintendo video game system when the Sega Genesis became available, because I had become used to the Nintendo game controller and did not want to learn to use the Genesis controller.


Network externalities:  An externality is a cost or benefit not born by those involved in a transaction. In some markets, the product or service becomes more valuable as the customer base grows.  Microsoft Office became more valuable as more individuals and companies utilized the software - sharing electronic files in a compatible format is often quite valuable.


The one common theme between each of these customer-related barriers to entry is that fixed costs incurred by the customer are involved.  Search costs are fixed costs, provided the customer finds the product or service purchased acceptable for repeat purchase. If the customer is willing to choose the same supplier or service provider for the next purchase, no search costs are necessary.  Likewise, the very name "switching costs" identifies that it will not be incurred if the customer chooses to stay with existing suppliers/service providers.


Network externalities are slightly more complicated.  In many instances, the benefit derived from the increasing customer base is that sharing becomes easier or more beneficial.  Sharing can occur between customers (e.g. Microsoft Office, EBay) or indirectly by providing feedback information through the supplier (e.g. Google, Netflix).  To share information effectively, users generally must become sufficiently familiar with the product or service.  The time invested in gaining familiarity is a fixed cost.  Although the cost of gaining familiarity with a competing product or service may be small, habits are hard to break.  I have not considered trying Bing to perform internet searches because I am in the habit of using Google, although I imagine it wouldn't require a significant investment of time and effort to become fairly proficient in using Bing.


When looking at barriers to entry from a customer-related sources, the label of the source as search costs, switching cost, or network externality is generally more important in evaluating the strength of the barrier than in identifying barriers that exist.*  In fact, I believe these labels may obfuscate important observations if they don't fit neatly into a discrete category. For example, switching costs can also serve as network externalities, as in the case of Microsoft Office.  What is important, however, is the identification of customer related fixed-costs already incurred.  Whether monetary or temporal, the presence of customer fixed costs provides a hurdle for any potential entrant's efforts to acquire customers.  The larger the fixed cost relative to the price of the product or service, the larger  the hurdle becomes and the more durable the company's advantage may be.


* I do note that the presence of search costs and switching costs likely serve to defend a company's existing market share,  while the presence of network externalities may suggest growth of market share may persist.

Revisiting Growth and Value Styles

The investing community has generally considered two investing "styles" of equity investments - growth and value.  These two styles purport to represent the types of stocks certain investors seek for their portfolios.  Value investors have been characterized as seeking low P/E, low M/B, and high dividend paying stocks. Growth investors, in contrast, are thought to target high P/E, high M/B, and low or no dividends.  A third style, "growth at a reasonable price," is also cited as an investment philosophy – one pursuing growth stocks but only at an appropriate price.  (This characterization implies that general growth investors care nothing about the price paid for their investments, something very peculiar indeed.)

The style labels of value and growth have troubled me.  I considered the terms growth stock and growth investor to be uninformative.  This is because, I thought, all rational investors are looking for one thing: a stock whose price is below its intrinsic value.  Notice  the word "value" contained within my preceding description, while the word "growth" is missing.  This simple thought led me to conclude that all successful investors were really value investors, some of which may mistakenly refer to their approach as a growth style.  Prominent investors (Warren Buffett comes to mind) have made eloquent observations suggesting growth is simply a component to the basic investment analysis process.  I agreed with this type of assessment.

Fairly recently I have become quite interested in the Austrian school of economics.  One basic tenet of the Austrian school is that the trends and results of complex systems such as our economy "emerge" and are not part of a linear process of cause and effect that can be easily be identified or forecasted in advance.  This belief in "emergence" is largely an accepted principle of complex adaptive systems, a term that describes our economy, among other things. Other complex adaptive systems include the stock market and natural ecosystems.  Complex adaptive systems form when an participants in an environment exhibit interdependencies, i.e. when each individual’s actions depend on the actions of others.  Often times, emerging results in complex adaptive systems can be observed in the form of power laws. A famous power law is the "80/20" principle, where 20% of the units can explain 80% of an effect.  More on power laws can be found here.

How do the idea of emergence and complex adaptive systems relate to my thoughts on growth investing?  Based on what I’ve learned from complexity, I’m more and more convinced that growth investors must analyze the likely emergent results produced by the system, as opposed targeting individual company for analysis.  Said differently, a growth investor must analyze the relevant ecosystem first to identify an emerging winner, whereas a value investor can select individual companies to evaluate. One could label this as a simple top down vs. bottom up difference, but I would disagree.  In the growth style, the analysis is done only at the top, or system level.  There is no “down” necessary for a “top down” analysis.

Importantly, the tools used for analysis are likely different between growth and value investors.  As mentioned above, complex systems often do not exhibit linearity.  There is no clear cause and effect when the environment becomes nonlinear.  These systems do demonstrate predictable patterns, however, such as power law distributions.  Successful growth investing may depend less on fundamental analysis behind a convincing investment thesis than timely recognition of an emergence pattern.  Value investors, on the other hand, stay more in a linear environment.  Microeconomic principles can be applied when cause and effect is clear.  Fundamental analysis and industry dynamics are easier to apply in this environment.  A rigorous investment thesis is more likely to succeed for value investors.