Theories on firm mortality


The mortality of companies


Investigating the exponential age distribution of firms


The mortality of companies

We examine a comprehensive database of more than 25 000 publicly traded North American companies, from 1950 to 2009, to derive the statistics of firm lifespans. Based on detailed survival analysis, we show that the mortality of publicly traded companies manifests an approximately constant hazard rate over long periods of observation. This regularity indicates that mortality rates are independent of a company’s age. We show that the typical half-life of a publicly traded company is about a decade, regardless of business sector.

(Our results shed new light on the dynamics of births and deaths of publicly traded companies and identify some of the necessary ingredients of a general theory of firms.)

A. Public companies and lifespan

Publicly traded companies are among the most important economic units of contemporary human societies. (Alchian 1972Coase 1988Hannan and Carroll 1992, 2000Jovanovic 2001


<- As of 2011, the total market capitalization of firms in the New York Stock Exchange was 14.24 trillion dollars, comparable to the entire gross domestic product of the USA.

While researchers have devoted considerable attention to the distribution of firm size (Simon and Bonini 1958; Stanley, Amaral, Buldyrev, Maass, Leschhorn, Salinger and Havlin 1996, 1997a, 1997b; Axtell 2001), the distribution of firm lifespan has been the subject of far fewer studies (Coad 2010). Thus, despite the availability of much quantitative information, our understanding of the way public companies live and die remains limited.

B. Datasets, Definitions, and Survival biases

Arguments addressing the statistics of company lifespans hinge on the interpretation of the meaning of the death event for a company. In this paper, definitions of ‘birth’ and ‘death’ are based on the sales reports available in the Compustat database. 

-> firms may ‘die’ through a variety of processes: they may split, merge or liquidate as economic and technological conditions change.

i) Datasets

Data on publicly traded companies were obtained from the Compustat North America and Compustat Historical databases,  covering the period of 1950–2009 and contain most financial information for North American and overseas American Depositary Receipt firms reported in their income statements and balance sheets, filed to the US Securities and Exchange Commission. A total of 28 853 publicly traded companies are included in the database. From these, we excluded 2292 that did not report any sales over the 60-year timespan. We also noted that 6868 companies were listed (alive) either in 1950 or in 2009, with 160 of those companies reporting sales for the full 60-year span of the dataset.

ii) Definitions

We define ‘birth’ to occur not at a company’s founding, but rather when it first reports sales in the Compustat database. We take ‘death’ to occur in the year when a company stops reporting sales. For each company, we define lifespan to be the total number of years for which the company reports non-zero sales.

<- There are a number of companies that fail to report for several years between years of activity. Such cases of re-entry are not counted as additional new births or deaths; the additional years are simply added to the total lifespan.

-> This definition is similar to the Bureau of Labor Statistics Business Employment Dynamics measures of entries, which include mergers, takeovers and industrial reclassification

-> This broad definition of death will affect the conclusions we can draw from our data, as an instance of firm death does not necessarily connote failure (Carroll & Delacroix 1982).

* As a metric of mortality that is closely related to lifespan, we use the term half-life, defined as the time it takes for half of the firms in a given cohort to die (following the above definition of death).

* For survival analysis, this half-life corresponds to the age t by which the cumulative mortality fraction M(t) = 0.5 (50%).

iii. Survival biases and subsampling

The historical data on Compustat database do have problems of survival bias (Ball & Watts 1979): almost no firms die in the first 20 years of the dataset. To account for the effects of this bias, we ran our analysis both on the entire dataset and on a set limited to firms reporting sales between 1975 and 2009. A comparison between analysis of the entire sample and this reduced set suggests that the effect of survival bias is limited. This is likely because the first 20 years comprise a very small proportion of the entire dataset.


C. Theory of the firm

There is a great diversity of perspectives on a theory of the firm, focusing on different aspects of their costs, organization and evolution. 

Economists such as Coase [1988, 1937, 1960] and Williamson [1991; 1985] proposed that firms exist in order to minimize (positive) market transaction costs involved in the production of goods and services. In situations when there is particular specificity of goods and services exchanged between two economic agents, such transactions may be best organized internally to an organization rather than negotiated in the open market (Alchian 1972, Williamson 1985 ).

-> As such, firms may split, merge or liquidate in response to economic agents evolving new and better ways of dealing with the various costs and revenues of production and exchange (Arrow 1969; Grossman and Hart 1986; Hart and Moore 1990; Tadelis 2002). 

-> Therefore, at least on the average, the merger of existing companies should be approximately neutral in terms of the balance between costs and benefits. ??

(However, this relatively simple picture becomes more complex in the light of behavioural studies of the impact of decision-making and management practices on the growth and viability of actual firms.)

D. Organizational ecology

In the framework of organizational ecology, companies are seen as units of selection in markets and their longevity is the result of their successes of learning and adaptation in these environments (McPherson 1981, 1983).

Organizational ecology has often suggested that the mortality rates of firms are age-dependent (Hannan and Carroll 2000).

liability of newness (Stinchcombe 1965): young establishments experience higher mortality rates, supported by observation of US manufacturing plants (Dunne, Roberts and Samuelson 1989), Argentinian and Irish newspaper companies and other types of businesses (Carroll & Delacroix 1982).

<- Theoretical grounding draws from the adaptive requirements of market entry; it takes time for young companies to gain the competencies and build relationships that will ensure their ability to survive (Hannan 1998; Freeman, Carroll and Hannan 1983). + New companies are likely to be smaller and less experienced and thus more susceptible to market shocks (Hannan 1998). Knott & Posen [2005] stress the evolutionary character of these arguments by suggesting that liability of newness is evidence for market-based selection.

However, liability of adolescence is also seen. In a study of West German business enterprises, Bruderl & Schussler [1990] find that companies are, in fact, protected from mortality in the immediate period after founding.

<- This liability of adolescence likely results from the buffer a firm acquires via its capital endowment at birth [Hannan 1998], which is also a characteristic of firms that have recently entered financial markets. <- As their initial capital stock is expended, less profitable companies become more vulnerable to environmental changes in market conditions.

And liability of senescence or obsolescence: mortality rates increase as companies age.
1) liability of senescence: as companies age, they accumulate rules and stagnating relationships with consumers and input markets that render them less agile and that re-configuration is increasingly expensive (Henderson & Clark 1990);
2) liability of obsolescence: environmental requirements change over time and that, although firms may improve in competence and efficiency with age by becoming more specialized, these specific adaptations also increase the companies’ risk to new kinds of external shocks that will inevitably beset them.

** Coad [2010a] has argued that these assorted liabilities constitute small deviations, at the tails, from an aggregate lifespan distribution that is generally well approximated by an exponential distribution. <- This proposition has been confirmed in Italian, Spanish and French firms (Navaretti, Castellani & Pieri 2014

<<- the statistical patterns of firm entry and exit will affect the distribution of firm sizes in any given year and set its form and temporal stability (Coad 2010bAmaral et al. 1997a).  <<- thus a better understanding of the mortality risk of firms is necessary to generate new insights on the empirically observed scaling regularities in firm size frequency distributions??? (Amaral et al. 1996; Axtell 2001)

(the existing literature on firm survival has often focused on tracking small samples of firms in specific industries. )


We confirm the hypothesis of an approximately constant mortality rate, finding that the exponential distribution of firm lifespans holds across business sectors and causes of mortality. We apply survival analysis to estimate in a variety of ways that the firms in our dataset have a half-life of approximately 10 years, regardless of age.





A. 新しさの不利益(liability of newness):古い企業に比べて、若い組織は、多様な困難に遭遇することにより、消滅率が高い。

Stinchcombe 1965 によれば、理論的に四つの事由に起因する:1)教え込む前任者がいない;2)有効かつ能率的な役割やルーチンを構築精錬するには時間を要する;3)成員間信頼を構築する時間;4)他の組織との安定した関係がいない

Freeman, Carroll and Hannan 1983の実証が裏つける:1800-1980における476のアメリカの労働組合、1800-19807大都市の地方新聞社2768社、1951-1979アメリカ半導体産業企業1159社 → 組織の年齢が若いほど、消滅率が有意に高かった。

Carroll 1983 は組織の消滅率を扱った52の先行研究のデータを分析した結果、46において年齢との負の関係を示していた。


Carroll & Delacroix 1982Swaminathan 1996: 1800-1900アルゼンチンの新聞社

Singh, Tucker, House 1986:1970-1982トロントのボランタリーソーシャルサービス

Amburgey, Kelly, and Barnett 1993: 1771-1963フィンランドの新聞社

Halliday, Powell,, and Granfors 1987: 1870-1920アメリカの州立法律事務所

B. 若年期の不利益(liability of adolescence):組織の創成期には消滅率は低く、年齢をたるに従い徐々に上昇し、再び下降する

Bruderl and Schussler 1990: 1980-1989西ドイツの企業

Mahood 2000: 1976-1986アメリカの中小企業

Kale and David 1998: 1985-1994アメリカの建設業界

基本負の関係が予期されるため、新しさの不利益を否定するではない、初期に一時的に消滅率が低い状態が発生する論拠、組織の創成期において保有されている資源に求める(組織が有する創業時の財務的資本や人的資源が有意に消滅率を下げる (Bruderl, Preisendorfer, and Ziegler 1992))(外部組織との関係の欠如やルーチンや役割の構築の費用という生存に対する負の影響を緩和する (Fichman and Levinthal 1991))。→ しかしその資源が枯渇する

C. 加齢の不利益(liability of aging):新しい組織ほど生存に有利、年齢と消滅率が正の関係

-> 新しさの不利益は年齢とともに変化する組織の規模をコントロールしないという統計処理の不備により生じた小ささの不利益(liablility of smallness)(Barron, West, and Hannan 1994)

Barron et al. 1994: 1914-1990ニューヨークの信用組合、時間依存変数として各年の総資産額を採用するとこにより、年齢と消滅率の正の関係を実証

Ranger-Moore 1997: 1813-1985ニューヨークの生命保険会社、年齢とともに変化する総資産額を規模の変数としてコントロールした

Barnett 1997; Barnett and Amburgey 1990: 1877-1933ペンシルブァニア州の電話会社

Baum and Mezias 1992: マンハッタンのホテル

Carroll and Swaminathan 1992: 1975-1990アメリカの醸造業者

i) 時代後れの不利益(liability of obsolescence)

環境の変化により、組織の有する能力が環境の求める能力と一致しないこと <- 組織には経路依存性があるため、創成期の環境に適合した組織能力を一度獲得すると、以後引き続いて、新しい環境に対応するように自身の能力を変化させることが困難である (Hannan and Freeman 1984)

ii) 老年の不利益(liability of senescence)(Carroll and Hannan, 2000)

組織の加齢に伴い増加する、組織内の様々な規則、ルーチン、組織構造の堆積により不利益 <- 規則やルーチンは一過性ではなく耐性を持つもの、古いの上に新しいものが積み重なる。その結果、複雑に絡み合った規則、ルーチン、組織構造は組織有効性や能率性を阻害し、安定した環境であろうとも、変化の激しい環境であろうとも、組織に対して費用を発生させる (Barron et al. 1994)

D. ABCの矛盾

Barron et al. 1994の規模コントロール説 <- しても、依然として、新しさと若年期を示している研究が存在


Hannan et al. 1998 先行研究において年齢と規模は過度に単純な変数として扱われており、該当個体群への参入方法と個体の規模の分布を考慮していない <- (その解釈は先行研究の結果を必ずしも十分に説明していない)

** 年齢は、正と負の両者の影響も代替している

<- Thornhill and Amit 2003 はカナダの倒産した企業について調査、古い企業は環境の変化による知識の陳腐化が原因、新しい企業は経営上の知識や財務管理能力の欠如が原因


つながり+知識 -> 新しさの不利益 

他の組織との安定した関係は、資源獲得において有利に動き (Pfeffer and Salancik 1978) + 役割やルーチンは組織の問題解決能力のストックとして、組織能力を形成する (Nelson and Winter 1982)

摩擦 -> 老年の不利益

liability of senescence (Barron et al. 1994; Inkson, Pugh, and Hickson 1970)


資源賦存量 -> 創成期の不利益を緩和 

(Fichman and Levinthal 1991)

環境変化 -> 時代後れの不利益

liability of obsolescence

** 加齢の不利益を示している研究対象は、すべて、多額の資本を必要とする業種(銀行、保険、て電話会社、ホテル)/ 新しさの不利益を示しているは設立に際し、比較的大規模な資本を必要としない組織

前者は慎重である、生存を担保するの知識とつながりを出来るだけ入手しようとする -> 十分なを確保できなかったは、設立を思いとどまる可能性高い -> 新しさの不利益を蒙ることが結果的に少ない、主に摩擦の影響のみが作用している

E. 実証の方向




Investigating the Exponential Age Distribution of Firms

The age distribution to be an interesting feature of industrial structure. The age distribution may be of direct interest in theoretical models of the firm size distribution, as well as providing indirect information on a number of phenomena such as entry rates, survival rates, and possibly even the age of technology used in production.

However age distribution of a population of firms has barely been investigated in empirical work (only Coad and Tamvada 2008 focus on census data covering around 700’000 small scale firms in India & Segarra et al. 2008)focus on a sample of about 85’000 firms taken from the Spanish Mercantile Register).


Figure 2 shows that, even in a sample of small businesses, these firms have very different ages. Most firms are relatively young, but some are extremely old.

The shape of the distribution appears to follow an approximately straight line with negative slope over most of the support. Given that the y-axis is expressed in logarithms, this straight line of negative slope suggests that an exponential distribution would be a valid approximation of the empirical age distribution.

In later sections of the paper, however, we focused on situations in which the exponential gave only an imperfect representation. In contrast to the exponential benchmark, we observed that young establishments seem to be especially numerous, the oldest firms seem to be exceptionally long-lived, and at the disaggregated level of certain specific sectors we can observe a particularly irregular age structure

<- i) Concerning the large number of young establishments, it could be that this excess weight in the age distribution corresponds to over-entry by inefficient establishments who exit shortly afterwards (the case of ‘hopeful monsters’); excess entrepreneurship undertaken by overoptimistic entrepreneurs. (Santarelli and Vivarelli 2007) ->> To the extent that departures from the exponential benchmark among young establishments represent over-entry, then the exponential age distribution could be used to gauge the magnitude of this phenomenon.

<- ii) Departures from the exponential benchmark in the case of the oldest firms also have a ready economic explanation, in that certain long-lived firms, and especially family firms, do not pursue economic rationality in the sense of maximization of expected profits, but instead they may seek to maximize their chances of survival (e.g. by pursuing risk-averse strategies).

<- iii) the exponential distribution may not always be a valid heuristic at the disaggregated level of individual sectors, e.g. the international airline industry. In contrast to the smooth shape observed at the aggregate level, the age distribution of this particular sector is much messier and displayed conspicuous multimodality.

** The exponential is still a useful benchmark for understanding the age structure of industries. For example, we suggest that theoretical models of firm entry, exit, and industry evolution would do well to generate an exponential age distribution as part of their output, even though the empirical data is not exactly exponentially distributed.

(other candidate distributions such as the Pareto, or intermediate cases (between the exponential and the Pareto) could also be fruitfully investigated.)



Firm Turnover and Competition





Fortune 500 Turnover


We find that, while annual turnover on the list has, on average, increased since the early 1980s, it doesn’t quite mean what many people think it means.

It’s easy to paint a narrative around these numbers that coincides with the Great Moderation and the productivity revolution of the 1990s and early 2000s. But reality isn’t so simple.

1. Turnover among big companies is not a new phenomenon.

The late 1950s, as mentioned, experienced moderately high levels of turnover (at least compared to subsequent periods). Prior research has revealed considerable churn among big companies in the early decades of the twentieth century as well.

2. Higher turnover in the 1980s did appear to reflect value creation as corporate conglomerates, ravaged by inflation and competition, were taken apart and remade into separate, more efficient companies.

But, in the 1990s, higher turnover reflected (a) methodological changes in how the Fortune list was compiled, and (b) a mergers and acquisition boom, concentrated in a handful of sectors, that destroyed perhaps as much value as it created. /Turnover is less a broad economic trend than a discrete temporal and sectoral phenomenon


Fortune 500 changes reflects:

1.a kaleidoscopic process of sectoral change and greater efficiencies at the level of individual firms

2. some less sanguine economic developments, which includes the downside of higher volatility—the high M&A volume in the late 1990s included the largest number of the worst deals of the past thirty years—and the deleterious implications for consumers and households.

3. it appears as if performance among the Fortune 500, as measured by return on equity, did not necessarily improve and, if anything, became more volatile over time.


Schumpeter’s ghost: Is hypercompetition making the best of times shorter?

At the center of Schumpeter’s theory of competitive behavior is the assertion that competitive advantage will become increasingly more difficult to sustain in a wide range of industries.   (recently resurfaced in the notion of hypercompetition.)

This research examines two large longitudinal samples of firms to discover which industries, if any, exhibit performance that is consonant with Schumpeterian theory and the assertions of hypercompetition.

We find support for the argument that over time competitive advantage has become significantly harder to sustain and, further, that the phenomenon is limited neither to high technology industries nor to manufacturing industries but is seen across a broad range of industries.

We also find evidence that sustained competitive advantage is increasingly a matter not of a single advantage maintained over time but more a matter of concatenating over time a sequence of advantages.


Update 2017/9/28

America’s Top 50 Companies 1917-2017


The Fall, Rise and Fall of Creative Destruction
What’s the lifespan of a company in the age of startups and tech disruption? Longer than it used to be.

the 5 most valuable corporations of today (by market capitalization, not assets) were all founded in 1975 or later, three of them since 1994 → there has been some creative destruction going on (Joseph Schumpeter “Capitalism, Socialism and Democracy” )
→  whether there is more or less of it happening than in the past
[+] In the business community, especially in and around Silicon Valley, there is a widespread belief that we live in an age of mind-boggling economic upheaval and change.
[-] but economists have been churning out research for several years now that seems to show a decades-long slowdown in almost every indicator of business dynamism.

(E.g.:  “Declining Business Dynamism: Implications for Productivity?” + “The Secular Decline in Business Dynamism in the U.S.” by Ryan A. Decker, John Haltiwanger, Ron S. Jarmin and Javier Miranda; “Declining Business Dynamism in the United States: A Look at States and Metros” by Ian Hathaway and Robert Litan; + “The Rise of Market Power and the Macroeconomic Implications” by Jan De Loecker and Jan Eeckhout. Ben Casselman’s New York Times article last week on the startup slump is a nice nonacademic summing-up.)

a brief history of how this age-of-upheaval story got started

1] Creative destruction is a view of capitalism that had fallen out of fashion in the 1950s and 1960s. In “The New Industrial State,” published in 1967, John Kenneth Galbraith described the U.S. economy as dominated and steered by 200 or so gigantic, permanently profitable corporations.

2] Forbes 70th anniversary issue was a blaring announcement that a new era had dawned. There were pages and pages of lists showing shifts in the ranks of the country’s 100 largest companies from 1917 to 1945 to 1967 to 1987.

+ In 1986, Richard N. Foster, a partner at the consulting firm McKinsey, had come out with a book called “Innovation: The Attacker’s Advantage” that described how giant, successful companies were blindsided and sometimes destroyed by what he called “technological discontinuities.”
(moments when the dominant technology in a market abruptly shifted, and the expertise and scale that the companies had built up suddenly didn’t count for much == organizational obsolescence)

3] In 1990, management scholars Rebecca M. Henderson of MIT and Kim B. Clark of HBS published a now-classic article describing the “sometimes disastrous effects on industry incumbents of seemingly minor improvements in technological products.”

+ in 1995, a young HBS professor named Clayton M. Christensen gave the phenomenon a convincing story line and a name that would stick: “disruptive innovation.” (expanded and expounded upon them in his 1997 classic, “The Innovator’s Dilemma.” ← Christensen studied technological shifts in the computer-disk-drive industry and refine his observations -which were quite similar to what Foster had seen in other industries- as disruptive innovation)
→ backlash in 2014: There was little evidence that disruptive upstarts, or companies that disrupted themselves, consistently won out

(Background: Corporate raiders shook up big companies in the 1980s, forcing mergers and breakups. In the 1990s, several not-very-old technology companies blasted into the ranks of biggest and most valuable corporations.)

4] In a 2001 bestseller “Creative Destruction” Foster & Sarah Kaplan →  documented this upheaval with a striking chart showing the “Average Lifetime of S&P 500 Companies” declining from more than 75 years in the early 1930s to between 25 and 35 years in the 1960s and 1970s to about 15 years in 2000
← problems: it’s about time spent on the S&P 500, not corporate lifetime + it’s calculated simply by taking the inverse of the churn rate of the index (the percentage of companies entering and leaving each year) to arrive at an estimated average tenure +  the managers of what was then called the Standard Statistics Composite Index, and had a lot fewer than 500 companies on it, were as diligent and systematic about adding and removing companies in the 1930s as they have become since

The destruction tendency is wrong!!
(It is true, though, that the corporations that survived the terrible 1930s went on to rule the economy for decades. And the downward trend in S&P 500 tenure is apparent even if you start in 1960, as Foster did in a 2011 report (“Creative Destruction Whips Through Corporate America”) for Innosight, the consulting firm founded by Christensen) 

But, it’s apparent that not a lot had changed since 2000. When Innosight redid the chart for a report last year, it became apparent that average tenure had actually risen since then.)
スクリーンショット 2017-09-28 23.00.10
For more than 15 years now, companies have been staying on the S&P 500 for longer. 

+ Turnover on the Fortune 500 — a list of the largest U.S. companies by revenue published since 1955  — is also often cited as an indicator of corporate upheaval → no clear trend (Turnover rose from the early 1960s through 2000. Since then, it’s down)
スクリーンショット 2017-09-28 23.05.44
(One caveat is that before 1994, the Fortune 500 consisted only of industrial companies, and the addition of service companies after that makes before-and-after comparisons a little suspect.)

+ turnover among the 50 biggest corporations assembled by Victor Manuel Bennett & Claudine Madras Gartenberg (Business School) shows a similar pattern
スクリーンショット 2017-09-28 23.10.42

=>> The great wave of upheaval that began in the 1960s has given way to a period of corporate consolidation and relative stability
(→ a pause than an end to the upheaval? → the past 10 to 15 years as the calm before another technology-induced storm presaged by Silicon Valley’s surfeit of billion-dollar startups and an increase in mergers-and-acquisitions activity? )

The 2012 paper WHAT DOES FORTUNE 500 TURNOVER MEAN? → also question whether the kind of turnover we’re talking about here really is all that reflective of economic change and progress:
→ Departures from and additions to such lists are often driven by waves of mergers and acquisitions that are more about rearranging corporate assets than creation or destruction => “Turnover is less a broad economic trend, than a discrete temporal and sectoral phenomenon.”
→ historical research showing waves of corporate churn in the 1920s and the turn of the 20th century that seem to have been at least as disruptive as those of the 1980s and 1990s

=>> modern capitalism produces and probably requires a lot of creative destruction → But this isn’t a relentless, ever-accelerating process. It goes in waves → For about 15 years now we’ve been in a lull, and it’s not at all clear when or how it will end.

=> Something happened around 2000 that made it easier for top companies to stay on top


In the 2000s, a series of academic papers showed that corporate America had become a much less comfortable place for incumbents:
→ L.G. Thomas & Richard D’Aveni found big increases in profit volatility among manufacturing companies from 1950 to 2002.
→ Diego Comin & Thomas Philippon found a similar increase in the volatility of sales growth and other metrics. Many other studies delivered comparable results.
→ D’Aveni: the “Age of Temporary Advantage,” or of “Hypercompetition,”  or Clayton Christensen: the age of “disruption”

Victor Manuel Bennett and Claudine Madras Gartenberg took the volatility measures from the above papers and a few others, added some of their own, and updated them all
=> “We are able to replicate prior results suggesting that from the beginning of our data, through roughly 2000, sustainability of competitive advantage was decreasing steadily. Interestingly, however, we find a pronounced reversal of that pattern after 2000.”   (profit and sales volatility show more of a plateau after 2000 than a steep fall, but in any case something changed after 2000)

explanations by Justin Fox

1] Regulation.
→ One possibility is that the deregulation of several major industries from the 1970s through the 1990s led to more volatility, and with the end of that deregulatory wave after 2000 things settled down.
→ Another is that the piling on over the years of safety, environmental, land-use and other regulations by federal, state and local governments has given advantages to big incumbents.
→ Yet another is that by easing up on antitrust enforcement and other efforts to protect smaller businesses from bigger rivals, the government has made life easier for the big guys.

2] Capital markets.
→ The number of publicly traded companies in the U.S. is way down from its 1996 peak. Same goes for the number of initial public offerings. → Something seems to have happened — maybe because of regulatory changes, maybe not — to make public financial markets less congenial to newcomers and to corporations in general → making life easier for the biggest corporations

→ the pressures of the financial market and a preoccupation with corporate financial metrics have left most businesses “afraid to pursue what they see as risky innovations” and focused instead on cutting costs

3] Technology.
→ The Internet and other technological advances seem to be leading to the creation of winner-take-all markets in which that winner becomes really hard to unseat. → IT giants keep gaining ground and striking fear into rivals and potential rivals.


On the other side of the increasing volatility research, (Thanks to data that the Census Bureau began releasing a decade ago, economists can now track what they call “business dynamism” in ways they couldn’t before) researchers found that most metrics of dynamism and upheaval in American business have actually been declining for decades, with the downturn steepening after 2000.
→ Fewer new businesses are being launched + the average age of businesses is increasing + job creation and job destruction are on the wane +industries are being consolidated + fast-growth businesses are rarer

(Before 2000, the decline was most pronounced in the retail and service sectors, which has been proven largely been good for productivity (John Haltiwanger) → new national chains armed with new technologies attack local retailers as the incumbents →plenty of upheaval in the top ranks of the business world in the 1980s and ’90s)
=> All of that activity seems to have peaked, however, a year or two after the stock market did in 2000. Measures of big-business volatility began to drop.
=> High-tech start-up activity and what economists call the skewness of growth—how quickly the fastest-growing companies in a sector are outpacing the median company—declined below the levels of the mid-’90s and stayed there.
=> Most worrying of all, the burst of productivity growth that started in 1995 and is widely attributed to the use of new information technologies also seems to have ended in the early 2000s (→ technological change came in sudden bursts, with fast-growing new firms providing much of the impetus)

[+] It’s possible, of course, that the business-dynamism numbers fail to capture some of the economy’s actual dynamism
→ In the technology sector, many upstarts have in recent years opted to sell themselves to Google or Amazon and do their disrupting as part of an already large organization
→ because several of the metrics are based on job counts, what we’re seeing may be less a decline in dynamism than the rise of new, technology-intensive companies that simply don’t need many workers




Papers on Firm death



The case of disappearing firms: Death or deliverance?

The case of the disappearing firms: Empirical evidence and implications

Death is not a success: reflections on business exit



The case of disappearing firms: Death or deliverance?

Many theories about performance, competitive advantage, legitimacy, and leadership rest upon a core assumption that firms, at least some of them, have long, perhaps limitless, life-spans. Long-term survival is not seen as merely a random outcome or an unattainable goal.

This paper surveys a broad set of empirical findings about firms’ life-spans. It is consistently revealed in the empirical literature that the VAST majority of firms, even large firms, survive relatively short periods.

Specifically, this paper makes three contributions to the organizational literature.

1. clarifies the importance of firm’s life-spans as a pivotal research topic
2. appraises a broad range of research about the survival of firms
3. presents general themes about observed life-spans in relation to concepts, measurement, and theoretical issues

A. Motivation, definitions, conceptual and empirical issues involving life-spans of firms


In thinking this way scholars subscribe, perhaps inadvertently, to a ‘meta-theory’ about success and failure. If a firm’s life ends, it must be the result of defects, of competitive mistakes, of managerial failures, inferior resources, and so on. Put simply, the failure of firms is not random, accidental, normal or inevitable, as it is in human life-spans.

How many firms survive? How long do firms survive? Who survives? Is (long term) survival the planned result of first-mover advantages, or large size, or market structure, or competitive advantage, etc?


Although the usage of the term ‘firm’ may seem straightforward in conventional business parlance, pinpointing a research definition involves difficult issues.Are firms identical to business organizations? What about subunits and merger activities of larger organizations? How do size, control, and ownership affect the definition? Any of these definition aspects may have a material impact upon ‘counting’ firms.

According to US Census Bureau, ‘A firm is the largest aggregation of business legal entities (enterprises or companies) under common ownership or control … typically corporations, partnerships, Lilacs, or sole proprietors.’

iii. definitions-large firm and data

Under the Census Bureau definitions, there were roughly 6 000 000 firms in the United States in 2001.In empirical research, researchers often think exclusively in terms of very large businesses. There are a vast number of studies involving databases such as the Fortune 500, Fortune 1000, Standard & Pours, NASDAQ, Compustat, Product Impact of Marketing Studies (PIMS), Thomas Register, and so on. All of these sources focus exclusively on a few thousand very large firms, not 6 000 000 firms.

In a statistical sense, the 500 and the 1000 are essentially outliers. Such studies underestimate the number of competitors, hazard rates, disappearance rates, and exit rates, while over-estimating median life-spans, growth rates, role of mergers. Because of their unusual large size, their prosperity, their prominence in the business press, and their leading position with researchers, these firms are simply in a class of their own.

iv. definitions-disappearing and failure

In business parlance, disappearance is related to failure, dissolution, or exit, conceptually broader than failure, because failure only means bankruptcy, shutdown, dissolution, or discontinuance

In addition to outright failure, fundamental changes in ownership and management—changes in identity, mission and governance—cause firms to disappear even where brand names, assets, and operations superficially continue unchanged. Firms disappear through mergers, acquisitions, and divestments. (the two are highly correlated in empirical studies (Baldwin, 1998; Geroski, 1995))

Merged firms disappear in a most fundamental sense, because they lose their independence, they lose control over basic choices about as mission, they lose control over their finances. Although changes in ownership are often ignored by scholars, their consequences are significant for firm behavior and markets (Caves; 1998; Weston, Siu, & Johnson, 2001).

To summarize, the disappearance of a firm means that something basic has changed, an identity has been lost—whether or not a name or a brand continues to exist. Because some of these changes are difficult for researchers to identify, many studies over-estimate life spans, survival, hazards, and related variables. The rate of disappearance is sensitive to definitions, data bases, samples and measurements. 

vi. Empirical issues-Incomplete life spans

Some firms survive past the ending dates for the research, rendering calculations, such as average life-spans, incomplete and biased. -> A better measure, one seldom used, is median life-spans. When means are misleading, medians can be calculated to more accurately reflect the data. (not available for large firms)

B. Empirical research findings about disappearance and life-spans

We reviewed a total of 240þ relevant documents, covering a broad range of sources, including different academic fields (technology, entrepreneurship, Strategy, Marketing, Sociology, and Organization) as well as different forms of research, such as academic studies, public information, and independent reports. <- many concepts are related to firms’ disappearance rates, failure or mortality rates, exit, entry, and life-spans per se.

In general, these findings show that American business firms are numerous yet fragile. They enter and exit most industries in large numbers. But from a combination of poor decisions plus competition, the vast majority of new, entering firms soon disappear. But, this is not simply an issue about small firms. Even large firms do not enjoy a median life span anywhere near a human person’s life span.

And if one carefully reviews the research literature, it is difficult to square empirical findings against notions about sustainable advantage, corporate leadership, first-movers, or adaptive change.

i. Entrepreneurship/federal data

Using Census data, Birch (1987) claimed that two-third of net-new-jobs generated by the US economy were attributable to ‘small’ firms.

‘Small Business Data Base’ (SBDB, 1998) painstakingly tracks businesses in the US.

スクリーンショット 2016-10-28 0.21.20.png

Using the SBDB, Audretsch (1991, 1995a) studied the innovative propensities of small firms across a wide range of sectors. In general, about 44 per cent of larger startups survived 10 years, whereas only 31 per – cent of smaller startups survived ten years (1995, pg 92). The highest 10-year survival rate for any sector was 72 per cent for large instrument firms.


Boden (2000) studied three broadly defined industry groups—goods, services, and information. Boden estimated the following median life spans: 4.49 years for goods-producing firms, 4.50 years for services providers, and 4.47 years for information-technology firms. 

Dunne, Roberts, and Samuelson (1988b, 1989) used Census of Manufactures data from consecutive 5 year time periods between 1963 and 1982. They confirmed that exits are chiefly comprised of new, small firms. In addition, entry and exit varied greatly across time and across industries, suggesting industry-specific factors may govern these processes. During any 5 year time-frame, entry and exit are strongly correlated, showing that entrants and exits are mostly identical firms. Low market-share and small size were identified as the biggest predictors for exit. Significantly, DRS also noted that rates of entry and exit were increasing during each time period they studied and the relative size of entrants was decreasing.


Summary. Federal databases represent one of the most extensive, inclusive and reliable sources for research on business formation, entry, exit, survival, etc. New firms begin their lives under-financed and vulnerable; almost all entrants are too small—far below minimum efficient scale. Larger entrants have better survival rates and longer life-spans, but not as high or as long as one may expect. Median survival rates of 5 years or less are typical. Average survival rates running between 4 to 10 years are common.

ii. Technology studies

they share a preoccupation with technological change as the driving force explaining industry and organizational change. They study innovation, first-mover advantage, dominant designs, technological trajectories, discontinuity, creative destruction, etc. These studies are useful in our review because many of them include detailed accounts of the entire history of important industries.

Utterback and Utterback et al. (1978, 1996) put forward an explanation of the emergence of dominant designs. They found firms who tried to maintain a traditional business while simultaneously starting a revolutionary business nearly all failed.

Anderson and Tushman (1990); Tushman and Anderson (1986) studied several industries, including minicomputers (1956–1982), cement (1888–1980), glass (1893–1980), and airlines (1924–1980). Both entry and exit were erratic and unpredictable.

Later studies of airlines (after deregulation) and computers (after 1980) showed increasing rates of turnover, exit and outright failure (see Miller & Chen, 1994), consistent with an increasing-turbulence environment.

On a broader level, Gort and Klepper (1982) studied 46 industries. The average shakeout during Stage Two was 53 per cent. They also presented evidence showing that the average time required for an industry to reach its peak number of producers, has been falling. That finding, especially across 46 industries, lent additional support to the idea that change is accelerating, that exit rates and fatalities are increasing, and that industry life spans are probably decreasing.


Summary. Lifecycle theories find significant empirical support but not universal confirmation (Iansiti & Clark, 1994). life-cycle models entail a surprising number of valid empirical patterns: entry by innovators heralding creative destruction; large numbers of firms enter and quickly exit; one severe shakeout; and finally the endgame converges on an oligopoly. The life-cycle model inherently carries a implication that the vast majority of entrants to any industry must fail (or exit), especially over a long-term. Therefore, technology studies reinforce the case for short life spans that was shown through the Entrepreneurship data (above).

iii. Ecology studies

Population Ecology measures variables such as exit, entry, mortality, and hazard rates, and are hyper-vigilant regarding the conceptual and measurement issues.

Aldrich (1979, 1999) estimated US business formation and business failure between 1940 and 1962 at about 3 400 000 new firms created and 2 800 000 discontinued per year. He reported that between 1910 and 1980 the number of US railroads declined from 1250 to merely 10 major firms.. For railroad firms, both the expansion and the exit periods involved astonishing disappearance rates.

Many other studies highlight industries where long-term mortality rates have been exceedingly high, such as automobiles, airlines, telephones (Barnett, 1990), discount retailers, savings and loan banks, etc.

Carroll and Swaminathan, studied strategic groups in the American brewing industry (1991). In 1880 the record showed 2474 brewers in the US. By 1980 only 45 survived. Many of those who disappeared were regional or local.

Hannan and Freeman (1987) investigated newspapers in the San Francisco Area.During the period they studied, 1840–1975, about 2179 total papers were founded,about 200 remained. In addition, they studied semiconductor manufacturers.1197 entrants between 1946 and 1984, 302 firms in 1985. Half of the entrants lasted less than 3 years. This study is especially important because many of the entrants were relatively large firms, or divisions of large firms. It shows that large-scale entry does not guarantee survival or long life.

Dowell and Swaminathan (2000) studied the US Bicycle industry between 1880 and 1915, when the industry experienced tremendous turbulence. 607 firms in 1898 had just 14 firms in 1904, a reduction approximating annihilation of a whole generation of firms in just 7 years.

Baum (1996) reviewed 20 years of PE research. The results are superficially consistent with notions about organizational ‘senility and rigidity’: whereas initial increases in age are correlated with decreased failure rates, ‘old’ age is often correlated with increasing failure rates.

However, research on organizational change (‘structural inertia theory’) have produced only mixed results (Aldrich, 1999). It is accurate to state that mortality risks increase for very old organizations, but the source(s) of those increased risks remains unclear.

Summary. Population Ecology studies report high mortality rates, similar to other findings (above). Failure rates are correlated with newness, size, and market-niche density. Moreover, mortality rates increase in organizational old age.

iv. Entry, exit, and mobility in economics research

Geroski (1995) reviewed empirical research on entry, exit, and failure. He found that smallscale entry is widespread in most industries. Entry comes sporadically, in ‘bursts’ or ‘waves,’ that are not correlated across industries, bursts that do not result from macro-economic shocks. More importantly, most exits result from economic distress. Entry and exit are highly correlated (0.5 to 0.7). Apparently, for many industries, entry is easy but survival is hard.

Caves (1998) reviewed turnover and mobility among firms. Caves defined turnover as, ‘entry and exit + mobility+ changes in control’. 


/Mobility: He shows the variance of growth rates e diminishes with greater size.Mobility seems to be independent from overall growth rates, cycles, firm size, investment patterns, macro-economy, and directionality of demand changes in the industry – many gainers in contracting industries as well as many losers in expanding industries. The facts imply limited life spans for all firms, large as well as small.
/ Entry rates and survival: About a decade after entry, continuing firms are looking at a 5–7 per cent hazard rate (the expectation of failure during the next year). These hazard rates increase when studies include progressively smaller firms in their sampling.
/ Entry and exit through control changes: Economists usually regard control changes as having no economic importance. Caves argues that control changes signify more than simply name changes, that control changes have real and important economic consequences. Caves contends that control changes can lift the productivity of large plants and greatly expand the size of productive small plants. Control changes must be included as part of industry dynamics. This reinforces our argument that control changes are real strategic changes, that they signal the disappearance of firms. Moreover, an active market for control implies that inefficient or vulnerable firms will soon disappear.

Agarwal and Gort (1996); Agarwal (1997) analyzed survival rates for entrants to 33 product markets, including large 3435 firms. They found hazard rates rising until firms reached ages 18–22. Therefore in a mature market, age and survival follow the typical positive correlation, similar to what PE would predict. However, an increase in hazard rates eventually takes hold of firms. Agarwal and Gort also calculated the ‘Mean Residual Life of Firms.’ – given a specified age how many additional years of life can be expected. Residual lives ranged from 5.8 years to a maximum of 14.6 years.


Summary. Economist’s studies of entry and exit, although largely fixed on manufacturing industries, provide important information relative to firms’ life spans: a vast majority of entrants begin far below minimum efficient scale; entry comes in ‘bursts’ or ‘waves’ unrelated to measured demand that inter-industry differences in profits don’t account for. These findings are consistent with the vision of an environment full of uncertainties, surprising spurts of growth and dramatic reversals—an environment where long term survival is problematic.

C. Large firms’ life-spans, in comparison to small, firms or ‘average’ firms

Are large and long-lived firms merely freak events, or do they represent some normal trends?

In 1987 Forbes Magazine reviewed their first ‘Forbes 100’ list and compared it to their 1887 list. Of the original group, 61 firms had ceased operations, 20 had been acquired or fallen out of the top 100, and only 18 firms managed to stay in the top 100. But the 18 did not perform well, data shows only GE survives. 

According to Census data, death rates ran high, 8 per cent, 10 per cent, and 9 per cent for firms employing 500 persons (include all US large firms) during the years 1995, 1996, 1997, respectively. Considering that these data cover mere one-year periods, these rates are quite large.


– Dow Averages: Of 20 Dow Industrials listed in 1920, only 2 remain on the Dow today, ATT and GE. (Pierce, 1995),
– Fortune 500: One-third of Fortune 500 in 1970 ‘disappeared’ by 1983. During the 1980s, no fewer that 113 of the 500 firms were acquired (Collins, 2001).
– S&P: The S&P averaged about 1.5 per cent annual turnover in 20s and 30s, but in 1998 the turnover rate in the S&P 500 had increased to 10 per cent. Annual S&P turnover, as a rolling 7-year average for whole 20th Century, is increasing. The average lifespan of S&P companies has been falling since about 1930, to less than 15 years today. (Foster & Kaplan, 2001).
– Geus (1997) reported findings from internal studies at Shell. His information placed the average lifespan of multinational firms at only 40–50 years.

Private research

Collins (2001) looked for firms that made a transition, from ‘good to great.’, Starting his research in 1980, his team eventually cut 1435 firms to merely 11. 

Foster and Kaplan (2001) compiled a database of 1000 large firms in 15 industries to search for patterns (did not include diversified companies or industries with overwhelming dominant firms, such as Autos) across four decades. Only 160 of 1008 companies survived from 1962 to 1998.

Weston show that the GNP share of the largest 200 firms has declined since 1970,ranges from 30 per cent to 40 per cent in US.

Mueller (1986) investigated a sample of 1000 large firms, 1950–1972. He observed stable market leadership in only 44 per cent of industries he studied. Out of 1000 firms, only 583 were still operating in 1972; 384 had been acquired.  

In addition to first-mover research, studies show that dominant shares decline: Shepherd (1997); Baldwin, (1998); Geroski (1995); Caves (1998); Caves, Fortunato, and Ghemawat (1984); Davies and Geroski, (1997); Elzinga and Mills (1996). Ferrier and Smith (1999). Specifically, Weiss and Pascoe (1983) found industry leaders dethroned in 39 per cent of industry segments they studied.


Perhaps mergers and acquisitions provide a back door escape from the specter of disappearance, but only for big firms . (according to the SBA (1998, 2000) only 2.6 per cent of US firms were involved in mergers)

Healy, Palepu, and Ruback (1992) studied post-acquisition performance in the 50 largest mergers US between 1979–1984 using lots of complex controls. They found that merged firms did more restructuring than comparable firms. Large firms did not just acquire firms; they also divested units, creating new firms. In 1990s divestitures represented about 35 per cent of M&A activities (Weston et al., 2001). Many studies report limited gains to acquirers and rapid divestment of acquired firms (Anslinger, & Copeland, 1996; Bradley, Desai & Kim, 1983; Caves & Porter, 1978)).

Dunne, Roberts, and Samuelson (1988a, 1989) used the Census of Manufactures, covering 1963–1982 387 two-digit industries. In general, diversifiers did not enter with new plants, they bought existing capacity. Diversifying entrants obtained high initial market shares, grew faster after entry, and had higher survival rates than small firms.

Baldwin (1998) reported many firms who entered by acquisition, soon exited—10 per cent in the first year. The cumulative exit rate of acquirers was about 60 per cent after 9 years. After 10 years, the hazard rates of Greenfield startups, acquisitions, and continuing firms all converge around 5 per cent.

Summary. surveying across many academic fields we find consistent indications that failure rates are increasing, even for large firms, that large firms face more turbulence and more challengers today than 50 years or 100 years ago. These trends all imply shrinking average life-spans. Based upon the research cited above, we could venture a ballpark estimate that medium-size and large-firms are approximately 20 years old and they can probably expect to survive another 20 yearsAll in all, large firms do not occupy a separate universe where marginalization, merger oblivion, failure, bankruptcy, and dissolution do not apply.

D. Themes, Puzzles, and Implications of Disappearance and life spans

Theme 1. The odd behavior of business firms

It’s widely known that a vast majority of small firms enter haphazardly, operate at an undersized, inefficient scale; and they fail (exit) at a prodigious rate. There is a strong case that newness, small size (in employment or capital) and associated inefficiencies all contribute to rapid failure. A majority of these firms cannot continue operations for 5 years, much less provide income during an entrepreneur’s working lifetime.

Theme 2. Over time the failure rates of acquired units converge toward the failure rate of new entrants

Diversified firms and large-size entrants, especially those with related experience, incur lower hazardrates, perhaps on the order of half the hazard rate of newly-founded firms. Even so, empirical research finds excess, illtimed entry and high exit-rates for subsidiaries, divisions, and strategic business units. As studies like Biggadike (1979), Yip (1982), and Robinson et al. (1992) found that the expected response to entry (increased output, ads, retaliatory pricing, etc) was highly selective or even entirely absent.

-> The results of real mergers, acquisitions, and allied activities do not comfortably fit a strategy framework (as Porter, 1980) or popular visions of entrepreneurship (Bhide, 2001).
-> Given large firms’ experience, their financial muscle, their vast core competences, giant strategic assets, and so forth—why aren’t large firms more successful at diversifying entry?

Theme 3. Entrants cannot resist an impulse to join an industry shakeout

Theme 4. The bigger they are the harder they fall: The MBA and the NBA

As a result of competitive mistakes and economic distress, large firms have an unimpressive average life-span. Despite their size, their vast financial and human resources, average large firms do not ‘live’ nearly as long as ordinary Americans. Therefore, setting out sustainable competitive advantage represents an elusive goal as achieving the combined goals of high growth, highperformance and long term survival is truly RARE.

(In statistical terms, outliers are usually viewed as a ‘problem.’ They violate assumptions etc. Therefore, outliers are routinely discarded because they distort variances and central tendencies.What would be the relevance of studying high-performance or sustainable advantage among US auto firms now, when only two firms remain—Ford and GM?)

Theme 5. Disappearance and design

As industries age, the proportion of ‘disappeared’ firms rises compared to the number of continuing firms. In a mature oligopoly, life-spans stabilize and the variance of life-spans becomes tighter.This pattern is tied to the shape of growth curves, industry-life-cycles, oligopoly, and empirical observation.

Theme 6. Time, and performance

Wiggins and Ruefli’s (2002) by using Compustat PC Plus, created a huge sample of 6772 (large) firms, from 40 industries, plotted across overlapping time periods up to 25 years. Their dependent variables were Tobin’s Q and Return-on-Assets, defining ‘sustained competitive advantage’ as a 10-year period of above-average performance for either dependent variable. During the entire 25 year period, about 5 per cent of firms achieved one 10 year stretch of superior ROA returns. Only 2 per cent of firms achieved any 10-year period above average Q. If the period is extended to 20 years, only 4 firms met the Tobin’s Q criterion! Although their study does not directly tell us anything about life-spans (they only studied 25 year survivors, not firms that disappeared), it does illustrate the consequences of longer time periods for organizational variables. Consistent with other authors, such as Mueller (1986) and Baldwin et al. (1995), their data show a broad pattern of regression toward the mean

Theme 7. Increasing turbulence

Although we did not review industry turbulence per se, we found considerable evidence pointing toward increasing turbulence, taking the form of increasing entry rates and increasing exit rates across a broad range of industries. It is not much of a stretch to suppose that increasing globalization, technical advances, and added pressures from investors could combine to create a very dangerous competitive landscape, one where thoughts of sustainable advantage are purely fantasy


The case of the disappearing firms: Empirical evidence and implications

Stubbatt and Knight’s idea: if the overwhelming majority of organizations have relatively short life spans, then the meta theory is misguided in believing managers make a difference.

In order to establish the life span of most organizations, they review ‘243 relevant documents’ that provide empirical accounts of organizational survival and disappearance.

-> They conclude that very few firms survive even a few years and that even successful firms rarely get beyond four decades. Further, the life span of firms is shrinking in the face of competitive pressures.

–>> The high probability of failure facing new ventures is already well established by entrepreneurship text  and organizational ecology. Organization theorists, similarly, for some time have observed the high ‘hazard’ rates confronting new firms and have explored the liabilities of newness and of adolescence. 

–>> the most interesting observation, however, is that almost all firms have brief lives and that established and successful corporations are subject to high failure rates. Is it true?

A. The Notion of ‘Disappearance’

For them, the endpoint of a life span is ‘disappearance.’ It covers many things from bankruptcy, mergers and acquisitions or the conversion of a closely held company to a limited corporation. 

<- It is important to separate these different meanings of disappearance because Business failure implies that managers were ineffective or perhaps irrelevant, while Mergers and acquisitions, on the other hand, might be the outcome of managerial prescience.

-> Sometimes, a firm acquires or merges with another in order to shift its scale or scope of operations, or in response to changing social expectations. For example. Green and Black, an extremely successful ‘fair trade’ cocoa company ‘disappeared’ when it was acquired by Cadbury Schwepps as the latter company repositioned itself to capitalize on the emerging consumer movement to support socially responsible businesses. Similarly, Price Waterhouse and Coopers & Lybrand merged to gain the geographical scale necessary to service the very largest transnational corporations.The entrepreneurship literature also talks about ‘serial entrepreneurs’ who start up businesses, bring them to success and sell them.

–>> Our key observation here is that ‘disappearance’ is not an unambiguous term. Firms disappear for different reasons. If our concem is to infer the role of managerial agency, then we need to distinguish between disappearance arising from poor or irrelevant management (what we might simply call organizational ‘death’), and disappearance that results from managerial prescience (organizational ‘deliverance’).

e.g. PricewaterhouseCoopers is the largest global accounting firm, ‘Legally,’ it began life in 1998. But, for us, PricewaterhouseCoopers is the consequence of managerial prescience on the part of the two legacy firms. These legacy firms did not simply ‘disappear’; they delivered themselves into a more effective firm for the 21st century. 

B. The Extent of ‘Death’

 40 per cent of the 100 largest consulting firms in the world existed in one form or another 50 years ago. All 10 of the world’s largest law firms pre-date 1950 and four of them trace their origins into the 19th century. Here we are defining survival as continued commercial activity irrespective of mergers or changes of name or ownership form. 

Some of sources cited by Stubbart and Knight hint that the life span of many organizations may well be longer than they suggested.

i. Foster and Kaplan (2001) ‘s study, conducted by McKinsey & Co., reports on 1008 companies over a period of 38 years only 160 remains. However,only 249 firmswere included in the database in 1962,the survival rate of firms in the 1962 cohort was 160 of 249 firms, or 64 per cent.

ii. In more scholarly but dated analysis by Mueller (1986) of the largest 1000 firms in the US between 1950 and 1972, he found that 58.3 per cent were operational over the full period and that only 1.9 per cent had been liquidated. The remaining 38.4 per cent had been acquired. Moreover, it is rtot the case that acquired firms were failing: ‘We conjecture that from 1952 to 1972 relatively few of the companies that were acquired faced immediate bankruptcy had they not been merged’. On the contrary, an earlier study by Mueller found that acquired firms were equally profitable to similar non-acquired firms.

->> sufficiently large numbers of established firms do survive for lengthy periods, and that managers probably have something to do with which firms survive and which do not.

Nevertheless, firms may differ as different survival and disappearance rates across industries. The professional service firms operate in very low capital intensive settings and survival may be easier than in industries where capital is more pronounced. 

Another possible source of variation in organizational survival and performance is ownership form. Most of the materials show bias to the limited liability, publicly traded corporation, while other ownership forms exist (Greenwood & Empson, 2003; Hansmann, 1996). Recently, Miller and Le-Breton Miller (2005) highlighted how family-controlled businesses ‘vastiy out survived’ non-family competitors. They report data on forty large, family-controlled businesses, of which more than half ‘had survived for over a century’. 

And international comparisons. Stubbart and Knight pragmatically restricted themselves to US data, but the literature on ‘varieties of capitalism’ (e.g.. Hall & Soskice, 2001) would suggest there may be significant variations between countries in the life cycles of organizations

->>Is the US, more vulnerable to high failure rates because of its widely assumed focus upon short-term results (e.g.. Miller & Le-Breton Miller, 2005)? Do transnational differences in capital markets affect longevity? Do intemational differences in the importance of the social dimension of economic transactions play any role in organizational performance and survival?

C. Bring on Death?

If we conceive of organizations as utility maximizing humans, then, yes, they ought to live forever. But if we adopt a different analogy, namely, that organizations are tools, similar to machines, as suggested by Weber (1964), then organizations ought to last only until their functional utility is exhausted. 

In fact, this is precisely how business corporations were originally conceived. They were designed as limited purpose entities and, until relatively recently, most corporate charters were granted not only for a limited time period, but also for narrowly specified purposes. From this point of view, it is hardly surprising that business organizations do not ‘outlive’ individuals. If business organizations are conceived with the underlying metaphor of a tool used by investors to maximize shareholder wealth, then longevity might be considered an undesirable outcome.

So, why is the eternal life of an organization considered to be an attractive goal today, Put another way, who benefits from creating the illusion that organizations are sentient organisms with the capacity to choose their life span and their scope of action?

<- It is the managerial class, the MBAs and CEOs that now populate business organizations that benefit most from changing the underlying metaphor of business organizations away from the idea that they are tools and toward the perception that they are ‘corporate persons’ without constraints on life or purpose.


Death is not a success: reflections on business exit

This paper is a critical evaluation of claims that business exits should not be seen as failures, on the grounds that sometimes they correspond to voluntary liquidations, or because they are /successful learning opportunities. We reiterate that the vast majority of business exits are unsuccessful. Drawing on ideas from the organizational life course, we suggest that “death” is a better word than “failure” to describe the phenomenon of business exit – we underline that it is not helpful to consider business exits as successful events.

Business exit always relates to unviable businesses – whether they be ‘relatively unviable’ when taking into account the entrepreneur’s outside options, or ‘absolutely unviable’ in the economic sense of being unable to cover its costs. Viable businesses that remain in operation even after the entrepreneur leaves (e.g. trade sale or initial public offering (IPO)) are not, in fact, cases of business death, but cases of business continuation.


In some cases, such as an IPO or an acquisition involving the sale of the start-up, entrepreneurial exit can be considered to be a success. Brander et al (2010, p4) write that “using exits as a measure of success is standard in the venture capital literature.” 

<- This kind of successful entrepreneurial exit, according to which the business continues operations after the exit of the entrepreneur but under new management or with new investors, should be conceptualized as a case of business survival, not business exit.

-> We would agree that this type of exit should be seen as a success. However, given that our paper is not concerned with entrepreneurial exit, or investor exit, but our unit of observation is the business, we can sidestep this category of events. We therefore distance ourselves from the standard approach in the survival literature that considers merger and acquisition (M&A) to be a form of exit (e.g. Schary 1991; Cefis and Marsili 2006; Bhattacharjee et al 2009; Balcaen et al 2011). 

(Although reincorporation and change of legal form may constitute a death and re-birth in the way these events are recorded in some national statistics databases, this is not a meaningful death/rebirth in an economic sense (Harada 2007 p403; Hoetker and Agarwal 2007 p447), and statistical offices recognise this and are working on ways of no longer coding a change of legal form as a death and subsequent rebirth. )

We argue here that voluntary closure can be characterized as ‘relatively unviable’: the case where the business has failed to be a viable economic entity when the entrepreneur considers her other outside options, even if it generates enough revenue to cover its costs./The business has perhaps played a useful role in the past, but now the opportunity cost of the business remaining in operation is too high to allow it to continue.

We define ‘absolutely unviable’ with involuntary closureas the case when a business fails to cover its costs even when we leave aside issues of the entrepreneur’s opportunity cost. 

While involuntary business closure corresponds to bankruptcy, voluntary business closure refers to liquidation, which can be described as either a ‘harvest’ liquidation or a ‘distress’ liquidation (Wennberg et al, 2010).

-> Of these two latter outcomes, harvest liquidation is considered to be more successful than distress liquidation. Harvest liquidation corresponds to the liquidation of a successful business, for motivations such as retirement, or perhaps the natural winding-down of projects that had always been thought of as short-term projects.

–>> In our view, the distinction between voluntary and involuntary closure is not always very helpful. 1) self-reported evaluations undertaken by unsuccessful entrepreneurs are likely to be strongly affected by cognitive biases . 2) many business exits that are classified as voluntary closures would have been classified as involuntary closures had the business closure taken place shortly afterwards. 

Technology Spreading


The impact of public mechanical clocks on economic growth

Are Smart Phones Spreading Faster than Any Technology in Human History?


In a past post about short-termism, we introduced a picture that shows the shortening diffusion time of new technologies from telephone to internet access, which means the period of competitive advantage is shrinking for companies.

Today we go through two articles about the topic of technology spreading.


1.The public mechanical clock in 13th century

Lars Boerner, one of my lecturer in the EH department of LSE, wrote a paper about the impact of public mechanical clocks on mid-century Europe’s economic growth.

European cities that were quick to install mechanical clocks enjoyed greater growth than late adopters. However, it takes some time for the effects from fundamental innovations of this type to be realised because the technology must be accepted both culturally and socially and then applied to related economic activities.

We focus on the later part.

Starting in the late 13th century, clocks started spreading simultaneously across Italy, Germany, and England, and then spread until 1450 in most other Western and Central European Countries, including Belgium, the Netherlands, France, Spain, Poland, the current Czech Republic, and Scandinavia

The diffusion curve of the mechanical clocksevergninifig2

The spread followed the s-shaped distribution curve that is typical for new technologies. The distribution weighted by the total population size started slowly, then rose exponentially, and finally slowed down again, reaching a saturation point in 1450 with a relative decreasing adoption rate.

-> The economic use of clocks was a slow process of adoption.

<- Building a clock in a town was motivated by prestige and not by economic needs – towns did not forecast ex post benefits as an economically efficient application.
<- The use of clocks for coordination activities, such as market times or administrative town meetings, can already be observed during the 14th and 15th centuries, the use of clocks to monitor and coordinate labour processes evolved only slowly, in particular during the 16th century.
<- Finally, a cultural adoption reflected in the daily cultural and philosophical thinking of the time can be observed from the middle of the 16th century

The conclusion here is that for the implement of new technologies, it is indeed important to let societies learn to use them and adapt to them to reap the full benefits.


2.Smart phones spreads faster than any technology in human history

Today, we have the fastest ever diffusion speed in the human history: software apps are reaching tens of millions of users within weeks.

<- Underlying these developments: the unprecedented speed at which mobile computers are spreading.

Presented below is the U.S. market penetration achieved by nine technologies since 1876. Penetration rates have been organized to show three phases of a technology’s spread: traction, maturity, and saturation.


It took almost a century for landline phones to reach saturation, or the point at which new demand falls off. Mobile phones, by contrast, achieved saturation in just 20 years. Smart phones are on track to halve that rate yet again, and tablets could move still faster. 

-> Those technologies with “last mile” problems—bringing electricity cables or telephone wire to individual homes—appear to spread more slowly.

Smart phones have gone from 5 percent to 40 percent in about four years, despite a recession. [1]




In late 2006, the quarter before Apple announced its now-iconic iPhone, only 715,000 smart phones were sold, representing just 6 percent of U.S. mobile-phone sales by volume. Up to that point, the smart phone was spreading not much faster than personal computers had in the preceding decades, and more slowly than radio decades before.

That changed when Apple’s iPhone sold 1.12 million units in its first full quarter of availability, despite prices starting at $399. Year over year, the market share of smart phones almost doubled, to 11 percent of U.S. mobile-phone sales. Now Nielsen reports that smart phones represent more than two-thirds of all U.S. mobile-phone sales.


Lifespans of Firms


Creative Destruction Whips through Corporate America

Die Another Day: What Leaders Can Do About the Shrinking Life Expectancy of Corporations

BCG Classics Revisited: The Growth Share Matrix

Where Do Firms Go When They Die?

This Is a Fine Time to Be a Big Corporation

Can a company live forever?



Lifespans of Top companies are shrinking

Richard N. Foster, first in his book Creative Destruction with Sarah Kaplan in 2001, and then in a report in 2012 argues that “lifespans of top companies are shrinking” according to the study on S&P 500 index rotation.


“61-year tenure for average firm in 1958 narrowed to 25 years in 1980, and to 18 years in 2012, based on a seven year rolling averages.”

“If removal from the S&P 500 is not due to an acquisition [3], it can be a jolt that emphasizes the urgency of turnaround effort or it can be a prelude to delisting and the threat of bankruptcy.”

->” To survive and and thrive,  leaders must “create, operate, and trade” – build new divisions and trade mature ones at the pace and scale of the market without losing control their company. “

“‘creative destruction’, widely credited to Joseph Schumpeter, indicates that economic progress, in capitalist society, means turmoil”

-> “According to Foster, the life span of a corporation is determined by balancing three management imperatives:
1) running operations effectively,
2) creating new businesses which meet customer needs [1]
3) shedding business that once might have been core but now no longer meet company standards for growth and return.” 

<- The problem is that the innovation that is needed to create significant new businesses can often directly conflict with the operational effectiveness of the current business. Under these circumstances large companies slowly fall behind the pace of change of the economy.  [2]


A changing Business Environment 

In 2014, BCG revised its famous BCG’s growth-share matrix [4], as the world has changed:

since 1970, when it was introduced, conglomerates have become less prevalent, change has accelerated, and competitive advantage has become less durable. The business environment has become more dynamic and unpredictable, and market share has become less of a driver of and surrogate for advantage.

1. companies face circumstances that change more rapidly and unpredictably than ever before because of technological advances and other factors.

-> proportionately lower numbers of cash cows because their longevity is likely in many cases to be curtailed.

<- companies need to constantly renew their advantage through disciplined experimentation (to invest in more question marks, experiment with them in a quicker and more economical way than competitors, and systematically select promising ones to grow into stars), increasing the speed at which they shift resources among products and business units share is no longer a direct predictor of sustained performance.

<- new drivers of competitive advantage, such as the ability to adapt to changing circumstances or to shape them



Shrinking Life Expectancy of Corporations

“we are in an era of unprecedented technological disruption and change that only the most forward-looking companies will survive.”

To understand how the battle for long-term survival has changed and the implications for both challengers and incumbents, BCG analyzed patterns of entry, growth, and exit for 35,000 companies publicly listed in the US since 1950-2013 —with surprising results.

1.  corporate survival and death

BCG focused on companies exiting the public-company pool —whether owing to bankruptcy or liquidation, merger or acquisition, or other causes (exit years correspond either to the database (compustat) deletion date or to the last recorded instance of income or revenues.), and found is that public companies are perishing sooner than ever before


Since 1950, the total life span of companies (calculated as the period between founding year and exit year (irrespective of gaps in sales reporting)) and the length of time that their shares are publicly traded have significantly decreased.In fact, businesses are dying at a much younger age than the people who run them.

2.Rising Mortality Risk

Today, almost one-tenth of all public companies fail each year, a fourfold increase since 1965.

スクリーンショット 2016-10-21 15.28.47.png

The five-year exit risk for public companies traded in the US now stands at 32 percent, compared with the 5 percent risk they would have faced 50 years ago.

One might expect particular types of company, such as new entrants in the technology sector, to account for most of the observed shift. Surprisingly, however, our research shows that the surge in mortality risk is widespread: 

There are no safe harbors. Mortality risk grew relatively uniformly across all sectors of the economy.  [6]

Neither scale nor experience is a safeguard. Mortality risk also grew for companies of all sizes and ages. [7]

3. Why it happens

In the economic and venture-capital-funding booms of the mid-1980s and mid- to late-’90s, many smaller and younger companies entered the public markets.

-> These companies (less than 10 years old and had less than $50 million in sales) had a more than 25 percent higher risk of failure compared with the average company—likely owing to poorer quality (a lower bar for entry), few buffers against failure (lack of resources), and intense peer pressure (the smallest companies faced the highest competitive density).

–>> Those that endured grew into serious competitors for incumbents, driving up the death rate among medium-size and large established companies, which were often unable to react quickly enough to the disruptions wrought by these smaller upstarts.

—>>> Surviving incumbents then began to react by acquiring small-to-medium-size companies, again driving up exit rates in this segment while stabilizing turnover among large companies. [8]

4. A Growth-Endurance Trade-Off

BCG observed a surprising relationship between revenue growth and mortality:  accelerated growth correlates with shorter life spans, whereas companies with more moderate growth face the lowest risk. (This correlation holds true across the entire period of analysis and remains significant even when controlling for factors such as age, size, industry, and profitability.)


+ Over the period of our analysis, the average cumulative profits of public companies declined at an even sharper rate than corporate life spans.

-> rather than achieving their full potential in less time, the contraction of corporate life spans, on average, diminishes long-term value creation. Longer-lived companies thus appear to create more value than fast-growing, short-lived ones.


Publicly-traded firms die off at the same rate regardless of their age or economic sector

Log sales of some 30,000 U.S. Companies 1950-2009
(controlling for inflation and GDP growth)


This picture credited by Marcus Hamilton and Madeleine Daepp partly proves some assumptions BCG made above.

<- The relatively rapid growth of smaller companies near the beginnings of their lifespans account for the ragged lower portion of the chart, as well as the relatively steep initial sales increases. As companies reach maturity, their sales tend to level off.

Looking at a database of 25,000 companies from 1950 to 2009, they found that publicly traded companies die off at the same rate, regardless of the firm’s age or what sector it’s in. In their dataset, they found that most firms live about 10 years and the most common reason a company disappears is due to a merger or acquisition.

(However, the problem here is that they are only doing with listed time. Firms may take long time until going listed, and different industry may vary significantly. And can we simply recognize companies merged equally as being demised or perished?)


Fortune 500 story

Researchers Dane Stangler and Sam Arbesman have some different ideas by examining annual turnover in the Fortune 500 list, the first of which was compiled in 1955.


The number of companies leaving (and entering) the Fortune 500, which ranks the largest corporations by revenue, rose through the 1980s and 1990s, but is down since.


-> Stangler and Arbesman cite several historical studies that seem to show that this stuff comes in waves. The 1920s was a decade of high turnover in the ranks of the country’s 100 largest corporations; the 1940s saw very low turnover. Then again, turnover in a list of the top 500 or 100 companies isn’t the same as the average corporate life expectancy. It could be that the long-run trend there is in fact downward.

-> Big companies have been doing a better job of sticking around and thriving since 2000. Meanwhile, broader “business dynamism” research based mostly on U.S. Census data shows that, on the whole, bigger, older companies control more of economic activity than they did not just 15 years ago but also 40 years ago.

->>Big companies have gotten better at adapting to technological change, or that the particular technological changes we’re currently experiencing favor the big over the small?

see more on firm turnover rate and competition


Can a company live forever?

It is perhaps unsurprising that the country where people live the longest is also home to some of the oldest companies in the world.

<- In Japan, there are more than 20,000 companies that are more than 100 years old, with a handful that are more than 1,000 years old, according to credit rating agency Tokyo Shoko Research.

->  1. Professor Makoto Kanda, who has studied shinise for decades, says that Japanese companies can survive for so long because they are small, mostly family-run, and because they focus on a central belief or credo that is not tied solely to making a profit.

-> 2. Local factors could be another key to their success. Shinise focus primarily on the Japanese market, from Kikkoman’s products to small sake manufacturers, and they benefit from a corporate culture that has long avoided the mergers and acquisitions that are common among their Western counterparts.

Although there are exceptions to every rule, the most important factor for survival is an emphasis on innovation and reinvention.

-> However, innovation for the sake of it is not the goal, It is a focus on “little bets” that helps companies grow and keep up with the competition.

–>> Innovation in general is not always easy, however, especially for publicly listed companies that must balance the concerns of capital markets and shareholders, who demand quarterly profits and who are not necessarily interested in decades-long research projects.

Yet even if a company can innovate and conditions do remain favourable, immortality does have its downsides.

For instance, there is no real proof that age makes a company any more profitable than younger companies. On the contrary, evidence from the stock market actually suggests that age could be a hindrance.

Of the 74 or so companies that have stayed in the S&P 500 for more than 40 years, only a dozen or so have managed to beat the average, according to a study by consultancy McKinsey.




“companies that have embraced the “create, operate, and trade without losing control” strategy include IBM, GE and Johnson & Johnson. Each of these companies create, often by buying smaller companies in the market that are poorly positioned to expand their products internationally. They also free up capital by divesting business units at the same time–all while maintaining control of their margins and returns.”


“How fast do we have to change to maintain our position within our industry?” The pace of change varies by industry. That said, if one’s industry is changing more slowly than the pace of change in the economy, that industry itself will experience gradual “fade’ and often replacement by lower cost international competitors. Doubters only have to think about the steel, auto or paper businesses to find examples.


“mergers and acquisitions tend to increase when the economy is growing, leading to higher returns on equity.

When the market for deals and IPOs is depressed, as it has been for several years, demand for creative destruction builds up. This pent-up entrepreneurial energy and demand for deals typically gets unleashed when the economy is stronger.”


“The matrix helped companies decide which markets and business units to invest in on the basis of two factors—company competitiveness and market attractiveness—with the underlying drivers for these factors being relative market share and growth rate, respectively. The logic was that market leadership, expressed through high relative share, resulted in sustainably superior returns. In the long run, the market leader obtained a self-reinforcing cost advantage through scale and experience that competitors found difficult to replicate. High growth rates signaled the markets in which leadership could be most easily built.

Putting these drivers in a matrix revealed four quadrants, each with a specific strategic imperative. Low-growth, high-share “cash cows” should be milked for cash to reinvest in high-growth, high-share “stars” with high future potential. High-growth, low-share “question marks” should be invested in or discarded, depending on their chances of becoming stars. Low-share, low-growth “pets” are essentially worthless and should be liquidated, divested, or repositioned given that their current positioning is unlikely to ever generate cash.”


BCG used Compustat’s database (1950 to 2013) for our core data set and then matched companies with the S&P Capital IQ database to obtain additional data points. The data set includes foreign companies listed on US exchanges. 


Only the past decade saw a slight divergence in outcomes: traditionally stable oligopolies (such as the oil and gas industry) recovered the most, while mortality remains high in more dynamic industries (such as technology).


While smaller companies have always faced greater risk, even the largest companies are now facing higher exit rates. Company age only began to affect exit risk in the first decade of the 2000s, when turnover plateaued for older companies but continued to grow among younger ones.


Take the example of Compaq Computer. Founded in 1982, it went public and shot to success extremely fast. Unlike many of its peers (Altos Computer Systems, Corona Data Systems, Eagle Computer, and Osborne Computer), Compaq survived into the ‘90s, establishing itself as a serious threat to incumbents in the computer industry. Among the much older and larger incumbents it disrupted was DEC, which was sold to Compaq in 1998. Another giant, HP, saved itself from the same fate by buying up Compaq after the company struggled through the dot-com collapse. 



Yield and Credit


Hunt for Yield and Wild Goose Chases

Triple A, 30-Years? Mate, That’s a Bargain

Sinking Yuan Risks Dragging Down Banks

Silicon Valley’s Cool Kids Are Turning Into Squares


A rotation on investors’ flavour 

The demand for High dividend yield changes since the yield left the bottom.




The rotation might be another overreaction, but valuations for dividend stocks are still lofty.


Anyway, the looking-for-yield game seems not going to end

Australia’s government sold its biggest-ever bond, raising A$7.6 billion ($5.8 billion) in its first ever offering of 30-year notes (triple-A rating). The 3.27 percent yield is less than what the country paid for securities due in five years as recently as 2014, and buyers from abroad snapped up almost two-thirds of the issue.

<- That yield and credit score combination make the 30-year notes particularly attractive to global financial institutions, which have started to adopt new liquidity coverage ratios required by banking rules introduced after the 2008 crisis. [1]


Banks, however, are in the business of making money, not holding bonds for an unlikely run on deposits. So treasurers go through pains to try and squeeze some yield out of holding all those godforsaken securities. That’s become increasingly difficult since the European Central Bank and the Bank of Japan (in spite of its nation’s single-A credit grade) embraced negative interest rates.


Another negative-rate environment trait is that the average maturity of debt lengthens. There’s a body of academic work that has found a correlation between lower yields and longer government debt maturities.



While banks in China is suffering from a weaker yuan

Investors have good reason to worry as debt continues to explode and money keeps flowing out of the country. Goldman Sachs says a rising amount of capital is leaving China in yuan rather than dollars and the moves can’t be explained by market-driven factors.


While less expects stimulation on export, outflows in turn will further weigh on the yuan in a vicious cycle that could have serious effects on China Inc., especially given that the country’s corporations have taken on record amounts of foreign currency debt.


This year alone, dollar- and euro-denominated loans taken out by Chinese companies have reached $195 billion, bringing the total outstanding to about $650 billion. That’s equivalent to the total amount of subprime mortgage loans outstanding in the U.S. in 2008.

-> S&P Global Ratings said that if China’s corporate debt doesn’t stop growing, it could cost banks $1.7 trillion (the amount of extra capital they would have to raise.)

(That’s why the state just issued guidelines for how banks may swap bad debt to equity. One may argue lenders will not be happy to hold hundreds companies in sectors with little prospect of turnaround, such as steel or coal mining. But at least according to Oliver Hart, it can be an efficient way to solve debt issues.)


Start-ups also feel cold on less capital inflow

Capital going into start-ups globally has fallen for the last four quarters to reach $24.1 billion in the third quarter, a level not seen since two years ago, according to a KPMG and CB Insights report. The trend holds for early and late-stage funding rounds.


A big reason for this is that valuations are cooling for many private start-ups from the biggest such as Airbnb to smaller ones like Rocket Internet-backed Global Fashion Group.

<- Funds are loath to contribute more money to a start-up in a later round at a lower valuation since it looks bad for their own performance. So they’re pushing their companies to pay more attention to profit and cost, rather than just handing them more cash to let them pursue breakneck sales and user growth.Uber’s tactical retreat from a war of attrition in China is but one example.

However, VCs have secured promises worth billions from limited partners, such as endowments and foundations, who expect them to put money to work. [2]

-> “VCs may need to get out their comfort zones to invest in a broader palette of sectors.”




Those Basel III guidelines mean that banks need to hold enough liquid securities to withstand 30 days of the kind of liquidity crunch that was seen back then. The devil in the detail here is what constitutes liquid. And outside of good hard cash, triple-A government securities get one of the best regulatory treatments.


The life of a typical fund is between eight and 12 years, and investments are made gradually in the first half of that period.


A little more Economics Nobel


An Economics Nobel for Examining Reality

A Nobel Prize for modeling contracts

The Performance Pay Nobel



Private Prisons and Incentive Design: Critiquing Oliver Hart

Second Thoughts on This Year’s Economics Nobel Prize


Microeconomic stuff

“The research here is deep microeconomic stuff. It’s about incentives, and imperfect information, and long-term relationships. It’s about delicate strategic interactions between people who don’t know each other’s capabilities or intentions. But it’s related to lots of real-world economic issues — performance pay, mergers and acquisitions, bank lending and corporate structure.”


Modeling Contracts

“Contracts can be very hard to design if it’s hard for the person paying for a service to observe the provider’s effort. If you’re paying someone to produce a certain number of widgets, for example, then setting up that contract is quite easy.If enough widgets are produced, then the provider gets paid the agreed amount of money. But if it’s difficult to see how much effort someone is putting in, then writing a contract can be more difficult. This problem, the potential misalignment between a principal (the person paying) and an agent (the person providing the service) when there is asymmetric information (the agent knows how much work they are putting in and the agent doesn’t) is a problem that Holmström’s work centered around.

A good example of this problem is executive compensation. The shareholders of firms often want to structure the compensation of their chief executives to make sure the incentives of the CEOs are aligned with those of shareholders. This thinking leads firms to make executive compensation tied to the price of company stock. But Holmström’s theoretical work emphasizes that the contract should be based on information that is informative of the executive’s effort and performance. A company’s share price could move due to factors outside the executive’s control—say an increase in oil prices in the case of an oil executive. This thinking can also be expanded to situations where teamwork is important and therefore points toward pay that is more salary-based and less tied to share-price performance.

Hart’s work also involves contracts, but focuses more on what happens when contracts can’t be fully specified and are therefore “incomplete contracts.” The applications of his work often focuses less on contracts between individuals and more on contracts between firms. “


Modeling Contracts 2

“Suppose that you are a principal monitoring an agent who produces output. The output depends on the agent’s effort but also on noise. It wouldn’t be a very efficient contract to just reward the agent based on output since then you would mostly be responding to noise—punishing hard-working agents when the noise factors were bad and rewarding lazy agents when the noise factors were good. Not only is that unfair–if you setup a contract like this the agents will a) demand that you pay them a lot of money in the good state because they will be taking on a lot of risk and b) the agents won’t put in much effort anyway since their effort will tend to be overwhelmed by the noise, either good or bad. Thus, rewarding output alone gets you the worst of all worlds, you have to pay a lot and you don’t get much effort.

But perhaps in addition to output, y, you have a signal of effort, call it s. Both y and s signal effort with noise but together they provide more information. First, lesson – use s! In fact, the informativeness principle says you should use any and all information that might signal the agent’s effort in developing your contract. But how should you combine the information from y and s? According to some calculation you should put high weight on the factor which is relatively not noisy. [1]

When should you use absolute pay and when should you use relative pay? For example, sometimes we reward salespeople based on their sales and sometimes we reward based on which agent had the most sales, i.e. a tournament. Which is better? The great thing about relative pay is that it removes one type of noise, say economy. [2] And this work has lot of implications for structuring executive pay. [3]

But relative pay isn’t always better. If the sales agents come in different ability levels, for example, then relative pay means that neither the high ability nor the low ability agents will work hard. Tournaments work best when agent ability is similar which is why in sports tournaments we often have divisions (over 50, under 30) or rounds. Oddly, however, performance pay for executives rarely works like a tournament. As a result, CEOs are often paid based on noise.”


Some thinking 

From my view, the contract problem can be generalized to the following 4 situations:

pay fixed, high risk -> not bad but no incentive = cost is limited

pay fixed, low risk -> not efficient as no incentive

pay floated, high risk -> not efficient as luck affects = high or less pay / effort

pay floated, low risk -> best

–>> The best way is to find a signal of effort which lower the risk of misallocation and then pay floated according to it.

However, finding such a signal is difficult and according to Hart’s theory, the principle always has different expectation with the agent even though they seemed to get a consensus. -> There is always something out of expectation that they just can’t imagine.


From Coase to Tirole

“much of Hart’s recent work repudiates the importance of his most famous articles…rather the impact of a devastatingly clever, and devastatingly esoteric, argument made by the Nobel winners Eric Maskin and Jean Tirole.”


” a huge amount of economic activity (including the majority of international trade) is not coordinated via the market, but rather through top-down Communist-style bureaucracies called firms. Coase’s early answer is that something called transaction costs exist, and that they are particularly high outside the firm. That is, market transactions are not free. Firm size is determined at the point where the problems of bureaucracy within the firm overwhelm the benefits of reducing transaction costs from regular transactions._

<- Problems: 1) what a “transaction cost” or a “bureaucratic cost” is, and why they differ across organizational forms; 2) as the wonderful paper by Alchian and Demsetz in 1972 points out, there is no reason we should assume firms have some special ability to direct or punish their workers.”: 3) as Eric van den Steen points out in a 2010 AER, can anyone who has tried to order paper through their procurement office versus just popping in to Staples really believe that the reason firms exist is to lessen the cost of intrafirm activities?

Oliver Williamson: 

“some relationships generate joint rents higher than could be generated if we split ways, unforeseen things occur that make us want to renegotiate our contract, and the cost of that renegotiation may be lower if workers or suppliers are internal to a firm. It is not that everyday activities have different transaction costs, but that the negotiations which produce contracts themselves are easier to handle in a more persistent relationship.

Empirical support for informal “bureaucratic costs”: firms are larger in the developing world because weaker legal systems means more “unforeseen things” will occur outside the scope of a contract, hence the differential costs of holdup or renegotiation inside and outside the firm are first order when deciding on firm size.”

Hart and Grossman(1986):

What really makes a firm a firm is that it owns assets. Contracts may be incomplete – at some point, I will disagree with my suppliers, or my workers, or my branch manager, about what should be done, either because a state of the world has arrived not covered by our contract, or because it is in our first-best mutual interest to renegotiate that contract. There are relationship-specific rents, but the owner of the asset is in a much better bargaining position when this disagreement occurs.Therefore, the owner of the asset will get a bigger percentage of rents after renegotiation. Hence the person who owns an asset should be the one whose incentive to improve the value of the asset is most sensitive to that future split of rents.” [4]

Hart and Moore(1990)

“extends this basic model to the case where there are many assets and many firms, suggesting critically that sole ownership of assets which are highly complementary in production is optimal. Asset ownership affects outside options when the contract is incomplete by changing bargaining power, and splitting ownership of complementary assets gives multiple agents weak bargaining power and hence little incentive to invest in maintaining the quality of, or improving, the assets.”

(There are, of course, many other theories of the firm. [5])

Maskin and Tirole (1999)

“Incomplete contracts theories: there are always events which are unforeseeable ex-ante or impossible to verify in court ex-post, and hence there will always scope for disagreement about what to do when those events occur. But, agent don’t care about anything in these unforeseeable/unverifiable states except for what the states imply about our mutual valuations from carrying on with a relationship. Therefore, every “incomplete contract” should just involve the parties deciding in advance that if a state of the world arrives where you value keeping our relationship in that state at 12 and I value it at 10, then we should split that joint value of 22 at whatever level induces optimal actions today. Do this same ex-ante contracting for all future profit levels, and we are done. Of course, there is still the problem of ensuring incentive compatibility – why would the agents tell the truth about their valuations when that unforeseen event occurs?

-> “Incomplete contracts don’t matter if we can truthfully figure out ex-post who values our relationship at what amount, and there are many real-world institutions like mediators who do precisely that.”
->> “If, as Maskin and Tirole prove (and Maskin described more simplyin a short note), incomplete contracts aren’t a real problem, we are back to square one – why have persistent organizations called firms?”


Incentive Design

Should the management of prisons be contracted out to the private sector? The owners of a private firm have a strong incentive to cut costs and improve productivity because they get to keep the resulting profits. If a public prison cuts costs, there is more money in the public treasury but no one gets to buy a yacht so the incentive to cut costs is much weaker.

Hart say this profit motive is the problem! Suppose that we care about costs but we also care about prisoner rehabilitation, civil rights, and low levels of inmate and guard violence.What we pay for is cheap prisons, but what we want is cheap but high quality prisons. If we can’t measure and pay for quality, then strong incentives could encourage cost cutting at the expense of quality.

The principle is a general one, a strong incentive scheme that incentivizes the wrong thing can be worse than a weak incentive scheme. One car dealer in California advertises that its sales staff is not paid on commission. Why would a store advertise that its sales staff do not have strong incentives to help you? The answer is clear to anyone who has tried to buy a car. High-pressure dealers who pounce on you the moment you enter the showroom and bombard you with high-pressure sales tactics may sell cars to first-time buyers, but the strategy is too unpleasant to win many repeat customers. Car dealers who rely on repeat business usually prefer a low-pressure, informative sales staff….In theory, a car dealer could have strong incentives and repeat business by paying its sales staff based on their “nice” sales tactics, but in practice it’s too expensive to monitor how salespeople interact with clients. Cheating by the sales staff would be difficult to detect and thus would be common. 

1. Are HSV correct that weak-incentive public prisons are better than strong incentive private prisons? Not necessarily. HSV assume that cutting quality is the way to cut cost. But sometimes higher quality is also a path to lower costs.

2. HSV may also underestimate how well quality can be measured. Measuring output pays off more when incentives are high. Unsurprisingly, therefore, private prison companies and government purchasers have made extensiveefforts to measure the quality of private prisons.

3. Finally, don’t forget that weak incentives reduce the incentive to cut costs but they don’t increase the incentive to produce high quality! Public prisons might use their slack budget constraints to offer high-quality rehabilitationprograms, or they might instead offer prison guards above-market wages. Which do you think is more likely?


Finally, I just find someone goes even further than my comment on last blog.

“Think of their work as consisting of three steps.

1. Identifying some real-world complexities that affect how businesses operate. For example, output may result from both effort and luck. Output may be joint. A worker’s job description may include more than one objective.

2. Construct a mathematical optimization model that incorporates such complexities.

3. Offer insights into designing appropriate compensation systems, including when to outsource an activity altogether.

In the eyes of the mainstream economics profession,step 2 is extremely important. Without it, you either do not get to step 3, or your claims in step 3 lack reliability and credibility. Step 2 is why Hart and Holmstrom earned the Nobel Prize.

In my view, step 2 is unnecessary. If anything, it tends to get in the way, often creating a barrier to doing step 1 properly, because economists limit themselves to what is mathematically tractable. I think that Hart and Holmstrom sometimes (often?) made good choices in step 1, and that is what accounts for the value of where they arrived at in step 3.

I offer a number of asides that go from step 1 to step 3 directly. I think in terms of a dynamic process of trial and error. A manager tries an approach to compensation. As long as it seems to work, it persists. Once it gets gamed too much by the employees, something happens–the manager makes changes, the manager gets fired, or the firm goes out of business. Another point is that I believe that managers closer to the problem do a better job of solving it.”

For example

“Think of the firm as a team in which the output of any one individual is difficult to value. Consider a computer programmer working on part of a bank’s software system. No one can state precisely the value to the bank of the particular section of code that the programmer works on. All that we know is that the bank cannot pay programmers too much, or else it would be unable to make a profit. and it cannot pay programmers too little, or they would choose to work elsewhere.

If it is possible to attach a precise value to a particular segment of work, then it is possible for that work to be broken out of the firm and outsourced to the market. Thus, if a bank can assign a precise value to a particular software system, it has the option of contracting with an outside firm to build the software for an agreed-upon price.

In short, when the value of different tasks can be isolated, specialization will tend to take place between firms, coordinated by the price system. When the value of a particular task is difficult to measure, because its value varies a great deal depending on how it is combined with other tasks, specialization will tend to take place within a firm, governed by instructions.”


I doubt if one could get to step 3 directly without using some abstract toolkit like math, and you can not just let the market tell you everything. But I do agree that economists should move some focus out of step 2.



“Suppose you write a contract where the agent is paid a wage, w=B0+By*y+Bs*s where Bo is the base wage, By is the beta on y, how much weight to put on output and Bs is the weight on the s signal–think of By as the performance bonus and Bs as a subjective evaluation bonus. Then it turns out you should weight By and Bs according to the following formula:

c is a measure of how costly effort is to the agent and so it also makes sense that the higher is c the less weight you put on performance incentives and the more on the base wage.

r is a measure of risk aversion for the agent. When r is zero:

(which means the agent is risk neutral and in the world where you put all the risk on the agent.

If r>0 then you don’t want to put all the risk on the agent because then the agent will demand too much so you take on some risk yourself and tamp down By and Bs and instead increase the base wage which acts as a kind of insurance against risk.)

-> It says that you should put a high weight on y when the s signal is relatively noisy and a high weight on s when the y signal is relatively noisy.”


“Suppose, for example, that sales depend on effort but also on the state of the economy. If you reward based on absolute sales then you are rewarding a lot of noise. Once again, that has two bad effects it means that you have to pay your agents a lot since you are imposing risk on them and it means that they won’t work that hard since they know they will be paid a lot when the economy is good and hardly at all when the economy is bad so in neither case do the agents have strong incentives to work hard. Suppose, however, that you have a relative pay scheme, a tournament. Now you have removed the noise coming from the state of the economy–since all the salespeople face the same economy and since there is always a first, second and third place the agent’s now have an incentive to work hard in good or bad times. Not only do they have an incentive to work hard you don’t have to pay them much of a risk premium since more of their pay is now based on their own effort rather than on noise.”


“In particular, executive pay often violates the informativeness principle. In rewarding the CEO of Ford for example, an obvious piece of information that should used in addition to the price of Ford stock is the price of GM, Toyota and Chrysler stock. If the stock of most of the automaker’s is up then you should reward the CEO of Ford less because most of the gain in Ford is probably due to the economy wide factor rather than to the efforts Ford’s CEO. For the same reasons, if GM, Toyota, and Chrysler are down but Ford is down less then you might give the Ford CEO a large bonus even though Ford’s stock price is down. “


Baker and Hubbard (2004) provide a nice empirical example: when on-board computers to monitor how long-haul trucks were driven began to diffuse, ownership of those trucks shifted from owner-operators to trucking firms. Before the computer, if the trucking firm owns the truck, it is hard to contract on how hard the truck will be driven or how poorly it will be treated by the driver. If the driver owns the truck, it is hard to contract on how much effort the trucking firm dispatcher will exert ensuring the truck isn’t sitting empty for days, or following a particularly efficient route. The computer solves the first problem, meaning that only the trucking firm is taking actions relevant to the joint relationship which are highly likely to be affected by whether they own the truck or not. In Grossman and Hart’s “residual control rights” theory, then, the introduction of the computer should mean the truck ought, post-computer, be owned by the trucking firm. If these residual control rights are unimportant – there is no relationship-specific rent and no incompleteness in contracting – then the ability to shop around for the best relationship is more valuable than the control rights asset ownership provides.”


“The idea that firms in some industries are big because there are large fixed costs to enter at the minimum efficient scale goes back to Marshall. The agency theory of the firm going back at least to Jensen and Meckling focuses on the problem of providing incentives for workers within a firm to actually profit maximize;More recent work by Bob Gibbons, Rebecca Henderson, Jon Levin and others on relational contracting discusses how the nexus of self-enforcing beliefs about how hard work today translates into rewards tomorrow can substitute for formal contracts, and how the credibility of these “relational contracts” can vary across firms and depend on their history.”


Oliver Hart and papers on related topics


Oliver Hart, Nobel Laureate

Economics Nobel Rewards Theories Worth Building On


The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration

Property Rights and the Nature of the Firm





The following is the introduction about Oliver Hart’s papers by .

1. The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration

When one company should buy out the assets of another company?
When do mergers and acquisitions maximize business value?

(This paper is a  starting point for thinking about mergers, vertical integration, and other questions of corporate ownership and contract and control.  Bengt Holmström of all people wrote a very nice appreciation of the paper.)

Hart was able to figure out how ownership transfers influence earlier decisions to invest in the value of company assets.

For instance, if Bayer buys out Monsanto, the incentives for the former managers to add value may go up, and for the latter managers the incentives may go down. The success of the merger may depend on whether the gain here outweighs the loss.

-> if you own an asset outright, you keep a greater share of the proceeds from improving the value of that asset.  Ownership should thus migrate to those parties who have the greatest ability to improve value.

–> that is a very fundamental improvement on the Coase theorem, which suggests ownership won’t matter when there is ex post contractibility. 

—-> Hart showed that for ownership not to matter there must also be ex ante contractibility about value-improving investments at earlier stages in the game, an unlikely assumption to hold.

And Hart helped devise a technical language for analyzing when too many potential veto points in a business deal can block progress.


2. Property Rights and the Nature of the Firm

This is again a model and series of parables about ownership and the allocation of rights, but with some twists on the earlier Grossman and Hart piece.

-> The key point is to not allow inessential agents to achieve blocking power of value creation. 

->> If inessential agents owns potential blocking power, the surplus has to be split, resulting in some loss of value, due to a tougher bargaining problem, higher transactions costs, and a chance there won’t be enough surplus to cover the most significant investments

->>> Parties who create a lot of value should own things.


3. Takeover Bids, the Free Rider Problem, and the Theory of the Corporation.

(One of Alex’s most interesting papers is an extension of this work)

Why a lot of value-maximizing takeover don’t happen, or why it is hard to buy up a whole city block and renovate it.

-> Maybe not enough shareholders will sell as they hopes other would sell and the raider would push the value, and so a value-enhancing takeover doesn’t always happen.


4. Incomplete Contracts and Renegotiation

(This paper is connected to the Nobel Prize for Jean Tirole two years ago.)

How can you write a contract so a) parties will make the appropriate relationship-specific investments, and b) it doesn’t have to be renegotiated all of the time?

–>> value maximization within corporate endeavors and possible obstacles to such value maximization.


5The Proper Scope of Government: Theory and an Application to Prisons

Hart, again with co-authors, also wrote a seminal paper on when we should prefer government over private-sector ownership.

<- Most of us prefer to eat in private rather than government-owned restaurants because we believe we’ll get lower costs, tastier food, and more innovation.
<- At the same time, private prisons may not be such a great idea. Prison companies will try to cut costs (“The incentive to cut costs is too strong! “), but the result may be facilities that are insufficiently humane.

–>> Sometimes the apparently inefficient bureaucracy does a better job helping to meet social goals, because the government won’t have the same profit incentive to skimp on quality along various margins.

(You also probably wouldn’t want Air Force One owned by the private sector, though you do want it to be designed and produced by the private sector. )


6. An Analysis of the Principal-Agent Problem

A breakthrough and highly rigorous means of modeling the principal-agent problem.

(It is in Econometrica and quite hard for many people to read.)


7. On Shareholder Unanimity in Large Stock Market Economies

Whether all shareholders will desire that firms maximize profits if markets are incomplete and some firm shares also serves secondary “insurance” purposes of helping protect against adverse states of the world.

-> If you use the shares of a wheat-producing firm for insurance purpose, you would hope the value of the firm covary with the value of wheat in ways that differ from simple firm profit-maximization.


Congratulations to Oliver Hart!

一个非主流经济学学生的感想:Oliver Hart的得奖是自Ronald Coase和Oliver Williamson后的第三次颁给新制度经济学。新古典主义在这几十年用一种数学建模吞并的方式宣称了自己的霸权,认为只有数学才能解释和验证阐述世界的理论,认为建模即主流,拒绝即无稽。所以他们可以把这次的得奖称为主流中的主流,并在宏观经济学饱受争议之后让理论架构的微观经济学再一次成为自己的旌旗。但是不可否认的是,理论架构建模之前,并不是归纳法或者演绎法创造了这一切的源泉,而是真正贴切世界的敏锐的抽象的思考和发现。就像爱因斯坦所说,抵达真理的方向上并不是逻辑的道路,而是对经验共鸣的直觉。


Increasing Short-termism?


Maybe Companies Aren’t Too Focused on the Short Term

Evidence and Implications of Short-termism in US Public Capital Markets: 1980-2013

A Long Look at Short-Termism Questioning the Premise


Last week when we went through Montier’s criticism about shareholding value maximization, one thing we mentioned is “given the shortening lifespan of a corporate and the decreasing tenure of the CEO, many managers are willing to sacrifice long-term value for short-term gain”. The most fact that Monteir found is “declining and low rates of business investment.”

However, this week Tyler Cowen wrote an article to defend the short-termism.

First of all,

“Short-termism is said to plague all parties in the investment community, including investment managers, companies, and investors. However, it is very difficult to prove.”

However, a recent paper claims to “provide evidence of increasing short-termism in US equity capital markets over the period of 1980-2013, by using a ‘market discount factor’ estimated for publicly traded firms based on a capital asset pricing model”. (Which means it’s also a study for valuation.)  They find that “markets more heavily discount firms that have less financial slack, spend less on capital or R&D, have greater analyst coverage, or held by more transient institutional investors as well as pay their executives via more short-term compensation.And “short-term market valuations are significantly negatively correlated with future capital investment. “


However, Cowen argued that American corporations ought to be on short-term.

The following reasons are also combined with ideas from Michael J. Mauboussin’s report.


1. In information technology, betting on the future feels imprudent if change is rapid.

Academic research shows that the period of competitive advantage is shrinking for companies.


The world is speeding up. If the sustainability of a company’s economic profit is fleeting, there is less reason to place great value on the future. Planning so far out can involve a lot of expense and risk.

2. In information technology, the average life of a corporate asset is lowering.

“Production has shifted toward service sectors with relatively short asset lives, and that may call for a shorter-term orientation in response.”


Except rapid change, the more actual effect on investment comes from balance sheet, since shorter asset lives means higher depreciation.

-> So for many companies a contraction in time horizon is a proper response to economic reality.

And this is backed by “corporate governance tends to be better in sectors where asset lives are long than in sectors where asset lives are short. Where monitoring long-term investments is most relevant, corporate governance is the best developed. “


3. Companies tend to make big mistakes from thinking too big and too long-term

“for instance, a lot of mergers were based on notions of long-run synergies that never materialized.”

4. Given to high endurance of losing money by startups, investors are not ignoring the long-run prospects of the company.

“Amazon has a high share price even though its earnings reports have usually failed to show a profit .Many tech startups have high valuations even though revenue is zero or low.”

“During the dot-com bubble of the 1990s, there was too much long-run, pie-in-the-sky thinking and not enough focus on the concrete present.”

5. Compensation schemes for managers are more complex and more varied than in the past.

“Executive compensation has moved toward long-term incentives, boards of directors are more independent than in the past, and governance committees are “nearly universal.”

Moreover, a link between pay and short-termism is difficult to establish.

<- Academic research shows that CEO pay has closely followed the size of the firms in the economy independent of the form of remuneration.

6. Transient investors are not increasing.


And “while transient investors do take a short view, they are attracted to companies that provide lots of information events. It is these companies that appear most willing to trade value-creating investments to deliver short-term results.”


Anyway, “If public shareholders are placing too much short-term pressure on their companies for a good quarterly earnings report, companies have the option of boosting their value by going private, as has been the trend.”

<- By 2012, the number of U.S. public corporations was less than half what it had been in 1997, in part because many companies went private. See here.


Update 2017/10/21

Are U.S. Companies Too Short-Term Oriented? Some Thoughts Steve Kaplan University of Chicago Booth School of Business