Photo by Ernie McClellan / CC BY
The number one risk that we are facing as stock investors is the death of a company we are invested in. Often this kind of risk is discussed in relation with backtests of different strategies and the impact of survivorship bias. However, I found that an analysis of probabilities and implications for real-life stock portfolios was not readily available. Thus, I went on a fact finding mission to get a better understanding of my risk exposure covering the following questions:
- How big is the overall risk?
- What are the implications for a stock portfolio?
- Is the risk of failure related to a company's age and, if so, how?
Happy and less happy endings
An easy way to scare stock investors is to remind them of the great churn among listed companies. You may have heard a couple of times that life expectancy of companies is lower than you think respectively declining, e.g. from BCG and their study Die Another Day or even Nassim Taleb:
Of the five hundred largest U.S. companies in 1957, only seventy-four were still part of that select group, the Standard and Poor's 500, forty years later. Only a few had disappeared in mergers; the rest either shrank or went bust.
Unfortunately, Taleb and even more BCG stop the movie before we get to know about the ending's details. Parts of the audience do not mind skipping the final scene in anticipation of the discussion about the aftermath of these troubling findings, but let us press rewind to learn more about the delisted companies' whereabouts. There is comprehensive data available. Baker and Kennedy figured out
Takeover delistings are far more common than delistings due to financial distress .
In their sample of 3,954 delistings between 1965 and 1995 66% of the companies were taken over and 19% ceased due to financial distress. More recent studies confirm this finding . It is a minority of cases where shareholders have to write off their investment. Quite possibly they benefit oftentimes.
The base rate
On my quest to find the base rate of delistings due to financial failure, I browsed through a couple of studies using different sets of data leading to slightly different results. Therefore, I took the somewhat brave decision to calculate the base rate based on my own desktop research so that I could stay on top of the sources, knowing full well that the academics have access to much better data than I.
I took the total number of companies listed on U.S. stock exchanges between 1991 and 2014 from the Worldbank's database which in turn is fed by the World Federation of Exchanges. A cross-check with SEC data confirmed that the population matches the total of companies listed on the NYSE and the Nasdaq National Market.
The number of companies that filed bankruptcy under either Chapter 7 or Chapter 11 could be taken from the UCLA-LoPucki Bankruptcy Research Database (BRD) which, however, does only include companies with assets worth $100 million or more, measured in 1980 dollars. Obviously, this subset is not fully compatible with the population taken from the Worldbank database. Therefore, I estimated the share of companies above that threshold based on SEC data and eliminated the smaller companies.
Unsurprisingly, the number of bankruptcy cases fluctuates substantially by year with peaks marking the end of the dot.com frenzy as well as the financial crisis.
The chart shows data taken from the BRD as well as from BankruptcyData.com with the latter including all publicly traded companies, i.e. even micro caps trading over-the-counter.
Combining the BRD data with the adjusted Wordbank data, I found that the percentage of companies that go bankrupt had varied between approximately 0.2% (1994) and 3% (2009) with the average being 0.96%.
A back-of-the-envelope model
Don't let yourself fool by the decimal places. These are rough estimates, not accurate figures. Still, I could not resist doing some projections to gain a better understanding of the potential implications of this base rate.
Let's assume your investment horizon in the stock market is 20 years, which happens to be about the time by which I could retire. Our investment universe is a stock picker's nightmare in that any stock will return 10% p.a., so that $1,500 invested today will turn into some $10,000 by the end of the 20 years. This of course will only happen if the company in question manages to avoid going bankrupt on the way. The base rate of 0.96% can be seen as a negative lottery prize that is awarded to a couple of unlucky stock investors year in year out. Thus, this seemingly small number cumulates to 19.2% by year 20. Let us now take a look how portfolios with different degrees of diversification perform in this set up. The following histograms show the distribution of returns after 20 years for 100 simulation test runs of three portfolios where the initial investment of $1,500 is spread over 1, 20 and 100 stock positions respectively and the stocks are never sold:
The histogram for the one-stock-portfolio shows how Marc Cuban may have come to his conclusion that diversification was for idiots only: Your odds to take home the full $10,000 after 20 years are maximized if you suffice with one stock position. And surely, if successful, this move will make you look clever in hindsight. That is a big if, though, because about one out of five investors will lose his shirt trying this stunt.
Further, we can see how the deviations from the expected value (=$10,000 x (100%-19.2%) = $8,080) diminish the more diversified the portfolio is. Bankruptcy risk cannot be diversified away, but diversification increases the predictability of the outcome. For a broadly diversified ETF the 0.96% can be seen as an annual fee to the Reaper, leading exactly to the expected value by the end of the period.
Staying power -- is there such a thing?
Moving away from our model universe of uniform stocks, we find that in the real world no two companies are the same. Academics have spent a lot of time on the development of bankruptcy prediction models trying to find the magic metrics that would allow us to identify the companies that will go belly up next. I have seen some very impressive formulas during my research going way beyond anything I could describe, let alone handle. At the same time, bankruptcy prediction is already a by-product of any investor's due diligence. I even suspect a hands-on due diligence can work just as well as a complex prediction model, since both suffer from the same lack of access to tomorrow's data. Still, one can wonder if there is any evidence for staying power of certain companies.
The answer is not straight forward. The first problem comes with the question. Life expectancy is less relevant to shareholders than eventual outcome. As we have seen, these can be two different things. The second problem is that academic studies usually define a company's life as the time span between listing and delisting without considering how long a company has been around before. Daepp et al, for example, identify in their huge sample of 26,000 only 160 companies that managed to stay alive through the whole period under observation between 1950 and 2009 . At the same time the number of companies qualifying for the NYSE Century Index of 100+ year old companies is 372, i.e. more than twice as high.
By the way, longevity as a concept for an ETF proved to be short-lived. Invesco (NYSE:IVZ) launched an ETF on the NYSE Century Index in 2014 and retired the concept after less than 1½ years. The underlying idea has its merits, though, as there are two findings supporting the view that there can be safety in company age:
- IPOs are a risky business for investors. If an IPO fails, it fails quickly in most cases -- within some 5 to 6 years on average. The risk that a company goes bankrupt in its first five years post IPO is approximately 25% .
- The only paper that I came across considering the incorporation date when referring to company age found that there actually is a link between age and financial failure:
The negative relation between failure hazard and company age is consistent with the hypothesis that firms learn how to avoid bankruptcy as they get older. […] The results suggest the existence of a monotonically declining relation with age in the case of financial distress […] .
Two key properties of death are certainty and fear. In investing, we can avoid the certainty part by means of portfolio concentration. The trade-off is that this puts our principal at risk which adds significantly to the fear side of the equation. Therefore, you may want to prefer to play it safe.
There is also the option to trade fear in for certainty by means of diversification. An appropriate level of diversification is difficult to obtain for DIY investors that hold individual stocks. Admittedly, my portfolio of 23 stocks does not look competitive in this respect when compared to a broadly diversified ETF.
The average age of the companies in my portfolio is 42 as measured since incorporation date. However, a lot of fuzziness about a company's age remains, e.g. if business model and ownership structure changed radically or the company emerged from spin-offs, mergers and the like. Risk-conscious investors should stay away from IPOs as a measure of precaution. Otherwise, paying attention to a company's age since incorporation will not hurt either. It will not make any investment a sure thing, though. The case of Kongō Gumi, a Japanese company that closed in 2006 after 1,428 years of operation, warrants caution.
 Baker GP, Kennedy RE, 2002, Survivorship and the Economic Grim Reaper.
 Daepp MIG, Hamilton MJ, West GB, Bettencourt LMA, 2015, The mortality of companies.
 Loderer C, Neusser K, Waelchli U, 2010, Firm Age and Survival.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.