What the Sydney Harbour Bridge Taught Me About Investing

Author's Avatar
Jun 13, 2008
Something took place before the opening of the Sydney Harbour Bridge that has affected me to this day. It taught me the importance of stress testing before making any important decisions, particularly before deciding on any transactions in the stock market. This led me to develop a new rigorous approach for stress testing stocks before making any purchases.


The Sydney Harbour Bridge is one of the most famous and photographed bridges in the world. Affectionately dubbed "the coat hanger" because of its distinctive shape, it was first opened to public traffic on March 19, 1932. Just before the ribbon was officially cut, Captain Francis De Groot of the para-military group, the New Guard, galloped in on a borrowed horse and slashed it with his sword. This caused great amusement amongst the huge crowd and has become part of Australian folklore.


But something took place before the opening that had much more significance for me. In the month prior to the opening, its safety was given a practical test with ninety-six locomotives laid end to end along the railway tracks on both sides of the bridge. In engineering terms, the bridge was given a stress test by applying a load that was many times greater than anything that would be expected under normal operating conditions. Passing the test removed the fear of using the bridge because it might buckle or even collapse. The sight of these dark locomotives lined up under the massive arches of the bridge is an image that has stood in my mind ever since I first saw photos of this event as a small boy. Since then I have always wanted to test things thoroughly before taking the next step.


Test Locomotives Lined Up on the Sydney Harbour Bridge: 10 February, 1932


This doesn't mean that I don't like trying new ventures. Quite the opposite; many things I have done may be considered risky by some. Over my life I have spent time as a rock climber and mountaineer. I have competed in an iron man triathlon, run in the New York marathon, and recently went sky diving as a birthday present from my children. It is just that before I do these things, I check as thoroughly as I can everything that can go wrong and work with the biggest safety factor that I can. If the level of safety does not meet my requirements, then I won't go ahead.


Margin of Safety


Because of my habit of checking safety levels before starting any enterprise or activity, I was excited when I went to my first annual meeting of Berkshire Hathaway in the early 1990s and heard Warren Buffett talk about "margin of safety," an expression he explained that came from Benjamin Graham. I immediately thought of this concept as applying a stress test before making an investment.


After the meeting I went to Graham's monumental book Security Security first published in 1934 where I saw that he frequently used this expression. Graham continued its use in The Intelligent Investor, declaring that "margin of safety" is the central concept of investing and the secret of sound investing distilled into three words.


You can't get better recommendations than this: Benjamin Graham and Warren Buffett. Graham is a legendary analyst and writer often referred to as the Dean of Wall Street. According to Jason Zweig in his introduction to a recent edition of The Intelligent Investor, "Graham was not only one of the best investors who ever lived; he was also the greatest practical investment thinker of all time. Before Graham, money managers behaved much like a medieval guild, guided largely by superstition, guesswork, and arcane rituals."


The second person, Warren Buffett, is the chairman and CEO of Berkshire Hathaway and was a former student and employee of Graham. When Buffett first gained control of this company the stock was trading at around $7.00. Today it is over $130,000 making him both the world's greatest investor and its richest person.


The Two Factors Behind Rises in Stock Prices


The problem was that when I saw the expression "margin of safety" it was often in a vague setting. I wanted the concept to to be structured with a firm intellectual foundation having the same sort of rigor as applied to the Sydney Harbour Bridge and other civil engineering projects.


The first thing I noticed was that for the price of a stock to rise, one of two things must happen: either the earnings per share (EPS) must rise or the price-to-earnings ratio (PE ratio) must rise (or both). This has to be the case since price is equal to the PE ratio multiplied by the EPS.


The PE ratio is the amount that the market is willing to pay for every dollar of earnings per share in the company. It varies enormously between companies; in the case of members of the Dow Jones Industrial Index, it ranges from around 11 with companies such as Exxon Mobil to around 27 for McDonalds Corporation with an overall average of approximately 17 (not counting companies with negative earnings)


The other component of price movement is the growth of EPS. For example, if the PE ratio does not change, then the percentage change in price will, over time, be precisely the same as the percentage change in EPS.


Let's look at some calculations taken from the Scenario Analysis feature in Conscious Investor starting with a simple case where the PE ratio remains fixed and the earnings grow. Suppose that earnings per share grow by 15% per year over the next 5 years. It is follows that the stock price will grow by an average of 15% per year over the next 5 years. Now suppose that the PE ratio grows by 20% over 5 years but there is no change in earnings. This time the stock price will grow by an average of 3.71% per year over this period.


More complicated is the case where both EPS and the PE ratio grow. Suppose that the earnings grow by 15% per year over the next 5 years along with an overall growth of 20% for the PE ratio. To get the growth in the stock price it is not just a question of adding the two separate growths of the EPS and PE ratio which would be 15% + 3.71% = 18.71%. The growth in the EPS and the PE ratio combine so that the average growth in the stock price is 19.27%. In the other direction, if the PE ratio is anticipated to drop by 20%, the stock price will now only grow by an average of 9.98% per year over the next 5 years despite the eps growing at an average of 15% per year.


More complicated still is the case where the company pays dividends because they would have to be taken into account to calculate the total return, capital gains plus dividends. For simplicity, however, in this article we assume that the company does not pay dividends.


Since the aim is to establish a rigorous foundation for a margin of safety, it is now clear that it has to involve both the PE ratio and EPS. Interestingly, these are the subjective and objective components of stock prices.


Subjective and Objective


The two threads that run through all observations and analyses that we make in our lives are the subjective mode and the objective mode of viewing the world. Science, for example, is the great champion of the objective mode. It continually strives to design instruments and procedures that "filter" out individual opinions and responses (that is, subjectivity) so that we can see the "true" nature of what is being studied (that is, its objective properties). On the other hand, the arts want to give full scope to subjectivity so that we can directly experience the "inner" nature of what is being viewed or experienced. Paintings by Pablo Picasso are penetrating examples of an artist paying little regard to any restrictions imposed by objectivity. Yet his paintings have given pleasure to millions of people in countries throughout the world.


We may think that the stock market is the pinnacle of objectivity. After all, in the end it comes down to dollars and cents, to hard currency. But it is a different matter when we are talking about the decision process involved in making successful investments. Here it seems that what is most important is acknowledging that both subjectivity and objectivity have partnering roles. Consider the two components of the share price that we considered above, the PE ratio and EPS.


To a large extent the PE ratio is subjective since it depends on the perception of the market to the future success of the business. This really hit home to me some years back when I would track pairs of companies that would have very similar financial ratios, operating histories, and even earnings forecasts. Yet often one company in the pair would routinely trade with its PE ratio in the mid teens and the other in the mid 20s. It was difficult to find any objective reasons for this difference.


On the other hand, current EPS and EPS growth is predominately objective since it is calculated straight from the financial statements.


Of course the PE ratio is not 100% subjective since it is influenced by the actual performance of the business and the EPS is not 100% objective since, even under today's stringent accounting requirements, accountants have some flexibility in the way they calculate and present the company accounts. Nevertheless, it is useful to think of growth in stock price as a combination of subjective factors quantified through PE ratio levels and objective factors quantified through EPS levels.


Coming back to margin of safety, from the above we see that there are are only two areas that can go wrong for an investment. (Remember, I am still assuming that there are no dividends.) Either the PE ratio decreases (or doesn't grow as much as anticipated) or the EPS decrease (or don't keep growing at the pace that was anticipated). This means that we only need to consider the margin of safety in these two areas.


To do this I developed two new functions for use in Conscious Investor called PESAFETY for safe estimates of the PE ratio and ESAFETY for safe estimates of growth rate of earnings. The actual calculations are quite complex but it is possible to give an outline of how they work.


The function PESAFETY uses the current PE ratio and the average of the PE ratios over the previous four financial years. Starting with the current PE ratio, if the historical PE ratio is lower, then the calculations use this result to lower the estimate. If not, the estimate stays with the current PE ratio. Finally, a mean reversion is used in the downward direction as an extra margin of safety.


Turning to earnings, analysts whose job it is to make forecasts of earnings have a difficult task. Numerous studies show that there is no significant correlation between their 5-year forecasts and the actual growth rate. In other words, they don't add any value to the process of forecasting earnings over 5 years.


When you look at companies such as Ford Motor Company you can see just how difficult it can be to make such forecasts. (See the accompanying chart.) The historical performance is so variable that it is impossible to have any confidence in any forecasts no matter who made them.


As prudent "margin-of-safety" investors we can sympathize with them but we don't need to get caught up in their problems. As an investor, it is less important to make an accurate estimate of growth rate than it is to make an estimate about which we can be confident that the actual growth will equal or exceed it. In other words, we are looking for a worst case scenario or something close to it.


The safety level for the forecast for earnings per share calculated by ESAFETY uses the historical growth rate of earnings per share, the historical growth rate of sales per share, and the stability of earnings per share to make a conservative estimate of the future growth rate of earnings per share.


It is fairly obvious why the first input, the historical growth rate of earnings per share, is used since a simple future forecast is that the growth rate in the future is the same as in the past.


For the second input, growth of sales per share, there are two options; either it exceeds the growth rate of earnings per share or it does not. Consider the second case where sales per share growth is lower than earnings per share growth; this can act as a drag on future growth of earnings and so we use the growth rate of sales to lower the estimate of the growth rate of earnings. In the opposite case where the growth rate of sales exceeds the growth rate of earnings we give less importance to the sales component.


Regarding the third input, namely the stability of the growth rate of earnings, the higher the stability, the better historical growth rates act as a forecast for future growth rates. To integrate this finding into forecast, the lower the value of stability as measured by another proprietary function STAEGR, the lower we make the forecast.


Finally, over and above what was just described, we incorporate an overall margin of safety. This is in two parts: mean reversion in the downward direction and an absolute reduction.


As I said above, the actual functions are quite complex and all that is possible in a general article is to give an idea of how they work. Their strength and usefulness follows from two facts. Firstly, they are based on sound investment principles. Secondly, they have been thoroughly back tested. The outcome is that they provide forecasts that are low enough to provide a hefty margin of safety but not so low as to be meaningless.


Example of an Automatic Margin of Safety


Consider Abercrombie and Fitch (ticker: ANF), the specialty clothing manufacturer and retailer. We will apply the automatic margin of safety using the functions ESAFETY and PESAFETY. (All the calculations are taken from Conscious Investor. Other functions used are STRETD for calculating the average annual return based on various inputs and HGROWTH for calculating past growth rates based on best-fit calculations.)


Over the past 5 years its growth rate of EPS was 24.96% and its average PE ratio was 18.05. Its current dividend payout ratio is 13% which we will assume continues for next 5 years. Assume that the current price is $68.33 and the EPS over the trailing 12 months is $5.24. This means that the PE ratio over the trailing 12 months is 68.33 / 5.24 = 13.04.


If earnings continue to grow at 24.96% per year for the next 5 years and the PE ratio is 18.05, then the average annual return will be 34.32% as calculated by STRETD. This would be an outstanding result. However, even though it is based on the assumption that historical parameters will continue into the future, it would not be prudent to invest in ANF without "stress testing" the return forecast.


This is done via the functions ESAFETY and PESAFETY. When applied to the historical data for ANF they give a forecast of 15.97% for the growth rate and 11.55 for the PE ratio. By chance the drops from historical levels in both inputs are the same, around 36%. In other cases the drop is much larger, depending on the fundamental data of the company as described above. The important thing is that we have applied a rigorous and systematic technique to stress test the historical levels. With these new inputs, STRETD gives a forecast return of 14.46% per year over the next 5 years.


Of course, these are just automatic margins of safety and cannot take into account all the information that you may have gathered about the company. Nor can they allow for an investor's own risk threshold. However, they provide a carefully calculated reduction that can be modified as needed. For instance, if you are not confident that ANF will grow by an average of 15.97% per year over the next 5 years, then it is easy to put a lower value into the calculations in Conscious Investor to get more conservative outputs. The same applies to the PE ratio and other inputs.


Stress testing stocks before purchasing


Construction companies apply stress tests to civil engineering projects to ensure that they will be safe under extreme operating conditions and perform as expected. In some cases it is done through careful computer modeling and design. In others it is done through scientifically controlled experiments by actually overloading the structure as was done with the Sydney Harbour Bridge using rows of locomotives.


The same is now possible with stock market investments. Instead of loading them up with locomotives, we apply the functions ESAFETY and PESAFETY which have been carefully developed using sound investment principles, mean downward reversion, and extensive back testing. They are applied through a Safety button in the Scenario Analysis of Conscious Investor that implements these two functions and recalculates expected performance with an automatic margin of safety. The default expected performance assumes that both the past economic performance of the business and its market level in terms of the PE ratio will continue into the future. Once the Safety button is selected the ESAFETY and PESAFETY functions are activated as a thorough "stress test." Passing the test requires that the new expected return is still at a healthy level even though the earnings grow markedly slower that in the past or that the PE ratio is considerably lower than its current and past levels.


For stocks that pass the test, it is like locking in a worst case scenario but leaving the upside open for greater profit. In other words, it is like purchasing at a price that locks in a reasonable return but leaves everything open for much greater profit. It also helps remove the fear associated with many investments because you have already quantified what can go wrong and have allowed for it.

Also check out: