Written By: Julian Bishop
Julian Bishop of Sarasin & Partners explains the dangers inherent in current methods of assessing the varying risk profiles of equities and proposes a different approach
There are many aphorisms in the world of finance, which are casually repeated without thought. The maxim that “higher risks means higher rewards” is cited as if a law of physics, despite scant evidence that it is true. Notions such as “equity risk premiums” imply that investors in equities demand a higher return than risk free assets, although precisely what premium is appropriate is hugely debatable. While some equities are hugely speculative and some businesses are inherently risky, some are not. Some businesses are consistently profitable even through the toughest economic times. Some have proven over decades, even centuries, that they have perennial relevance. How should we think about an equity’s riskiness in this context?
Understanding the role of financial modelling
Many core tenets of conventional financial theory are mathematical or rely on simplistic models. These are well- intentioned efforts at valuing and explaining the world around us, and many of them are genuinely useful. When you buy an equity, you buy the right to a portion of that company’s future cash flows. Therefore it is mathematically logical to say that we should value those cash flows at today’s value to calculate the value of that equity. But maths’ usefulness in equity analysis ends there. What maths cannot do is tell you what those cash flows will be. And therein lies the problem. All investment, by definition, involves predictions about the future, and the future is inherently unknowable. Mathematics, equations and models are of limited use in analysing financial instruments such as equities, where the range of possible outcomes for each company is enormous.
Economics and investment are, at heart, social sciences. There are no absolutes. Investors and economists can never enjoy the scientist’s luxury of laboratory conditions. We cannot undertake double-blind placebo experiments, where one country is subject to quantitative easing while another, identical, subject is not. Even retrospectively it is never completely clear why what has happened has happened. Everything economists and investors see and do, both in the past, present and future is viewed through a murky lens, with the number of variables near infinite. Predictions and forecasts thus have a huge margin for error.
The exceptions that prove there is no rule
In investment, there are always interesting stories in extremis. Take the investor who paid 19 times the “forward” price-to-earnings ratio (P/E) for a struggling technology company in 2003. The P/E he or she paid for the 2015 fiscal year turned out to be about 0.12 times; that is, the annual profits today are about ten times what their entire investment cost back then. Was that forecastable under any reasonable scenario? Probably not.
The alchemy of conditions that created that company (Apple, if you hadn’t guessed) was hugely random. If Steve Jobs hadn’t seen Jonathan (now Sir Jonathan) Ive poring over his prototype of a design for a new computer (the now iconic Apple G3), and hadn’t promoted him on the spot, would the iMac and iPod – a hugely risky proposition at the time – have come to market? In many businesses, as in meteorology, the “butterfly effect” truly matters.
Another famous case is that of a small American beverage company that went public back in 1919. The IPO came at a price of $45 and records show the company earned $3 per share of profit the following year, so it floated at a forward P/E multiple of 15 times – a reasonably standard multiple at the time for a decent quality company. Using historical records we can calculate the correct P/E (the one which should have been applied to the stock, had investors at that point known the true trajectory of the company’s earnings): an investor who paid at over 3,000 times the stock’s then P/E ratio would still have received returns on their original investment that compounded at 8% per annum through 2013, assuming dividends were reinvested. This was, of course, Coca-Cola.
But for every success story like this there are a thousand examples of companies that haven’t made it. Many equity investments fade to zero. Some make their investors rich beyond any original expectation. The variety of outcomes is staggeringly wide. How can we value any equity in that context?
Can we challenge how risk is valued?
The discounted cash flow (DCF) model is a mainstay of equity valuation. As financial models go, I like it; it stabs at the heart of what an equity truly is – a claim over all of a company’s future cash flows, discounted back to their value today. Sadly, there are two primary and hugely significant complications. First of all, as the Apple and Coca Cola examples show us, as do all the companies who have ultimately gone bankrupt, what future cash flows of most companies will be is hugely uncertain. To complicate matters further, the correct discount rate is also highly debateable.
The discount rate used in a DCF is, under conventional financial theory, a reflection of the “risk free rate” plus an additional “equity risk premium” (ERP), itself adjusted for beta – a measure of volatility versus the broader market. The risk free bit is relatively simple – use the appropriate sovereign bond yield. The concept of an equity risk free premium is intuitively neat. Equities are, under most scenarios, higher risk than sovereign bonds, so you rationally require a higher return. But how ERPs are calculated is open to broad church of interpretation.
Even some mainstream methods of calculating ERPs crumble under scrutiny. As Aswath Damadoram, a professor at Stern School of Business at NYU, said in a 1999 paper, it is often calculated using the “historical risk premium approach.” In other words, by calculating the historical gap between equity returns and bond returns over long periods, we can, supposedly, derive the premium equity investors require for the additional risk. In the US, this gap is measured at between 6-8% over the period from 1926 to 1997 depending upon the methods used. Thus the ERP is 6-8% – simple!
This makes the assumption that investors received precisely the same return they required for the risk they took, an assumption that stretches credibility. It ignores the huge survivor bias implicit in the reality that the US stock market has been the best performing over most long periods, reflecting America’s tremendous economic success and the ability of its businesses to create value. It also implies that less successful investments (such as those in many other countries’ stock markets, where long-term equity returns have sometimes been below risk free returns) require a lower equity risk premium. It equates past success with future risk. Applying the same theory to property, it would suggest that those who have been lucky enough to own property in London would require a higher future return on their investment than those who buy a property in Liverpool, for example, where property price increases have been far lower. It withstands no critical analysis whatsoever, yet this is mainstream financial theory.
Taking a fresh approach to valuing investments
There needs to be a better way of thinking about risk when analysing equities. My view is that the true risks to an equity’s cash flows vary hugely and judging those risks is a hugely underestimated part of the equity analysis process. ERPs should vary substantially from stock to stock based on a qualitative assessment of their company specific risk.
Our “Franchise Power” theme looks for where we believe the true risk to the company’s cash flows to be extremely low, and the risk of sizeable capital loss substantially minimised. These are companies which are proven winners in their space, where the end markets are reliable and where the chances of disruption are minimal. These are Warren Buffet’s “economic castles”, “surrounded by unbreachable moats”. At current valuations we find today’s free cash flow yields (that which is truly available to return to shareholders each year) on these equities are typically in the region of 4- 6% – considerably higher than the approximate 2% yields on the same company’s 10-year corporate bonds.
While there are some theoretical additional risks to equity cash flows versus bonds, in these examples we believe they are far outweighed by the superior yields and, importantly, the potential for long-term growth. These are relatively non-speculative investments in ungeared companies which are in control of their own destinies. They are worlds apart from the far riskier, more speculative end of the equity spectrum, yet often they are valued using exactly the same framework.
In short, the market does a poor job in assessing the differing risk profiles of equities. There is a tendency to see all equities as one asset class of homogenous risk; a tendency we work hard to avoid. If an equity’s cash flows are only mildly riskier than bonds, why shouldn’t they be valued as such, and a minimal equity risk premium be used when discounting them back to their value today? At current levels, we believe such equities should form a key component of the prudent investor’s balanced portfolio, even in an uncertain world of both challenges and change.