Victor Yan

Introduction

The alluring prospect of engaging in investments is to capture returns on your initial capital, be that in the form of dividend distributions or capital gains. However, as famously termed by Milton Friedman, “there is no free lunch”. This understanding of tradeoffs has extended its way into financial literature with the idea that given a well-diversified portfolio, in order to achieve higher return, one must be willing to bear greater risk in the process; conversely, one who incurs higher risk should be compensated fairly for doing so. 

This is perhaps lesson number one in finance across university courses. For example, the following is a screenshot taken from the first lecture of Investments (FNCE30001):

When asking a lecturer on how we would proxy risk, textbooks and slides typically do so via an asset’s historical volatility (often denoted by sigma). This, coupled with an understanding that people on average are of the risk-averse type compel any investor to attempt to “maximise reward while mitigating risk”. Subsequent lectures on portfolio management will prescribe the Markowitz method of allocation asset weights to yield a risky portfolio with a maximum achievable Sharpe ratio to achieve this goal.

However, as often is the case, assumption-based academic theory is not a legitimate substitute for practical application in the real world. This article attempts to communicate the shortcomings for those who subscribe to this naive interpretation of risk, and offer an alternative perspective on the matter. 

Risk is a Two Sided Coin

The first observation to gather from an academic framing of risk is that high volatility is bad and low volatility is good. This connotation purely attributes the volatility calculus to equal the greater prospect of losing money.

In part, this is true, but only one side of the coin. One has to question why the higher prospect of gaining money through elevated volatility is never considered. To aptly illustrate this point, consider the following two stock price evolutions over time:

Which of these stocks would an investor seeking to minimize volatility pursue? Based on an eyeball test, Stock A’s peaks and troughs are less exaggerated than Stock B – thus it has lower volatility and should be favoured on a risk basis under academic interpretation. However, we can appreciate a situation where the higher volatility of Stock B has led to relative outperformance on returns that looking back upon, was regrettable to disqualify on the basis of higher volatility.

Of course, this is not to recommend investment in higher volatility stocks such as pre-feasibility mining explorers, but demonstrates a countercase to where the upside of volatility can profit the investor. 

As suggested by Benjamin Graham in “The Intelligent Investor”, the first objective of the investor is to ensure capital preservation, or in layman’s terms, to reduce your chances of losing money. Thus, perhaps a better way to proxy risk would be to tailor volatility to fit a one-sided profile on the downside. This is otherwise known as the “Sortino Ratio”, where we take only historical observations that fall below a minimal acceptable return, and compute standard deviation accordingly.

This aptly corrects standard deviation to account only for the left-sided tail of the risk distribution, while laying free the right-sided tail where capital gains can appreciate. After all, I rarely see people complain about windfall gains of 30% instead of 5% because it was mathematically more volatile. However, as we will explore later in the article, reliance on historical data can come with its designated limitations.

What does Behavioral Economics Say?

Having established a “downside-risk” adjusted measure of standard deviation, this is also arguably a better way to understand and compute risk when we consider the teachings of prominent behavioural economists Kahneman and Tversky, who famously won the 2002 Nobel Memorial Prize in Economics for their seminal 1979 paper: “Prospect Theory: An analysis of decision under risk”.

The nuts and bolts of Prospect Theory follow a descriptive analysis into how humans actually behave under games of wins and losses. Based on controlled studies, individuals are shown to display an asymmetrical perception towards winning and losing in what we call “loss aversion”, where losses are felt more strongly than gains. In many cases, the dominant aversion against losing is so strong that people choose against taking a bet even when the odds are mathematically stacked in their favour. The Youtube channel Veritasium conducted an experiment on a coin toss game for asymmetric payoffs that illustrates the point:

What does this imply for the conventional interpretation of risk? Given an altered form of volatility that concentrates purely on downside risk, the idea of loss aversion lends greater credence to the fact that we should focus greater attention on losses than gains when we assess risk. The prescription here is to take risk aversion one step further to loss aversion. 

A subscriber to academic theory would pundit that Prospect Theory is irrelevant towards generating the optimal portfolio of risky securities to maximise return and minimise risk. After all, Prospect Theory deals in subjective utilities and not hard risk and return metrics. In fact, the “separation property” taught in textbooks posits that a fund manager would offer each one of his clients the same risky portfolio optimised by maximising the Sharpe ratio irrespective of risk aversion, before adjusting to each client’s risk preferences by the longing or shorting of risk free assets. 

But none of this happens in real life. The responsibilities of fund managers rarely deal in the realm of academic psychology; catering to the utilities of each individual client is nearly impossible. Testing for utilities in the real world is often an arduous task, with a prohibitively high error rate. Instead, funds are structured with mandates that roughly categorise a specific type of investor, The simplest example would be the likes of superannuation funds where an investor subscribes to a plan ranging from conservative to balanced to aggressive. 

Hence, one must question when this supposed “adjustment to each client’s risk preferences” actually occurs in the portfolio construction process? Instead, funds create their own demand by enlisting the product first before advertising for prospective clients who shop for a fund with the appropriate risk and return characteristics in accordance with their appetites. These risk preferences are captured instead, by revealed preferences from choice of fund by clients – which also explains the legal requirements for funds to distribute items such as Product Disclosure Statements in order to eliminate information asymmetry risks. In other words, it is the onus of the clients to know about the fund’s characteristics, not the other way around as suggested in academic classrooms. One can imagine how much whiter Peter Lynch’s hair might get if he had to accommodate his fund to each client from a pool of thousands on top of monitoring the thousands of positions he held.

If managers do not in fact tailor for each client’s risk preferences post-optimal risky portfolio construction, then there is an argument that they should do so when constructing the post-optimal risky portfolio. Hence, lending in from our measure of risk that factors only downside deviation, some factor greater than 1 may be applied to adjust for the loss aversion we descriptively see in human behavior. This is particularly important for funds with unstable shareholder bases that also feature thin defenses in short redemption clauses and lock in periods, as a down swing in the portfolio exacerbated by loss aversion and short-termism by shareholders may result in the high likelihood of angry, emotional customers barking down the manager’s telephone. 

The Past of the Future

Another qualm regarding the academic interpretation of risk by volatility presides in its utilisation of historical volatility. This introduces the implicit assumption that the past is a good proxy for the future. Overall, this reliance is fairly dangerous if the business fundamentals have been subject to change over time. In fact, funds are legally required to disclose in their documents that “past performance is not indicative of future performance”.

Take a company like Tencent for example. Once starting out as an online messaging service with the likes of QQ to the superapp WeChat, the business is not comparable to what it is now, where the grand cash cow of the business stems from its ownership and equity stakes in the gaming universe. While Tencent may have been a core service provider in communications 10 years ago, the same company is more aptly described as an investment firm behemoth comparable to the likes of Berkshire Hathaway where Buffett and Munger utilise their insurance business float to fund prospective investment areas (in this case, Tencent utilises their recurrent cash flows from their gaming operations). Simply put, the economics of the business, and hence its volatility profile, is non-comparable to its past. After all, companies are not stale entities but living organisms subject to change over time (and sometimes these changes can be of a discrete nature in the case of situations such as M&A activity). 

There have been certain “blow-up” episodes chronicalled throughout investment history for entities that rely on the past to predict the future. Perhaps most notably, this was detailed in the downfall of Long-Term Capital Management (LTCM) in 1998 – founded by Salomon Brothers bond trader John Meriwether and led by the likes of Nobel Prize laureates Myron Scholes and Robert Merton of the Black-Scholes-Merton model fame. Dealing in bond arbitrage on heavy leverage (this article won’t go in depth on their strategy but if you want more detail on this, read When Genius Failed by Robert Lowenstein: https://en.wikipedia.org/wiki/When_Genius_Failed), the fund evaluated its assets on a VaR (Value-at-Risk) assessment based on the idea that historical volatility would persist into the future. This resulted in a fantastical meltdown in August 1998 when Russia devalued the ruble and declared a moratorium on its Treasury debt, leading to a flight to quality that led to widening of spreads when quality liquid investments and less liquid, low creditworthy investments – the opposite to LTCM’s position where they were betting on these spreads tightening. This led the hedge fund to suffer massive losses of nearly $4 billion, which required an orchestrated bailout by the United States government to prevent the losses from spilling into the markets (considering LTCM was holding roughly 5% of the fixed income market globally at the time). 

As an overall note, as detailed in Nicholas Taleb’s works, people fail to accurately account for the occurrence of negative black swan events – outcomes multiple deviations away from their distributional averages. As observed by Anthony Bolton, “investors underestimate the likelihood of rare events happening when they haven’t happened recently, while they overestimate them when they have”.

Logic Over Numbers

While investing is often heralded as an activity for the mathematically inclined, this often predisposes businesses to be thought as a matrix of numerical inputs and outputs rather than businesses itself. Over time, subconscious habits manifest themselves where the analyst places greater emphasis on the numbers than common sense logic. This is not only naive for the reasons stated in this piece, but also arguably boring as the analyst is relegated to a role as a glorified number plugger. It is much more interesting instead, to think of risk not as a silver bullet number of volatility, but as the dimensions of business risk subject to the operational characteristics and environment of the company, and financial risk subject to the leverage of the company through deliberation of preferred shares or debt instruments. 

This not only provides a logical mental model into disseminating the sources of risk for a company, but provides useful information for the marginal investor to understand the business at hand. For example, through understanding operational risk, one can grasp a view on the  cyclicality of the business, degree of operating leverage, and price elasticity of their product. Investors can also analyze financial risk to evaluate the effectiveness of risk management from discretionary management decisions. While numbers are important, a diagnostic analysis of risk provides their context. 

Conflating Price and Value

Finally, an interpretation of volatility as risk can be misleading as this volatility is solely based on the pricing fluctuations of an asset. As those from the camp of value investing can attest to, price and value are not one and the same. Aswath Damodaran provides a good rundown on this concept here:

Simply put, just because one observes the fluctuations of an asset’s price, does not necessarily reflect the case where the price of the asset has necessarily changed. A well-documented case pertains towards spinoff situations, where on average we observe indiscriminate selling on these businesses without any negative change in its fundamentals. If you’re interested in understanding spinoffs in greater detail, refer to (https://www.unitunimelb.org/2020/08/16/the-spin-on-spinoffs/).

Another common example where price can deviate away from value occurs in stock splits, where existing shares are divided into smaller pieces. For example, one share that was worth $400 could be split into 2 shares worth $200. This is to primarily motivate a boost in liquidity for the stock to provide better access to capital constrained individuals who may not have the $400 to fork over, but also to play on the psychology framing where the average person is more comfortable with buying 2 shares of $200 than 1 share of $400. Multiple companies have employed stock splits, more recently Apple back in 2020 on a 4-to-1 basis, translating its $540 share price to $540/4 = $135. Tesla also underwent its own stock split last year on a 5-to-1 basis from a sky-high share price of $2230 down to $2230/5 = $446. (Funnily enough I managed to prank a friend convincing him the company’s share price tanked 80% on the day). 

But question if the fundamental value of the company has changed or not? Before and after the split, Apple was still making its key products, had the same debt risk, was based in the same locations, and engaged in the same projects. What is effectively happening with a stock split is that a pizza that was once divided into 4 pieces is now divided into 8 pieces instead. But at the end of the day, a pizza is still a pizza. This is a common example of how flows, not value, can affect asset pricing. Again, Aswath Damodaran explains this in greater detail here: http://aswathdamodaran.blogspot.com/2020/09/illusion-perception-and-reality-stock.html.

Thus, using volatility as a measure of risk can lead to adverse decisions against finding the best value stocks – those that are priced at depressed levels from their intrinsic worth. For the value investor, volatility is not something to be fearful of, but something to relish as it opens the doors to find attractive opportunities. 

This is most lucidly expressed by Michael Burry in his April 2001 shareholder letter, which I will paste an excerpt below:

Conclusion

While it is simple, and even intuitive at times to understand volatility as risk, doing so can lead investors to take a narrow understanding of what is a complex problem. Risk represents the yin to reward’s yang, and requires a nuanced interpretation beyond a simple sigma figure that frankly challenges even the most seasoned of investors. Ultimately, university material paints the landscape for us to understand fundamental objects and relationships in finance. This is an important first foundation in learning, but it should not serve as a placebo for understanding. It is ultimately the student, investor, or reader’s job to continually push the envelope, and question the status quo (the contents of this article included). After all, thinking is much more interesting than memorizing.

Victor Yan is the Vice-President of Publications for UNIT – University of Melbourne, specializing in financial investments and economic policy.

References

Anthony Bolton: Investing Against the Tide

Benjamin Graham: The Intelligent Investor

Bryan Lim: University of Melbourne Investments Lecture 1 Slides

Daniel Kahneman: Thinking Fast and Slow

Howard Marks: The Most Important Thing

Michael Burry: April 2001 Shareholder Letter

Robert Lowenstein: When Genius Failed

Mark Meldrum: https://www.youtube.com/watch?v=N614C4gKVdI&list=PLM9WI-4yn8BKekRABsV7OgBCwgFi_ELrl&index=28

Vineyard Holdings: Monopolies with Scale (Tencent)

Aswath Damodaran: http://aswathdamodaran.blogspot.com/2020/09/illusion-perception-and-reality-stock.html

Veritasium: https://www.youtube.com/watch?v=vBX-KulgJ1o&t=329s

The Plain Bagel:

Disclaimer: The views expressed in this article are solely that of the author’s, and do not necessarily reflect the position of UNIT nor the University of Melbourne. The advice given is general in nature and does not consider an individual’s personal financial circumstance. Transacting off this information is done so at one’s own risk, and individuals are encouraged to consult a finance professional before making investment decisions based off of this article.