Most investment fund managers attempt to quantify risk through a variety of numerical measures. When discussing risk in their portfolios, they will use terms like �standard deviation,� �Beta,��Sharpe Ratio,� and many others to give the listener a sense of how much risk the overall portfolio has. While simple and convenient, these measures are incomplete at best, and many times inaccurate.They lull their users into a false sense of security. Is measuring risk in terms of volatility (as these gauges do) the best way to assess risk?
I favor the use of a businessman�s approach which blocks out the market �noise� of short-term price movements. I base my risk management around one central theme: buying an outstanding business at far less than its intrinsic worth. If done often enough throughout one�s investing career, this discipline in risk management has proven to produce far superior results than use of any simple volatility ratios has. The most pointed example that most readers would be familiar with is the track record of Warren Buffett. Throughout his storied career which has produced market-beating results for forty-seven years, Mr. Buffett has never used volatility measures and regards them as useless (see the Appendix to �The Intelligent Investor,� by Ben Graham for more on Buffett�s views on this topic). The key to his success is to stay away from doing stupid things and to buy great companies at forty to fifty cents on the dollar or less.
The theory behind using single numbers to quantify the risk of a basket of securities is taught everywhere in the financial community, from top business schools to the CFA program and Wall Street itself. The drive to nail down risk into a single number or set of numbers stems from the natural human instinct to have �control� over the uncontrollable. It is unsettling to think that one may not have any control over a number of factors that affect the individual investments that are in a portfolio. Placing a number on a portfolio gives the manager more assurance that he has some control over his investments, and their outcome is relatively assured. As history has taught us in extreme cases like Long-Term Capital Management in 1998, relying on single number gauges of risk can lead to disastrous results.
I will acknowledge that there are many users of these ratios that produce fabulous, consistent results exactly as expected. I would offer the proposition that in a majority of these cases however, that the manager utilizes a superior investment philosophy which gives him an �edge� over other market participants. I would venture to say that this �edge� is what provides the consistency and performance, rather than a superior skill in utilizing risk ratios. After all, with over 90% of the investment community measuring risk in terms of volatility, how much of an edge can a manager have over another manager utilizing such gauges? Inherently, the more people that use a particular philosophy for investment, the less effective that philosophy becomes. This goes for risk management measures as well.
The inherent flaw with all volatility measures of risk is that they are backward-looking. They measure what has happened in the past, and presume that the future will hold similar results in terms of how volatile an investment or a set of investments will be. Users of these measures were very impressed with Long-Term Capital�s results going into 1998. The firm had posted consistently extraordinary results since its inception, with the largest monthly loss of only 2.9% before 1998. By any measure of volatility, the fund was the �toast� of Wall Street, and demand to invest in the fund couldn�t be higher. What the firm and ultimately its investors failed to possess was the common sense to realize that when you carry leverage ratios of 30-1 and higher, you may be carrying a little more risk than some theoretical number can measure. After all, �markets can remain irrational longer than you can remain solvent.� Further, when you put on �hedged� positions in numerous countries around the world which essentially have the same characteristics (for example in LTCM�s case, buying lower-grade bonds and selling government paper betting that the spreads will tighten) no single number can measure your risk when spreads move against you in all the countries at the same time.
The comfort that comes from single-number measures of risk gives users of these equations a false sense of security that �everything is covered.� Portfolio management progresses along fine for most of the managers that utilize these techniques, until something extraordinary happens. In other words, the ratios work until they don�t. Often there is little forewarning, and even less time to react to rapidly changing market conditions. The portfolio insurance debacle during the Crash of 1987, illustrates the speed with which things can change, and how little control all market participants have in the process. The user of volatility ratios would �freeze like a deer in headlights� under those conditions. Many would sell or flatten out their positions to manage their �risk,� or volatility. However, the rational, disciplined businessman would most likely engage in buying under such conditions. After all, in his view, he can pick up the same assets as the day before, but like in the Crash of �87, he could pick them up at 25% off. Those that had the guts to buy after the Crash were handsomely rewarded with hefty returns over the coming decade. Those that grew overly-concerned with the volatility and got out of the market altogether, missed the fire-sale prices. Most likely, they didn�t reap as large of returns over the coming years.
Most ratios fail to take into account that �everything important in financial history has taken place outside of two standard deviations.� Long-Term Capital again exemplifies the lack of control that all investment managers have when using these ratios. The management at the firm in fact believed that it would take a �ten-sigma� event for them to lose money:
According to these same models, the odds against the firm�s suffering a sustained run of bad luck- say, losing 40% of its capital in a single month- were unthinkably high (so far in the worst month, they had lost a mere 2.9%). Indeed, the figures implied that it would take a so-called ten-sigma event- that is, a statistical freak occurring one in every ten to the 24th power times- for the firm to lose all of its capital in one year.<<p>
As we came to find out in the fall of 1998, these �models� created by Nobel Prize winners and Wall Street geniuses were a little off. The firm collapsed in the face of excessive leverage and illiquidity in getting out of their positions. No magic set of numbers accurately assessed their risk, and their investors paid a dear price.
A more precise critique of volatility measures of risk can be made when discussing �Beta.� Seeking to limit �risk,� a manager which uses Beta to measure the overall portfolio, will often seek to limit the Beta of the portfolio, and thus provide consistency of return which is uncorrelated to the overall stock market. In adjusting the portfolio, he will often discard stocks that have displayed heightened volatility, so as to keep the overall portfolio in line with the fund�s Beta objectives. If a number of stocks in the portfolio have displayed choppy market price action, wholesale changes will be made to bring in more consistent stocks which don�t have such volatile price action. Does measuring past price history of a stock and its volatility adequately measure the risk of the stock?
To put my critique more bluntly, I will ask a simple question: �if a stock�s price drops from $40 down to $10 in a short period, and its Beta increases in the process, is the stock more risky at $10 than at $40?� To cite a concrete example that has occurred in the market over the past year, I will point out BJ�s Wholesale Club. The stock traded above $40 in early 2002, and subsequently dropped to a little under $10 in March of 2003. By any volatility measure, this 75% drop in the stock price was much more volatile than the overall market�s drop of a little over 20% during this period. As a backward-looking measure, the Beta of the stock increased significantly during this period. Classic Beta philosophy would suggest the stock is more volatile than the market and thus is more �risky.�
I am quite confident to say that there were numerous fund managers that noticed this increase in Beta and subsequently dumped the stock on the way down. For those that got out early in the move, they were spared a great deal of pain. However, for those that waited until the stock�s Beta reached an extremely high level (and thus increased the �risk� of their portfolio too much), they most likely got out when the volatility of the stock price reached its zenith. This most likely occurred in early March when the stock priced dropped 25% in one day down to $10.34. The stock price subsequently hovered around $10 for the next couple of weeks before advancing. Was the stock more risky at $10 than at $40?
It is easy to say in hindsight that those that got out around $10 because they couldn�t take the pain of the volatility of the stock price were wrong in their assessment of risk. The stock price ultimately advanced in the short-term, and they could have held out and sold at higher prices. In playing �Monday morning quarterback� however, the vital point would be missed. Even if the stock had subsequently dropped further, it was still less risky at $10 than at $40 no matter what the volatility measures said. This begs the question: �are volatility measures of risk the best way to assess the risk of a particular security or set of securities?�
The BJ�s Wholesale Club example points out the inherent flaws of gauging risk by measuring volatility. Oftentimes, volatility spikes when a particular security�s price is near a bottom. The panic overwhelms any rationality, and price moves tend to be the most exaggerated when the price is ready to take a turn. To cite other examples of this phenomenon from the stock market in 2003, observe the price action of stocks like Nautilus, Winnebago, and AFC Enterprises. In each case, there was a substantial adverse move in price at or near the bottom of a prolonged downturn. If you had been a staunch believer in Beta or other volatility measures, you would have passed on a great opportunity to buy these securities at good prices, or worse, you would have sold them.
Coinciding with this aversion that fund managers have with highly-volatile stocks and portfolios is the fact that volatility measures will lead managers to exclude the very securities that offer the highest return potential. Most market observers can agree that the securities with the choppiest price patterns tend to be smaller-cap issues that have fewer analysts covering them. Moves in the small-cap arena tend to get exaggerated due to lack of liquidity among market participants. These same observers will also agree that these small-cap issues are the securities that offer the highest return potential in the long-run if carefully selected. As great as the returns of the large-cap stocks has been over the past century, small-cap stocks have performed even better.
High-quality small-cap issues offer more growth potential for two main reasons. First, because they are less-followed by market participants, there is a greater likelihood that they are mispriced relative to their intrinsic value. With fewer �eyes� on them, smaller issues can offer substantial gains for buyers who realize stock price appreciation through a higher �valuation� on the shares. The second main reason that small-cap stocks offer more upside centers around the businesses themselves. Smaller stocks have more potential for growth than typical large-cap stocks. It is much easier for a company that has $500 million in sales to double than for a company that does $20 billion in sales. The small-cap stars of today are tomorrow�s S&P 500 companies and the stocks that will gain a wider following amongst Wall Street houses.
Many substantial gains can be reaped by carefully selecting these less-well-known issues and allowing them to perform for you. However, if measuring risk in terms of volatility, managers will eliminate the purchase of these future stars and will often �invest with a rearview mirror,� buying the stocks that have provided consistent returns in the past. Blinded by volatility gauges, these managers will pay premium prices on larger issues that have displayed more �smooth� price patterns. They forget the basic notion that good past returns borrow from the returns of the future. These same issues that have performed so favorably in the past relative to their benchmarks will tend to be the laggards going forward. To illustrate, I will cite the example of Wal-Mart. An incredible company, Wal-Mart has now become a cornerstone in most money-managers� portfolios because of its consistently high returns. However, will those returns continue? For the stock to perform as well as it has in the past, the company will need to be doing over $1 trillion in sales in 2013 and be trading at the same premium multiple that it trades at today. Is this feasible? I don�t like the odds!
The fundamental criticism I have with most volatility measures of risk is that they tend to be backward-looking. Managers get lured into a sense of comfort by getting their portfolio�s Beta or correlation numbers to their targeted level. They give the manager an illusory sense of �control� over the situation. By inherently leading managers to �invest with a rearview mirror,� these correlation numbers created substantial losses if used strictly in cases like the Nifty-Fifty and in the recent tech and telecom bubble. By strictly screening for stocks that offered the highest return per unit of risk as defined in the recent past, these ratios led their users to invest in overvalued securities right at their peak. The ensuing losses were substantial.
By relying on volatility ratios too much, managers put too little weight into more common-sensical gauges of risk like leverage ratios, quality of the business, consistency and growth of the earnings, and price relative to underlying value. The solution to risk management is that there are no hard and fast rules which can accurately define the risk in a given portfolio or an individual security. Anybody that uses single volatility ratios to define risk most likely will be in for a rude awakening at some point during his career. As Warren Buffett said at this year�s annual meeting �if anyone uses precise figures in finance, be careful.�
For the fund I manage, Tarrach Holdings, LLC, I could tell you the Sharpe Ratio and the Standard Deviation, as it is calculated for me on a website that I use (Hedgeco.net). However, these numbers have little meaning to me, as I understand the flaws of relying on them. My style of risk management is more concerned with measures of risk which attempt to quantify stock price relative to intrinsic value. With limited diversification, I attempt to purchase a basket of stocks which are extremely undervalued relative to a range of values that I have placed on them. I attempt to buy dollar bills for forty or fifty cents. Additionally, I place a lot of thought into the dynamics of the business. I look for companies with huge competitive advantages, high returns on capital with little debt, rapid growth rates in �owner earnings,� and preferably, have high profit margins. Having a keen sense of both qualitative and quantitative measures of a business comprises my �first line of defense� in the risk management process. I live by the John Keynes method of investing: �one bet soundly considered is preferable to many poorly understood.�
As I assess risk in the portfolio, I don�t use any hard and fast rules for when to get out of a stock. I don�t use �stop-outs� to get out if the stock price falls. In general, I would use a falling stock price as an opportunity to add more shares of a good business at an even better price. I focus my attention on the business itself and its price relative to value. If the business is becoming more �commodity-like� or the share price advances to a point where I can sell my stake for 100% of value or more, I may engage in selling an issue. I don�t pay attention to the short-term �noise� that the market presents at every moment. With the quality of the businesses in my portfolio, I would not be disappointed if the stock market closed for several years- the daily ripples in the stock price is not my focus except to the extent that it may present opportunities. I focus on the long-term prospects for the business itself. If I buy a sound, growing business at a reasonable price, I am confident that the stock price will rise to reflect that growth. I hold the firm belief that the market is relatively efficient in the long-term, but in the short-term can display periods of extreme inefficiency. The past five years have been affirmative evidence of that short-term inefficiency.
As my last article, �A Critique of the �Top-Down� Approach to Investing,� pointed out, I don�t purport to have all the answers to investing. In fact, there is a great deal I do not know, nor do I believe anyone else knows. The same holds true with risk management. The attempt to boil down risk in a portfolio of hundreds of securities and derivatives into a single number does not make sense to me. Risk control is more than crunching numbers and saying that a particular �ratio� looks all right. It is about staying away from doing something stupid. In my book, the only way that can be achieved is through being disciplined about buying great companies at great prices.
While I cannot provide personalized investment advice or recommendations, I welcome feedback and observations by subscribers. You can email me at ctarrach@hotmail.com
Christopher Tarrach manages Tarrach Holdings. To learn more about Chris and his fund, CLICK HERE.
Disclaimer: Chris Tarrach is the president of Tarrach Holings, Mr. Tarrach’s columns are not intended to constitute investment advice or a recommendation to buy, sell, or hold any security.
