Market Volatility: Understanding the Risk

Given the recent swoon in the market, I thought this might be a good time to discuss volatility. Volatility is usually synonymous with risk. When you hear someone talking about risk, they are usually talking about volatility. But is it really the best definition of risk?

The market is generally assumed to return 10-12% per year on average, but each year can be +/- 32% around that average, or a range of -20% to +45%. That variation around the average is often called risk. And quite a range it is.

Volatility is typically defined as a statistical concept - it is the "standard deviation" (variance, etc.) of returns. In other words, one would calculate the average return on stocks from year to year, and then calculate how much actual returns deviate above and below from that average. Historically, for the market as a whole (the S&P 500, for instance), the standard deviation has been assumed to be around 16% with an average return of 10-12%. Most years will fall within two standard deviations of that average - thus the range of -20% to +45%. These numbers vary a little depending on who you talk to.

Market returns don't always fit within those parameters, and the standard deviation is not always 16% (such the boom and bust of the late '90s bubble), but it's a "useful fiction."

Of course, the volatility of the market is not the same as the volatility of individual stocks. The standard deviation on individual stocks is usually higher (25% for many large cap stocks). The reason the market has less volatility than individual stocks is because diversification reduces volatility. This is why investors are usually advised to "diversify risk".

It goes a step further. There are complicated formulas designed to link the returns on individual stocks, market returns, and volatility. One such formula, the "Capital Asset Pricing Model", is considered the "Bible" of business school finance courses. But does it work?

In order for volatility to work as a measure of risk, some assumptions must be made. One assumption is that stocks follow a "random walk", meaning that stock returns are random. If stocks are random, then returns don't follow any set pattern or rules, and cannot be predicted through analysis or linked to fundamental factors in any systematic way. Thus, if a stock moves up 10%, it is just as likely to move up another 10% as move back down 10%. Therefore, if this randomness is true, there are no reversions to the mean, no trends, and no valuation discrepancies.

Of note, this is a limited definition of the "Efficient Market Hypothesis" - a principle proposed by academics that basically means that it is not possible to predict stock price returns and that we should all just invest in index funds.

This idea is controversial, but there are many academic studies that are believed to back it up. Basically, it can be shown that the returns for investment professionals (such as mutual funds) follow a pattern that is consistent with the Efficient Market Hypothesis - that is, the pattern resembles a bell curve with an average that is close to (actually below when fees are factored in) the broad market indexes. That is, the average professional investor not only does not outperform the indexes, but charges extra for the privilege of performing in line with the indexes.

Unfortunately, none of these studies can actually "prove" the Efficient Market Hypothesis. It is a ""hypothesis"", not a theory nor a law. It cannot be proved. The idea that stock returns are random cannot be proved.

First, there is no such thing as a "random" event in the real world. That would violate the law of cause and effect. The concept of "chance" or "randomness" is a mathematical and statistical concept used for simplification and generalization. It is applied to events for which the explanation is either too complex or otherwise unobservable.

Second, "random" is not the same as "unpredictable". To summarize my first point, we use the concept of randomness to describe and predict that which we are otherwise unable to describe or predict.

Third, "chance" is really a proxy for "complexity." We live in a complex world, often too complex to predict.

Fourth, professional money managers are the market, for all practical purposes. In the aggregate, they determine the performance of stocks and the indexes.

Therefore the results of these studies are unsurprising: they effectively show that the indexes cannot outperform the indexes. The conclusions are defined by the assumptions, not necessarily by the data.

Let me take this a step further. If market returns are not "random", then maybe "volatility" is not an appropriate definition of risk. Perhaps "volatility" has been used instead as a proxy for "uncertainty". It is a very imperfect and limited proxy at that.

As investors, we are investing in an uncertain world. There are many things we cannot predict, but there are some things we can. While difficult to measure statistically, I believe that there are inherent tendencies in the market and in return patterns.

Specifically, the market as a whole follows a boom and bust pattern that is "manic-depressive". For instance, recessions have been around as long as civilization itself. Yet, as soon as there is a whiff of a recession, economically sensitive stocks take a nose-dive. Then, when the recession is over, these stocks rocket back up. Is this rational?

Individual stocks fall in and out of favor even when the underlying business dynamics haven't changed. If we know this, then a good way for profit is to buy stocks that are out of favor and sell stocks that are in favor.

Both of these strategies take a long-term perspective. They don't work if your time horizon is tomorrow or next week.

I do not believe volatility is a good measure of risk. It measures phenomena that I do not consider to be risk - like the silliness of the market - and it fails to capture very real underlying risks. Thus, investors relying on volatility measurements, like beta or standard deviation, are, in my opinion, not adequately protecting themselves from risk but may be protecting themselves from profit.

Thus, answering my question, I am not concerned about most kinds of volatility for my portfolios. If am right on any individual stock, I will be paid regardless of whether the stock goes down or up after I buy it. I don't really care about the path it follows between the purchase and the sale.

What about diversification? I do diversify. But I am not really interested in diversifying volatility.

To paraphrase Warren Buffett, I would rather have a high return that is lumpy ("volatile") than a low return that is smooth.

However, I will diversify my definition of risk - though I generally prefer to avoid this kind of risk rather than diversify it. I know that I am sometimes wrong or unable to perfectly predict the future, so I try to structure my portfolio to take advantage of return opportunities while limiting my exposure to my definitions of risk.

Thus, I will sometimes avoid stocks that may have great upside, but that for some reason I can't get comfort with the unknowns. I will buy stocks that appear to be highly volatile, but I have a high level of conviction in their ultimate value.