I recently looked at a SA article by a new contributor and, in passing, he commented (intelligently) on perceptions of risk and volatility by investors. Since Risk and Reward are the two most basic dimensions of concern to investors, I sought to find what else might have been written here on SA about the topic.
On the SA Home page there is an entry window box where a "symbol, author or keyword" will bring an appropriate response. So I entered "risk".
To my delight, the magic mirror on the wall told me: RISK is defunct.
Hurray! Good triumphs over evil!
A closer look at the response indicated that an earlier contributor organization used that word as a principal identifier, and sadly, they are no longer extant. (Don't ask why!) Further exploration indicated that, alas, there were no articles that dealt with the subject of risk explicitly. (Now ask - why not?)
Several references to various instablogs involved the term volatility, however. And there are a number of ETFs linked to volatility, and a VIX index that continually measures the condition in equities, at least as implied by options on the S&P500 index.
Since the two items, risk and volatility, are often used interchangeably but are significantly different, I propose to help readers explore the vital difference. In the process some profitable opportunities may occur to SA readers.
My definition (yours may well be different) is that risk is a human perception of the potential for the likelihood of events that will cause harm to the interests of the observer(s). That is, intentionally, as general as I can make it.
To begin with, risk (of any kind) has to do with getting hurt. For investors, that usually means losing money. Making money accidentally or by chance is not any part of risk.
Because risk is a perception, it only lives in the present and in the future.
If you decide to take a short-cut, on foot, across a busy highway, that decision as contemplated is a risky one, and is full of risk in its execution, until you are safely across. Then that risk is over, and no longer exists.
Looking back in time at such decisions may allow their evaluation as having been risky, but that recognition has value only in comparison with present or future decisions. A history of such comparable experiences might be useful in making decisions or choices in the future. The insurance industry follows that practice with an elaborate and well-established profession referred to as "actuarial science."
A key consideration in applying this approach is in identifying comparable circumstances. Actuarial sciences got its start in life insurance, where the absolute nature of death made circumstances of life expectancy very precise and comparable. Unfortunately, in the investment world there are few such certainties.
Equity investors look for "Return" from future price increases measured by earnings relationships - Price/Earnings ratios, or Price to Earnings Growth ratios. The "Risk" dimension typically is related to "uncertainty" about the price, rather than "uncertainty" about the earnings, or their growth.
Price uncertainty conventionally is measured by looking backwards in time to see its variances in units of time, typically expressed in statistical terms, like "standard deviation per day." This presumes chance is a constant across time.
Life insurance actuaries can be "dead right" (with mortality table updates) using such a presumption, but for investment analysts the parallel notion is dead wrong! Risk is not static, as many a failed-margin-call former investor can attest to.
Is it rational for a now $98 stock that has recently risen from $20, to be regarded only as risky as when it was at $50 on its way up, when at the time its last 52-week range was $98 - $20?
Furthermore, "normal statistics" utilize the "normal distribution" to infer probabilities of occurrence in things of interest, and have built well-accepted, elaborate and ordinarily useful terms and practices that rely on it. Including concepts like standard deviation.
Two problems are present in using this approach for investments. 1) The normal distribution is symmetric, containing both positive and negative departures from its mean, and 2) the historical distribution of stock price changes has been extensively shown to depart significantly - where it counts most - from the "normal". Stock price change distributions have "fat tails" (larger than "normal" extreme values).
Even if standard deviation was an appropriate measure of investment uncertainty, it cannot be calculated, and used as it often is, without including the positive outcomes to determine the mean value of the distribution. While those positive values are appropriate to the notion of uncertainty, they are not any part of risk, which includes only hurtful things.
Options investors look for "Return" from changes in price relationships to one another - "spreads" between contracts - in terms of uncertainty that are usually described as "volatility." The terminology comes about because by turning the usual options-price-determining formula around, and instead of having uncertainty as an input, accepting the existing prices as the inputs, the formula can be solved for uncertainty, or "implied volatility."
Historical norms of implied "vols" are usually expected to be returned to (the notion of mean reversion). In the process, changes happening in the option contracts' spread relations produce the desired net price changes.
Since options contracts are fundamentally insurance contracts - and thus risk-transfer arrangements - this thinking is well founded in its orientation.
So, ongoing indexes of implied vols are a useful gauge of the uncertainty content of stocks, if the options being used relate to stocks, or an index of stocks, like in the VIX. The VIX Index is a measure of the implied volatility of the S&P500 stock Index, logically derived from its options.
Stock investors have observed that when market prices generally get depressed for whatever macro reason, the VIX rises, often up into the moderate to mid double digits, from the low teens where it had shrunk while good times were rolling and prior concerns took a back seat and were becoming ignored.
But can the reverse be true? When the VIX is high is the stock market due for a rise? Or when the VIX is low, should stock investors fear an oncoming panic?
To test this out, we compared S&P500 Index prices in months following VIX prices, daily. To get the most meaningful relationships we looked for the biggest SPX (SPY) gains and biggest SPX losses in the multi-month periods following each of the daily VIX quotes.
Here are scatterplots of those relationships: First, looking for higher stock prices, and then, in the second picture, anticipating lower possible prices.
Since both of these efforts produced pretty poor relationships, we turned things around and made the same tests, but looked to see if the VIX Index might better be foretold by recent prior price changes in the S&P500. Here is what that produced:
These scatters make it pretty clear that the VIX is dependent on the S&P500's price changes, rather than the other way around.
There are over 100 Volatility market instruments, most of which the individual investor never sees. They are all derivations of the VIX index or detailed parts of it.
What is offered to the investing public is only half a dozen ETF products with any real market liquidity. Three of these are ETNs, backed by no other market support assurances other than the issuing institutions' creditworthiness. They are VelocityShares Long VIX ST ETN (VIIX) and iPath S&P 500 VIX ST Futures ETN (VXX), both playing off VIX futures with short-term expirations, and iPath S&P 500 VIX MT Futures ETN (VXZ), which has a mid-term focus to the expirations of the VIX futures on which it depends.
The volatility true ETFs are all straight short-term focus plays on the VIX futures. ProShares VIX Short-Term Futures ETF (VIXY) is the plain vanilla ETF, ProShares Ultra VIX Short-Term Futures ETF (UVXY) the (2x) leveraged ETF, and ProShares Short VIX Short-Term Futures ETF (SVXY) the inverse, or short position ETF.
The VIX itself, from which all these derive, being an Index, not an ETF, cannot be invested in directly. But the VIX does have options trading on it, used largely by market professionals. The options can be employed to create a "synthetic" instrument - buy an at the market call, sell a similar put, to be long the VIX, or the reverse for a bear position.
And since it is an optioned product we can derive market-maker price range expectations for it, as we do daily for over 2,000 other stocks and ETFs. Here is what the market pros are currently betting on for the VIX, and how their perceptions have been trending in the past 6 months.
(click to enlarge)
The options-based approach to uncertainty is superior to the historical, "normal statistics" approach to price volatility because it embeds forward-looking anticipations in its process. That way it reflects the dynamic, changing nature of risk and avoids the static presumption.
These price range forecasts provide an additional dimension to the VIX, since we can allocate the uncertainty that others are forced to work with into upside and downside prospects, separated by the current market quote. One part is the hurtful risk portion, and the other being the opportunity complement. Which is which depends upon the "long" or "short" posture the investor adopts.
Our standard analytical procedure is to examine how a subject's price has behaved in the past, following all varied levels of balance between those risk and reward prospects. There is no guarantee that future behavior will duplicate the past. But it probably helps in choosing between alternative investments to know which ones have had better odds of profitable prior experiences at present forecast proportions.
To simplify that process we create a single-number balance-metric we call the "Range Index." It is the current market price's position in a span where the low forecast = zero and the high forecast = 100.
The following table shows how Range Indexes for the VIX Index have been followed by S&P500 Index prices during the past 4+ years, including the 2008-9 market crisis. The numbers in the body of the table are annualized rates of SPX price change, to allow easy comparisons between the progressively longer periods. The 16 weekly-interval columns have a maximum of just under 4 months.
It seems evident that higher VIX Range Indexes are associated, in longer holding time periods, with higher rates of return and have odds of winning (not shown) better than 3 out of 4. But the number of these opportunities are limited. The table's second column tells that double-digit gain rates are constrained mostly to fewer than 100 instances out of 1100+.
These are worthwhile overall market opportunities, but we regularly see far more attractive potentials in specific stocks and ETFs on an every-day basis. The current VIX Range Index is 17, at the opposite (top) end of the Range Index scale where there are no big payoffs, excluding those very short holding periods.
The same kind of analysis on the Volatility ETFs does not produce anything competitive with either our minimum commitment standards, or alternative opportunities. Still, here is their current review:
(click to enlarge)