Volatility is considered the most accurate measure of risk and, by extension, of return, its flip side. The higher the volatility, the higher the risk – and the reward. That volatility increases in the transition from bull to bear markets seems to support this pet theory. But how to account for surging volatility in plummeting bourses? At the depths of the bear phase, volatility and risk increase while returns evaporate – even taking short-selling into account.
“The Economist” has recently proposed yet another dimension of risk:
“The Chicago Board Options Exchange’s VIX index, a measure of traders’ expectations of share price gyrations, in July reached levels not seen since the 1987 crash, and shot up again (two weeks ago)… Over the past five years, volatility spikes have become ever more frequent, from the Asian crisis in 1997 right up to the World Trade Centre attacks. Moreover, it is not just price gyrations that have increased, but the volatility of volatility itself. The markets, it seems, now have an added dimension of risk.”
Call-writing has soared as punters, fund managers, and institutional investors try to eke an extra return out of the wild ride and to protect their dwindling equity portfolios. Naked strategies – selling options contracts or buying them in the absence of an investment portfolio of underlying assets – translate into the trading of volatility itself and, hence, of risk. Short-selling and spread-betting funds join single stock futures in profiting from the downside.
Market – also known as beta or systematic – risk and volatility reflect underlying problems with the economy as a whole and with corporate governance: lack of transparency, bad loans, default rates, uncertainty, illiquidity, external shocks, and other negative externalities. The behavior of a specific security reveals additional, idiosyncratic, risks, known as alpha.
Quantifying volatility has yielded an equal number of Nobel prizes and controversies. The vacillation of security prices is often measured by a coefficient of variation within the Black-Scholes formula published in 1973. Volatility is implicitly defined as the standard deviation of the yield of an asset. The value of an option increases with volatility. The higher the volatility the greater the option’s chance during its life to be “in the money” – convertible to the underlying asset at a handsome profit.
Without delving too deeply into the model, this mathematical expression works well during trends and fails miserably when the markets change sign. There is disagreement among scholars and traders whether one should better use historical data or current market prices – which include expectations – to estimate volatility and to price options correctly.
From “The Econometrics of Financial Markets” by John Campbell, Andrew Lo, and Craig MacKinlay, Princeton University Press, 1997:
“Consider the argument that implied volatilities are better forecasts of future volatility because changing market conditions cause volatilities (to) vary through time stochastically, and historical volatilities cannot adjust to changing market conditions as rapidly. The folly of this argument lies in the fact that stochastic volatility contradicts the assumption required by the B-S model – if volatilities do change stochastically through time, the Black-Scholes formula is no longer the correct pricing formula and an implied volatility derived from the Black-Scholes formula provides no new information.”
Black-Scholes is thought deficient on other issues as well. The implied volatilities of different options on the same stock tend to vary, defying the formula’s postulate that a single stock can be associated with only one value of implied volatility. The model assumes a certain – geometric Brownian – distribution of stock prices that has been shown to not apply to US markets, among others.
Studies have exposed serious departures from the price process fundamental to Black-Scholes: skewness, excess kurtosis (i.e., concentration of prices around the mean), serial correlation, and time varying volatilities. Black-Scholes tackles stochastic volatility poorly. The formula also unrealistically assumes that the market dickers continuously, ignoring transaction costs and institutional constraints. No wonder that traders use Black-Scholes as a heuristic rather than a price-setting formula.
Volatility also decreases in administered markets and over different spans of time. As opposed to the received wisdom of the random walk model, most investment vehicles sport different volatilities over different time horizons. Volatility is especially high when both supply and demand are inelastic and liable to large, random shocks. This is why the prices of industrial goods are less volatile than the prices of shares, or commodities.
But why are stocks and exchange rates volatile to start with? Why don’t they follow a smooth evolutionary path in line, say, with inflation, or interest rates, or productivity, or net earnings?
To start with, because economic fundamentals fluctuate – sometimes as wildly as shares. The Fed has cut interest rates 11 times in the past 12 months down to 1.75 percent – the lowest level in 40 years. Inflation gyrated from double digits to a single digit in the space of two decades. This uncertainty is, inevitably, incorporated in the price signal.
Moreover, because of time lags in the dissemination of data and its assimilation in the prevailing operational model of the economy – prices tend to overshoot both ways. The economist Rudiger Dornbusch, who died last month, studied in his seminal paper, “Expectations and Exchange Rate Dynamics”, published in 1975, the apparently irrational ebb and flow of floating currencies.
His conclusion was that markets overshoot in response to surprising changes in economic variables. A sudden increase in the money supply, for instance, axes interest rates and causes the currency to depreciate. The rational outcome should have been a panic sale of obligations denominated in the collapsing currency. But the devaluation is so excessive that people reasonably expect a rebound – i.e., an appreciation of the currency – and purchase bonds rather than dispose of them.
Yet, even Dornbusch ignored the fact that some price twirls have nothing to do with economic policies or realities, or with the emergence of new information – and a lot to do with mass psychology. How else can we account for the crash of October 1987? This goes to the heart of the undecided debate between technical and fundamental analysts.
As Robert Shiller has demonstrated in his tomes “Market Volatility” and “Irrational Exuberance”, the volatility of stock prices exceeds the predictions yielded by any efficient market hypothesis, or by discounted streams of future dividends, or earnings. Yet, this finding is hotly disputed.
Some scholarly studies of researchers such as Stephen LeRoy and Richard Porter offer support – other, no less weighty, scholarship by the likes of Eugene Fama, Kenneth French, James Poterba, Allan Kleidon, and William Schwert negate it – mainly by attacking Shiller’s underlying assumptions and simplifications. Everyone – opponents and proponents alike – admit that stock returns do change with time, though for different reasons.
Volatility is a form of market inefficiency. It is a reaction to incomplete information (i.e., uncertainty). Excessive volatility is irrational. The confluence of mass greed, mass fears, and mass disagreement as to the preferred mode of reaction to public and private information – yields price fluctuations.
Changes in volatility – as manifested in options and futures premiums – are good predictors of shifts in sentiment and the inception of new trends. Some traders are contrarians. When the VIX or the NASDAQ Volatility indices are high – signifying an oversold market – they buy and when the indices are low, they sell.
Chaikin’s Volatility Indicator, a popular timing tool, seems to couple market tops with increased indecisiveness and nervousness, i.e., with enhanced volatility. Market bottoms – boring, cyclical, affairs – usually suppress volatility. Interestingly, Chaikin himself disputes this interpretation. He believes that volatility increases near the bottom, reflecting panic selling – and decreases near the top, when investors are in full accord as to market direction.
But most market players follow the trend. They sell when the VIX is high and, thus, portends a declining market. A bullish consensus is indicated by low volatility. Thus, low VIX readings signal the time to buy. Whether this is more than superstition or a mere gut reaction remains to be seen.
It is the work of theoreticians of finance. Alas, they are consumed by mutual rubbishing and dogmatic thinking. The few that wander out of the ivory tower and actually bother to ask economic players what they think and do – and why – are much derided. It is a dismal scene, devoid of volatile creativity.
Title:
On Volatility and Risk
Word Count:
1364
Summary:
Volatility is considered the most accurate measure of risk and, by extension, of return, its flip side. The higher the volatility, the higher the risk – and the reward.
Keywords:
Article Body:
Volatility is considered the most accurate measure of risk and, by extension, of return, its flip side. The higher the volatility, the higher the risk – and the reward. That volatility increases in the transition from bull to bear markets seems to support this pet theory. But how to account for surging volatility in plummeting bourses? At the depths of the bear phase, volatility and risk increase while returns evaporate – even taking short-selling into account.
“The Economist” has recently proposed yet another dimension of risk:
“The Chicago Board Options Exchange’s VIX index, a measure of traders’ expectations of share price gyrations, in July reached levels not seen since the 1987 crash, and shot up again (two weeks ago)… Over the past five years, volatility spikes have become ever more frequent, from the Asian crisis in 1997 right up to the World Trade Centre attacks. Moreover, it is not just price gyrations that have increased, but the volatility of volatility itself. The markets, it seems, now have an added dimension of risk.”
Call-writing has soared as punters, fund managers, and institutional investors try to eke an extra return out of the wild ride and to protect their dwindling equity portfolios. Naked strategies – selling options contracts or buying them in the absence of an investment portfolio of underlying assets – translate into the trading of volatility itself and, hence, of risk. Short-selling and spread-betting funds join single stock futures in profiting from the downside.
Market – also known as beta or systematic – risk and volatility reflect underlying problems with the economy as a whole and with corporate governance: lack of transparency, bad loans, default rates, uncertainty, illiquidity, external shocks, and other negative externalities. The behavior of a specific security reveals additional, idiosyncratic, risks, known as alpha.
Quantifying volatility has yielded an equal number of Nobel prizes and controversies. The vacillation of security prices is often measured by a coefficient of variation within the Black-Scholes formula published in 1973. Volatility is implicitly defined as the standard deviation of the yield of an asset. The value of an option increases with volatility. The higher the volatility the greater the option’s chance during its life to be “in the money” – convertible to the underlying asset at a handsome profit.
Without delving too deeply into the model, this mathematical expression works well during trends and fails miserably when the markets change sign. There is disagreement among scholars and traders whether one should better use historical data or current market prices – which include expectations – to estimate volatility and to price options correctly.
From “The Econometrics of Financial Markets” by John Campbell, Andrew Lo, and Craig MacKinlay, Princeton University Press, 1997:
“Consider the argument that implied volatilities are better forecasts of future volatility because changing market conditions cause volatilities (to) vary through time stochastically, and historical volatilities cannot adjust to changing market conditions as rapidly. The folly of this argument lies in the fact that stochastic volatility contradicts the assumption required by the B-S model – if volatilities do change stochastically through time, the Black-Scholes formula is no longer the correct pricing formula and an implied volatility derived from the Black-Scholes formula provides no new information.”
Black-Scholes is thought deficient on other issues as well. The implied volatilities of different options on the same stock tend to vary, defying the formula’s postulate that a single stock can be associated with only one value of implied volatility. The model assumes a certain – geometric Brownian – distribution of stock prices that has been shown to not apply to US markets, among others.
Studies have exposed serious departures from the price process fundamental to Black-Scholes: skewness, excess kurtosis (i.e., concentration of prices around the mean), serial correlation, and time varying volatilities. Black-Scholes tackles stochastic volatility poorly. The formula also unrealistically assumes that the market dickers continuously, ignoring transaction costs and institutional constraints. No wonder that traders use Black-Scholes as a heuristic rather than a price-setting formula.
Volatility also decreases in administered markets and over different spans of time. As opposed to the received wisdom of the random walk model, most investment vehicles sport different volatilities over different time horizons. Volatility is especially high when both supply and demand are inelastic and liable to large, random shocks. This is why the prices of industrial goods are less volatile than the prices of shares, or commodities.
But why are stocks and exchange rates volatile to start with? Why don’t they follow a smooth evolutionary path in line, say, with inflation, or interest rates, or productivity, or net earnings?
To start with, because economic fundamentals fluctuate – sometimes as wildly as shares. The Fed has cut interest rates 11 times in the past 12 months down to 1.75 percent – the lowest level in 40 years. Inflation gyrated from double digits to a single digit in the space of two decades. This uncertainty is, inevitably, incorporated in the price signal.
Moreover, because of time lags in the dissemination of data and its assimilation in the prevailing operational model of the economy – prices tend to overshoot both ways. The economist Rudiger Dornbusch, who died last month, studied in his seminal paper, “Expectations and Exchange Rate Dynamics”, published in 1975, the apparently irrational ebb and flow of floating currencies.
His conclusion was that markets overshoot in response to surprising changes in economic variables. A sudden increase in the money supply, for instance, axes interest rates and causes the currency to depreciate. The rational outcome should have been a panic sale of obligations denominated in the collapsing currency. But the devaluation is so excessive that people reasonably expect a rebound – i.e., an appreciation of the currency – and purchase bonds rather than dispose of them.
Yet, even Dornbusch ignored the fact that some price twirls have nothing to do with economic policies or realities, or with the emergence of new information – and a lot to do with mass psychology. How else can we account for the crash of October 1987? This goes to the heart of the undecided debate between technical and fundamental analysts.
As Robert Shiller has demonstrated in his tomes “Market Volatility” and “Irrational Exuberance”, the volatility of stock prices exceeds the predictions yielded by any efficient market hypothesis, or by discounted streams of future dividends, or earnings. Yet, this finding is hotly disputed.
Some scholarly studies of researchers such as Stephen LeRoy and Richard Porter offer support – other, no less weighty, scholarship by the likes of Eugene Fama, Kenneth French, James Poterba, Allan Kleidon, and William Schwert negate it – mainly by attacking Shiller’s underlying assumptions and simplifications. Everyone – opponents and proponents alike – admit that stock returns do change with time, though for different reasons.
Volatility is a form of market inefficiency. It is a reaction to incomplete information (i.e., uncertainty). Excessive volatility is irrational. The confluence of mass greed, mass fears, and mass disagreement as to the preferred mode of reaction to public and private information – yields price fluctuations.
Changes in volatility – as manifested in options and futures premiums – are good predictors of shifts in sentiment and the inception of new trends. Some traders are contrarians. When the VIX or the NASDAQ Volatility indices are high – signifying an oversold market – they buy and when the indices are low, they sell.
Chaikin’s Volatility Indicator, a popular timing tool, seems to couple market tops with increased indecisiveness and nervousness, i.e., with enhanced volatility. Market bottoms – boring, cyclical, affairs – usually suppress volatility. Interestingly, Chaikin himself disputes this interpretation. He believes that volatility increases near the bottom, reflecting panic selling – and decreases near the top, when investors are in full accord as to market direction.
But most market players follow the trend. They sell when the VIX is high and, thus, portends a declining market. A bullish consensus is indicated by low volatility. Thus, low VIX readings signal the time to buy. Whether this is more than superstition or a mere gut reaction remains to be seen.
It is the work of theoreticians of finance. Alas, they are consumed by mutual rubbishing and dogmatic thinking. The few that wander out of the ivory tower and actually bother to ask economic players what they think and do – and why – are much derided. It is a dismal scene, devoid of volatile creativity.
Mail this post
Technorati Tags: risk, trading, volatility
