There is a natural volatility to financial instruments and the level of volatility itself fluctuates. A sophisticated trader can generate useful signals by studying and mapping volatility and its changes.
Volatility is easier to understand intuitively than to quantify. Or rather, it can be measured in several ways, which can lead to conflicting signals.
How do we measure historical volatility?
One measure of volatility is the spread in buy-sell quotes. If the spread is high, the instrument is volatile. The width of a spread is important for a day trader or jobber because it is a benchmark for minimum gain or loss. However buy/sell spreads are also a function of liquidity. If there is a high liquidity and depth, there is a small buy/sell spread. So, an instrument with large price swings could have a low spread.
This brings us to a second measure of volatility. An instrument, which fluctuates by a large amount from day to day, is volatile. We can look at the closing quote of Day One and compare it with the closing quote of Day Two.
However, an instrument could show small changes in closing prices while having large two-way intra-day movements, which cancel out. On a measure of closing price comparison, it would be as a low-volatility instrument, while being high-volatility on a third, commonly-used measure.
A third measure of volatility is the daily spread- the high-low range of the day. If this is expressed as a percentage of the closing price, it gives us a measure of intra-day volatility.
Thus, we may have an instrument with low spreads, low daily volatility and high intra-day volatility or different combinations of all these measures. These contradictory results occur when we deal with different timeframes since the same instrument may show different volatility in different timeframes. Derivative traders also encounter the concept of implied volatility.
What is implied volatility (IV)?
A derivative such as an option, are priced with premiums. In a "zero-volatility" instrument, that premium would factor in the interest rate (known), the current price (known) and the strike price (known).
In practice, underlyings are not zero-volatility. Hence the difference between actual premiums, and the premium theoretically derived for a zero-volatility underlying, must reflect the future expectations of volatility. This is IV. There are several fairly complicated measures of IV, which work "backwards" from theoretical option-pricing models such as the Black-Scholes model.
The difference between historical volatility and the IV is a pointer to market sentiment. In practice, a low IV scenario is generally perceived as bullish while a high IV scenario is seen as bearish. The Chicago Board of Trade (CBOT) releases a popular Volatility Index (VIX) which sums and averages volatility across the entire strike-price-chain of options.
The VIX is known as an "index of fear" because it tends to be low when prices are trending up and high when prices are trending down. From studies in Indian markets, (where derivative prices are available for far shorter periods), it seems we could construct a similar Indian index of fear because the IV-price relationship seems to work similarly.
Intra-day measures of historical volatility offer similar results. That is, in bull markets where prices are trending up, volatility tends to be lower. In bear markets, when prices are trending down, volatility tends to be up.
At market peaks and market troughs, there is often a shift in volatility accompanying the price trend changes. Thus, historical volatility can even be a leading or confirmatory indicator for market trend reversals.
For a trader, who is not comfortable with sophisticated mathematical calculations, the historical volatility relationship is very useful. It is easy to calculate and it may offer actionable signals.
In India historical volatility is meaningful only for the 100-odd stocks and three indices covered in the F&O segment. That's because daily volatility is constrained by circuit breakers for stocks outside the F&O group.
The main graph (Nifty Volatility 2000-2006) is a superimposition of intra-day volatility across daily Nifty prices since January 2006. We've taken a basic "high-low range as the percent of close" as our measure of daily volatility.
For example, if the Nifty registered a high of 3030 and a low of 2970 on a day when it closed at 3000, our measure would suggest a volatility reading of 2 per cent. This would probably be a more accurate measure if it was converted to logs but it is not necessary. A close look at the chart shows that volatility tends to be low when the market is climbing (July 2004- April 2006) and vice-versa, volatility tends to be high when prices fall (May-June 2006, Feb 2000-October 2001). Volatility also tends to be high close to market bottoms (May 2004, May-June 2006).
The second chart zooms-in on March-August 2006. Volatility was consistently lower during the bull run of March-April and it spiked in May-June as the market fell. Again, after the bottom in mid-June, the volatility has dropped while prices have risen.
Can volatility be turned into a trading tool? You would need to do more sophisticated analysis but it is possible. The average daily volatility across this six-and-a-half-year period was 2.09 per cent and the standard deviation was 1.37. Using these as benchmarks, we could look for signals, markedly higher or lower than the average.
This article was originally published on September 01, 2006.
Ask Value Research 
New episodes every Friday 
New Process Alert 
Seems like you've uploaded a wrong QR code. Please upload the QR code downloaded from MF Central platform.
