Implied Volatility for Stops ?

[Glenn]

Here are some thoughts about the use of IV for stops.

The whole point of IV is to estimate volty over a future period of time e.g. to the end of the currect options contract for the stock.
In that sense it would appear to offer something more 'intelligent' than ATR or historical volatility which only look backwards.

As I see it there are two ways in which IV could be used for stops.
One uses a mean IV and the other uses two separate IV's, one for Longs and one for Shorts.

Mean IV: -
Use the front month Options and pick the two strikes nearest to at-the-money.
For these two strikes find the IV of the Puts and Calls.
This gives you four IV's.
Take the average of them and that is the IV to use for long or short stops.

Two IV's: -
The logic here is that the IV of the Puts will always be higher than the Calls due to the event risk.
In essense this means that long positions have lower risk than short positions because the stop is nearer to the entry.
For long positions you would use the IV of the a-t-m Calls.
Conversely for short positions you would use the IV of the a-t-m Puts.

So when you look into IV it does seem that there is quite a performance overhead in using it.
To use either method means that you either need your data feed to be able to provide IV, or you have to calculate it, which is itself complicated - yet another overhead.

Glenn

 

[animal]

I'm not fully au fait with the historical discussion here regarding the use of IV but personally feel that the use of adaptive stops is very desirable.
Other questions arise to my mind when seeking a fully generalised rule.
In particular there are issues concerning the instrument being traded ie stock or index future. The time frame whether intraday or several days.
The price of index options is set by the price of the futures contract for the same expiry. So why not measure the volatility of the underlying (future), when trading futures.

In the case of stocks, if you wish to use IV derived from options, those options must exist and be reasonably liquid. This might severely limit the choice of stocks.
In the case of the UK market it would be to 70 or 80 anyway at a guess, from thousands.
On the Nasdaq with which I am unfamiliar, there are many more but my gut reaction is that similar restraints may apply. Please put me right on this.

To my mind a sensible approach is to use a standard statistic technique such as Standard Deviation applied over a suitable sample of whatever instrument is chosen eg 20 or 30 days in the case of multi day trades or 20 or 30 units in the case of intra day trades. Stops are then calculated against recent traded range plus or minus (say) 0.1 *std dev.
This has the advantage of objectivity and universality.

The volatility implied by options might be regarded as indicative of "smart money" and possibly future trends but also includes bid/ask spreads and other market demand related variables.

As an aside, I note that Tradestation global server has a function to provide an IV calculation which is update daily according to a chosen model as part of its housekeeping. I'm not able to comment further on that for lack of day to day experience with the product. That might be usable but a bit restrictive in an intra-day context.

At the end of the day there are players shooting for the stops to release trading opportunity and you don't want to get hit by them.
That implies increasing your risk tolerance at the same time as putting on a lifebelt.
That maps into similar issues affecting risk/reward ratios at the same time.