Probably the greatest discovery ever (2)

In the earlier post probably greatest discovery ever I said, as being a option trader, that probability is such a good thing especially on the strike picking because apart from reviewing all the technical indicators on the supports and resistances and the analysis on the open interest of other derivatives like the Futures and CBBC, the decision on a strike boils down to whether the strike will become ITM or OTM at the settlement day. Notwithstanding there are Straddle, Strangle, Butterfly...and many other option strategies but boiling it down to the most foundamental components there are only Long and Short while Long prefers ITM vs Short avoids ITM. Therefore the tool that helps predict whether a particular strike will go ITM is paramount to an option trade.

In that post I have slightly touched on the saying that in option trading shorting at a strike with p=0.1 of being ITM is usually safe in a normal market situation. Perhaps not every reader is familiar with option trading so I am going to explain a little bit that in the option trading world, some traders use option as a hedging of their positions in assets holding like stock while some just trade for profit without any need on the hedging of whatsoever. For the latter, profit could be from receiving the premium on their short position or the price difference after the settlement of their initial long position. Long position traders is betting on the strike(s) of their trades will go into in-the-money (ITM) so their profit can grow exponentially but Short traders just prefer the opposite because basically they are the counterparts of the Long traders despite not necessarily be of any of a particular trade concerned. Simply speaking when a Long trader makes money there must be a Short trader losing money in a particular trade because option trading is a zero sum game . Therefore Short traders never want to see the strike(s) of their position go into ITM.

I hope the above little explanation can shed some light on why I said a tool that helps to predict whether a particular strike will go into ITM is so important to all option tradings. That also explains why probability is so helpful because when the probability of a particular strike is known then it is just all about a trader's discretion on whether the risk of the trade is justified. In that earlier post I said luckily the probability of a particular strike in option trading is, unlike some other events which their probabilities are difficult to tell, rather easy to be determined, scientifically and objectively. 

For any seasoned option traders, they should be well versed in the jargon in option trading like Delta, Theta, Gamma and Vega...etc. Among these jargon, Delta is defined as the ratio of the change in the price of an underlying asset with the change in the price of a derivative or option. However on the other hand, it is also known as the probability. For example, a strike with 0.2 Delta means p=0.2 or 20% possibility being ITM. One must understand that the Delta of a particular strike does change along with the price movement of the underlying asset though so that probability is better seen as a snapshot of the moment concerned only. Therefore it is rather risky to make a trade base on the Delta reading of a strike to assume it will remain the same till the settlement day especially when there is still a long way to go. 

Considering the drawback of the Delta, some traders resort to the option pricing theory  to use the Implied Volatility (IV) to work out how likely a particular strike of an option will be exercised, ie., being ITM. In real life, IV is quoted by market makers who are usually more market intellectual to judge the riskiness of a particular strike so supposedly to be able to assign the corresponding IV. Basing on the quoted IV, traders can work out the likelihood of an option will be exercised at the expiration according to the formula of the pricing model. However, apart from other elements, the underlying asset price is also one of the variables used in the formula. Therefore using the IV on the calculation of probability literally has the same drawback as the using of Delta. Meanwhile, to the contrary to the Delta, market makers often quote a higher IV when the market is volatile but then when an higher IV is used to be the variable on the calculation in a specific market volatile moment the result tends to be fluctuant thus making it less reliable.

In light of the pitfall of the Delta and option pricing model, I prefer to refer to the historical data. For example for a spot month trading, the past record of the percentage of the monthly closing price vs opening price is used. These past percentage figures, say for the last 10 years, are ranked from smallest to biggest then it is easy to find out which percentage is within the 10% of the smallest as well as the 10% biggest. Those 10% smallest should be negative representing the most extreme adverse months while those 10% biggest are positive representing those extreme good months in the past 10 years. The thresholds right before these 10% smallest and 10% biggest biggest are what I am looking for and are used to become the risk determinants. That is to say, the strike is chosen at the same percentage of these thresholds comparing to the month opening price of any particular spot month trading. Naturally whether it is 10% smallest/biggest or whatever percent just rest on the risk appetite of the trader at their own discretion. 

The above methodology is the application of the probability based on historical data. This idea might not be most scientific but it has taken the history into account. Some argue that history is something happened in the past and it is not necessarily able to predict the future. As Mark Twain puts it that history doesn't repeat itself but it rhymes. Those good months or bad months could be as the result of different causes but the magnitude of their impact in the market were clearly marked in the history which can be used as a reference to whatever happening or going to happen in the future. For example if there will be a black swan event in this month which is not bigger in scale than the financial crisis in October of 2008 then it is quite reasonable to assume the plunge should not exceed the magnitude in that month which falls into the 10% adverse months bracket. Therefore any strike with smaller decline percentage than was that in October, 2008 should be a safe one no matter what the black swan event is.

Having said, the application of probability based on historical data doesn't come without pitfall of its own. The prerequisite of this methodology is the binomial distribution of the past data. Meanwhile even if this criteria is met, if the black swan event is so big and bigger than any past events throughout the statistical measured period, ie., 10 years in the example, then the assumption is simply not applicable. Luckily an as longer as possible period in question should be able to address the issue, largely if not completely. What still makes this methodology fails is that if there will be a mega crisis in scale which has never happened in human's history before like the outbreak of a worldwide nuclear war then the historic statistical data just fails to do their job. However, the point is, if there will be such a devastating nuclear war, then do we still exist? So do we still care about the option traded?