writing option contracts on the US futures markets

                                              by John R Pretorius



I live in Johannesburg, South Africa and make my living trading the US commodity futures and options markets. Recently I found myself increasingly specializing in options. More specifically, writing options with short time to expiry. The aim is to collect premiums and have the contracts expire worthless. This is a simple strategy that I hedge with protective stops in futures. It will never double my margin in a week, but it has provided me with consistently good profits and low risk.

The problem was that I was spending many hours each day trying to pick out the best option contracts to sell. I own an excellent program called OddsCalc that is designed, among other things, to quickly calculate the “odds” of an option expiring worthless. But I still needed to feed in each strike of each contract to generate a list of candidates with high odds. Then I had to go and find the premium, multiply it out by some obscure factor and only then could I hope to get SOME of the high premium, short time to expiry options that we are all looking for.

I have found a way of automating this process. I wrote a computer program to scan through the option prices every day and essentially do an “odds to expire worthless” calculation for each strike. This is the statistical probability of the underlying futures price not reaching the strike price by date of expiry. For those items with odds over a certain value (I use 70% at this stage) I then calculate the dollar premium and rank the items in terms of percentage annual return. The idea being that I am investing my margin forn days (the days to expiry) and receiving a return (the premium minus brokerage). The annualized return is thus an easy way of picking out the best possible options every day. If your criteria for choosing options are different, say only odds over 80%, then you would simply ignore those lines between 70% and 80%. Same if you only want to look at, say margins less than $2000, or time to expiry of less than 20 days.

I have run it now for a few months and the results are better than I had ever hoped for. I intended it for my own use, but I now realize that it might have some commercial value. So here comes the sales pitch right? Well yes and no. The good news is that I want to give it away for free. Not the program, but the results emailed on a daily basis. The program was never intended for use by anyone else and as a result it is cobbled together with the software equivalent of bits of wire and chewing gum. To get it into a form where it could be sold, or even given away, I would have to invest much more time on it, and that isn't what I want right now.



I am prepared to e-mail the output listing every day at about 9-am SA time (1-am CST in winter) to anyone who wants it - free. Why so generous? The reason is that I have stumbled onto another idea in options trading that I believe will have enormous value. It will take at least six months to develop (I wrote that in 1999 - still not finishedJ), and meanwhile I would like to build up a prospect list of potential customers for what will be a valuable (as in expensive) item. I need to find people who are already familiar with futures and options trading. If those potential customers are making lots of money in the meantime, so much the better. Hence the free e-mail. I genuinely want people to get hooked on using my products.


Checking Results

It is possible that some of the conversion factors have been entered incorrectly. These factors change occasionally and data vendors also change the basis of their reported prices. I sometimes miss these events, or pick them up only after some days or weeks. The daily prices also contain random errors. My data sources are no better than anyone else’s, and even the best data vendors make mistakes. You should check the numbers before entering any trade. If it looks too good to be true, then it probably isn’t, and with a bit of experience, you will probably get a feel for what is reasonable.


Report Columns: (see example)

Contract My abbreviation of the option contract name and month.

Strike The strike price and type (P=put C=call)

Future Price The closing price of the underlying future as at the first date on the heading. Please remember that the futures contract month is not always the same as the option contract month. For example Jan T-Bond options are based on Mar T-Bond futures. Same for currencies.

Expiry Date The Last Trade for Options date or Option Expiry date.

Volatility  The volatility is the prime factor in determining the probability or “odds” of reaching or not reaching the strike before expiry. Statistical Volatility (SV) is a measure of the actual price movement of the underlying futures contract. Implied Volatility (IV) is a number which is derived from the option price. It is an attempt to quantify that portion of the premium that is attributable to the price “jumping around”. My calculation is the SV based on the front month of the underlying, and is presented in the normal way as the annualised Standard Deviation of return expressed as a percentage. It differs from most published work in that it uses only the last 10 trading days as a sample. This is generally considered to be very short-term, however I think it tends to more readily highlight contracts that are "settling down" and where the actual volatility is dropping while the implied volatility (option premium) is still high.

Premium Points The option price in the form quoted by your broker. Remember that this is the price yesterday (or at the date on your report). Because of time decay, it will not be the same today even if the future price and IV stay the same.

Premium Value “points” means something different for each contract. I have done the conversion into dollars for you.

Approximate Margin The report is ranked in terms of Return on Investment (ROI) “Investment” in this case is the margin that we are putting up. The margin on options is calculated at the exchanges using a very involved formula called SPAN that I am not able to duplicate (see my reasons in the FAQ section). Fortunately for our purposes, the futures margin serves just as well - hence the word approximate.

Days to Expiry The other important factor in determining the probability of reaching the strike by expiry date. Also to annualize ROI

% Annual Return Calculated as (($Premium - $30) / Margin) x (365 / Days) x 100. The $30 is for brokerage. A high value in this column neatly encapsulates all the things we are looking for - high premiums, low margins and low days to expiry. By sorting on this column - highest first - the most profitable options to write automatically appear first.

% Odds to Miss The statistical probability that the option will expire worthless based on volatility, days to expiry and distance between the future price and the strike.

Volume Traded Use this to keep away from illiquid markets and thinly traded contracts. Note that the volume on futures contracts is always one day late, so for instance a volume of zero does not mean that the contract did not trade on the day we are looking at.

Last Trade Date on which last trade took place. Avoid contracts that have not traded the previous day.



I am frequently asked if it would be possible to send the list in a different sequence; lowest days to expiry first, or highest odds first. The simple answer is that this would destroy the most powerful feature of my report - the ROI encapsulates all the things we are looking for - high premiums, low margins and low days to expiry. By sorting on this column - highest first - the most profitable options to write automatically appear first.

However, it is very simple toresequence the report yourself using the built in features of MS Word. Simply click on Table then Sort. Select your preferred column as “Field n” and choose Ascending or Descending.

Or how about using MS Excel to extend the report with your own calculations. While in MS Word, click on File then Save-As. Type in a name for your file. Under “Save as Type” choose “MS Dos Text”. Load Excel and click on File/Open. Type in the name you used to save the file. For the dialogue that follows, choose Delimited/Tab and Finish. The report is now available in spreadsheet form for you to sort or recalculate to your own specifications.



A good website for futures & option prices, specifications and expiry dates is http://www.bohlish.com  My best source of links to other trading related sites is SmotAss  I get futures margins from optionsXpress

Queries and feedback

I would welcome feedback - criticisms, suggestions or questions by e-mail to johnpretorius@postnet.co.za


"Remember the report is sorted by return on investment, so the high premium, short time to expiry items will automatically be listed first"


Frequently Asked Questions

Well, maybe not that frequently, but I promise that they have all been asked at least once, and answering them is a convenient way of getting across some of my ideas.



Q. You list a June Corn contract. There is no such thing. I can only find a July contract?

As I point out in my booklet, there are many markets where option contracts exist without an equivalent futures contract. In the above example, there is a June Corn option (expiring on 22nd May) and a July option (expiring on 19th June). There is no June future, so the July future is used as the basis for both of them. All very confusing I know, but you need to understand this if you are going to trade options.


Q. I have a query with regard to some futures prices which you quote…

The problem stems from the fact that certain markets also trade a night session. Some downloaded prices are for the day session only and some for a combination of day and night. Neither is right or wrong, only different - just to add a bit more confusion to an already confusing business.


Q.  Might there be a limitation of liquidity if the volume of specific option strikes fall below a certain level?  Might we define that level?  I imagine different markets might have different personalities but look to your guidance for this type of information?

Liquidity certainly is a problem, but I am not able to give you any specific guidelines. In one of his books Joe Ross gives the minimum previous days volume at 10 contracts. Not very scientific but I am afraid that is the best I can do.


Q.  If the premium money received is below an amount such as $300 or so. I'm not sure that the low payments warrant entering any trade at all?

Remember that my listing is based on return on investment. A "low" premium of say $70 on a margin of $1000 for 1 day is 1500% (after deducting $30 for brokerage). A "high" premium of $1000 on the same margin for 180 days is only 200%. If the risk (in the form of  %odds) is the same, then the first choice is much better than the second. That's what my system is all about.


Q. Might there be a cut-off point below which the amount of option premium received via a sale is too low to consider?  Might this value be somewhere around $300?  The commissions play a role and there is always the possibility of slippage or even errors.  What I have been doing with your Option Scan data is the following:

   1.  Sort it by premium value.

   2.  Delete all the options with premium value lower than 280.

   3.  Re-sort it via % Annual return.

This elimination of the low value premiums does help to scan through all of your dat  My next step may be to delete certain markets.  For example I'm not comfortable with any of the Dow Jones options.  By this type of sorting using Microsoft Excel I'm able to better narrow the field of potential options sales.

I don't necessarily agree with your cut-off at $300. Remember that I am already factoring in the low return by including an amount for brokerage. The low premiums that appear on my list are there on their own merits.

What I do like about your method is that you are using your own criteria for selecting the items that YOU consider important.


Q. Why does the Corn 2300C ( 7th down on the 13 May list) not work out according to the above formula My calculation says (($63-$30)/($405))/8*36500=372, yet you show 366. The first 6 options on the list work out OK. Also, the point value for the Corn 2300 call is shown as 12, it should be 1.2.

The Corn option price quoted as 12 is in fact shorthand for 1 and 2/8ths of a point. At $50 a point the premium is $62.50. The %ROI is then 366.126

As to whether it should be specified as 12 or 1.2, I have decided on this notation after a great deal of thought. I eventually concluded that the least confusing method was to use the strike price without any decimals and then to quote the futures and option prices in the same dimension. Attempting to put decimals anywhere very quickly becomes a nightmare. Try it.


Trading method

Q. How much of the report do you make use of?

I never print more than one page of the report, and I never look at more than about the first 20 lines. The 3 pages that I send out are for those people who want to use my data as a basis for their own analysis, probably using MS Excel as explained in my booklet.


Q. How do you use OptionScan?

My strategy is basically to put a future buy stop (for a call) at the strike plus the premium. If it gets hit, then I am breaking even except for brokerage. I then have to worry about the price going back down below the strike and whipsawing me. This is the one cause of loss. The other is gapping or going lock limit past the stop. See “Trading Rules” further on in this FAQ.


Q. What criteria do you use to select or reject items from your list?

I filter from the top down. I will take Odds-to-miss of 70% (some people only take 80% and above). I look for volume of 100 and above, and days to expiry between 5 and 20.


Q. Do you take account of “catastrophic events” such as 9/11?

History shows us that the stock market is occasionally capable of dropping dramatically in one day, however it never goes up so dramatically, and I can’t think of a scenario where the majority of shares suddenly increase by 10% in one day. I am therefore I am happy to sell calls on Stock Indexes (DJ, SP etc), but I avoid selling puts.

Some agricultural commodities (coffee is a good example) tend to have weather related scares at particular times of the year. These take the form of sudden massive increases in the price when the crop seems to be in danger of being harmed by something like a hurricane. Again I can’t think of a contrary scenario where these commodities suddenly drop in price – Brazil is not going to wake up one morning to find that the coffee crop has mysteriously doubled overnight. So I am happy to sell puts on these contracts, but avoid calls.

Having said that, there is also a group of statisticians who believe that such contrary scenarios are not only possible but inevitable – the stock market will double one day and coffee will halve. See “Black Swans” further on in this FAQ.


Q. If you have a stop in the futures that is hit what do you do next? Do you then exit both the option and the futures at the same time at the next available opportunity?

No. If you are going to exit the trade by buying back the option, then you need not buy (for a call) the future. If on the other hand you hedge with a future stop (which I do), then when it gets hit, you use it to cover the option at expiry - provided that the option is still in the money on the expiry date.


Q. I am extremely interested to hear how you protect your trades, if at all. I mentioned to you that I have been trading "naked" and have been watching the 10 minute delayed prices. Well that would not have helped me in Jap.Yen yesterday; thank goodness I was flat the market. A protective stop, depending on where one put it, would only have helped partially as it would not have catered for the gap upwards (you would only have been filled within yesterdays price range)?

Not an easy business this. The Feb. Jap. Yen 8900 call, before the exchanges opened yesterday, had an 88,3% odds to miss, yet it opened in the money. I find such situations rather frightening and as a result am not comfortable trading at the moment. Interested to hear your comments should you have the time.

No easy answers either, I'm afraid.

 I set up my option trades with a protective stop at about the strike plus the premium (for a call). This means that when the stop gets filled, I have broken even or at most lost the brokerage. Remember this is a trade that has gone wrong. On a futures trade that has gone wrong you don't get the luxury of breaking even.

O K that isn't the whole story. The first danger is that the future price now turns around and goes back out of the money. You can sell the future again or hedge it with a sell stop (for a call) on a far month. This creates a spread on the future when it gets below the strike, effectively freezing the price at the strike. The stop price is the strike plus or minus the gap between the near and far month. The other danger - as you point out - is that the price gaps past your protective stop.

In both these cases you end up losing money. But these events happen in only a small percentage of trades.

For me options are still more forgiving. In futures, if you get it right you make money, else you lose money. In options, if you get it right you make money, else you probably break even.

I find that I am using my printouts as a general starting point rather than for specific trades. Yesterday for example, there is no way I would have put in an order for an 8900 call on yen without first checking the ruling future price. Having found that it was up, I would immediately have been looking at the 9000 call.

I took a trade in yen. Fortunately I thought it looked bullish, and I sold 8500 puts last Friday at 22 points for $275.

I take your point about protective stops getting taken out. But again my solution is to trade options. You can position them far enough away that they only get taken out when your trade goes wrong. In futures trading, your stops get taken even when the price is generally going your way. Either that or you have to move them so far away that when the trade goes wrong it really hurts.

My aim is to make 10% of my margin per month. I haven't got there yet, but I believe that it is very reachable at reasonable risk levels.


Q. I believe that in your own trading you use a system with very specific rules. What are they?

Here are the rules of 2 trading systems based on OptionScan:



Trading Rules - System 1

Sell the option on a market order.

Protect the position with a futures stop order. For a call option this will be a buy stop at a price calculated by adding the option premium to the strike price. For a put option this will be a sell stop calculated by subtracting the premium from the strike price.

 If the stop is hit then:

For a call option place a new futures sell stop at a price calculated by subtracting half of one days range from the strike price. For a put option place a new futures buy stop at a price calculated by adding half of one days range to the strike price. Provided that the new stop is not hit, then the result will be break-even minus brokerage and any slippage - i.e. a small loss.

If the new stop is hit then buy back the option on a market order. The result will be a loss equal to the cost of this buy back plus half the days range plus brokerage and any slippage (the premium that we originally received has been offset by the loss of the futures trade moving between the first and second stops). The loss is typically equal to about 80% of the original premium value - fortunately it doesn't happen very often.

If the first stop is not hit, but the option expires “in the money” (above the strike for a call or below the strike for a put) then we need to cover the futures position on the expiry date or shortly thereafter. This entails buying a future contract for a call or selling a futures contract for a put. This is in order to fulfil our contract giving the buyer of our option contract the right to buy (call) or sell (put) a future contract at the strike price. The result is a profit equal to the total premium minus the loss on the future trade (the difference between the cover price and the strike price) minus the brokerage for the extra futures trade. This profit spreads evenly between -5% and 90% of premium value.

If the first stop is not hit and the option expires "out of the money" we do nothing. The result is a profit equal to the whole premium.


Trading Rules - System 2

The rules are the same as above, but we sell a “strangle”. A strangle is a call option and a put option of the same contract for the same month. Choose both from the OptionScan listing.

There are now two initial stops, calculated by adding and subtracting the “total premium”. This is the sum of the premiums from both legs of the position.

If we need to exit the position by buying back the option (when the initial stop was hit and the future price has now gone back out of the money) then we buy back both options.


New Project 

Q. What is your new project?

My new project involves three major enhancements to the present model.

The first is what I call an "expectation scenario". It is based on "expectation payoff" - if you have a 60% chance of winning $20 and a 40% chance of losing $10 then the expected payoff is (60/100*20) + (40/100*-10) = 8. This is statistically sound provided the choices are mutually exclusive. Basically I build a matrix of all the paths that the underlying future can take from today until the option expiry date. Each cell contains a probability and the endpoints have a payoff (profit or loss). Using a modified random walk technique, it is possible to arrive at the expected payoff for each strike/contract given the premium, days to expiry and volatility. This one number will enable different strikes of different contracts of different expiries all to be compared - not only for return, but also for risk. The results will throw up out-of the-money, in-the-money, strangles, straddles and what-have-you all ranked on a common denominator.

I had hoped to find a theoretical basis together with equations for at least some of the areas covered, but there seems to be surprisingly little, except in the area of random walk. However, I still hope to find something in the area of  "derivative probability" - I even have to invent my own terms for this stuff. Although I have a degree in statistics, I am by no means an academic, and trying to do research is a mission of its own - academia is driven by different goals to mine. Making money in options trading doesn't feature highly on the average untenured professor's to-do list.

The possibility of finding some all-encompassing formula seems to be diminishing, but I am probably going to take a pragmatic engineering type approach to solving the problem. I shall build a large matrix and then use brute computer force to crunch the numbers. It may not be the most elegant solution, but I am reminded of a quote from WilliamGallacher in his book The Options Edge - "It isn't rocket science, but hey the thing flies".

The second development involves taking a view on the direction of the underlying price. One of the absurdities of option modeling theory (Black Scholes et al) is the assumption that the price of the underlying futures or stock at expiry will be the same as it is today. The above model, however calculated, conforms to the overall rules of normal distribution statistics. If in addition we also have a statistically based view on where the underlying futures or stock price might be at expiry date, then it is very simple to incorporate this into the model.

So to the problem of predicting price. At one stage I was tending towards the belief that  futures prices were far more random that most people believed, and becoming more so as thousands of computer models jerked the price around. My belief was reinforced by the large percentage of my capital that I lost trading futures. I now have reason to think otherwise. I have seen solid evidence of accurate projections, especially in the areas of Elliott wave and other pattern based studies. The problem with most of these projections is that their entry signals tend to be based on the thrust of breaking through support and resistance levels. Why is this a problem for an option trader? Well, a future trader can make money by setting stops and waiting for the breakthrough. It doesn't really matter if it happens tomorrow or in two weeks time. Also, the amount that it goes up/down or the timescale is not really critical - sure it is great if we hit a powerful trend, but if we have good exit signals, who cares when or where the exit is so long as it makes money. For options we need to know how much the move will be and by when. So even the best technical indicators do not help answer the sort of question we need answered - “Will the price of Dec Crude be nearer to $18.00 or $22.00 in 47 days time?”

As an alternative to technical indicators, I began looking at projections based on seasonal history. For the last year I have subscribed to the monthly spreads and seasonals newsletter from Moore Research - http://www.mrci.com . As an example of what they do, the following trade is recommended in the June 1999 edition: Buy Sep 10yr T notes on approx. 7th June and sell on approx. 16th July. They then give a detailed breakdown of the same period for the last 15 years. It shows that the trade on the same dates would have been successful in 14 out of those 15 years giving an average profit of $1408 per contract. They do the same for another 14 futures and another 15 future spreads. They publish their results, and it is remarkable how consistently accurate they have been. Great stuff, but unfortunately the Sep T Note option doesn't expire on 16th July, nor is 7th June necessarily going to be the day when T Note options have nice juicy premiums. So I shall do my own seasonal analysis based on historical futures prices. Instead of looking at the most optimum periods of the year, I look at the actual period between now and expiry date. I shall also do some basic pattern and regression analyses to select only those years that best conform and “bend” the others into shape.

 The third element uses similar seasonal and pattern techniques to project the implied volatility, which I believe is more cyclical - and hence more predictable - than the futures price.



Recent questions (since Jan 2005) 

Q. Your strategy of writing options is not a low risk strategy even though it may appear to be due to the black swan problem. (Ask Victor Niederhoffer). An excellent book to read on this subject is “Fooled by randomness” by NassimTaleb.

I never claimed that my system was low risk. Like most responsible system developers I strive to maximise the reward/risk ratio as best I can.

I read an article “The Black Swan: Why don’t we learn that we don’t learn” by NassimTaleb which is probably an extract from his book. His premise is that systems such as financial markets do not conform to Gaussian distribution and are subject to “wild uncertainty” and that for these systems past data is totally useless in being able to predict what he calls “surprises” – huge outliers that in financial markets take the form of massive price movements in a short space of time. Fine, but then perhaps we should not be trading any markets. It always amazes me that this particular argument seems to be reserved only for option sellers along with the obligatory reference to Niederhoffer.

Victor Niederhoffer will go down in history as the fund manager who lost $130M, much of it his own money, in two spectacular option selling trades in 1997. What is often forgotten is that in spite of his education and experience, he took a huge risk that was the equivalent of betting the farm on two rolls of the dice. But is he held up as an example of irresponsible gambling? No way. Somehow “option selling” gets the bum rap. If he had taken the same huge risks in the stock market and lost, would conventional wisdom now say “Ah remember Niederhoffer – trading stocks is a dangerous mugs game”? I think not.


Here are some questions put to me by Jeremy Lyell on 13th May 2005


I am finding your daily scans very interesting and am putting together a trading plan that hopefully will work for me. Some questions, please :

Hi Jeremy

I'll try to follow each question with an answer below. I will use examples from OptionScan today (2005/05/12 calc on 2005/05/13), part of which I have appended below

* with your 10% monthly goal in mind, do you base your capital requirement per position on the margin requirement alone ? I understand that this will always allow sufficient capital to enter a covering futures position where required, but it leaves little margin for error if the position retraces. I am thinking of allowing 2x margin per position with a 5% monthly goal.... 60% pa would not be too shabby ;o).

Firstly, remember that the margin that I use (futures) is already overstated vis-a-vis the options SPAN margin. Secondly, I tend to look for some portfolio diversity by not committing my whole capital to one trade at a time. Let's look at an example: The June Crude 4950 call has a futures margin of $5434. I looked up the SPAN margin at my broker. It is $5585, but your account will be credited with $540 (the received premium) leaving a net $5045. Say my starting margin was $13000. After selling the 4950c, it would show $7955. We need to commit $5434 of that to the futures hedge stop leaving $2521. Now the question is, if I want to enter another trade, do I look for one where the $2521 will cover both the options and futures margin - i.e. one that has an approximate margin of $1250? The conservative answer is yes. My approach is a bit more gung-ho. I might for example take the June Copper 14200C on the assumption that Crude and Copper are uncorrelated. The SPAN margin is $2269 ($2869 - $600 premium). My margin account would reflect $7955 - $2269 = $5686 That's enough to cover either one of the futures stops, but not both. Whether you consider that dangerous trading is up to you.

* do you favour strangles where possible ? I have heard it said that to write an option that produces premium income = or > than 50% of the margin is a good target, but can most often only be achieved using a strangle (which if I understand correctly is very favourably treated under SPAN margins).

 Premium income as a percentage of margin is a meaningless number because it doesn't take time into account. For example the list today has the Sep Coffee 14000 call with a premium of $2958.75 on a margin of $3500. That's 84%. Great, but expiry is in 91 days time.  So in terms of annual return, it's nowhere near as good as the June Crude 4950C with a 10% ratio, but 4 days to go.

 Yes, strangles are a good way to maximise the trading model that OptionScan is based upon, and this is largely because of the favourable SPAN margins. It is very difficult for me to get these margins on a daily basis. Firstly it needs the rental of a computer program, but even worse is a finicky updating procedure that I just cannot spare the time for at this stage. However let's look at an example from today's listing using actual SPAN margins. I am able to do a SPAN calculation online at my broker (Xpresstrade).  Today the first 2 lines give a perfect strangle. The margin for June Crude 4950C is $5585, and for the 4750P is $4268. The margin for the strangle is $5471, and bear in mind that your account would be credited with $1030 ($540 + $490) premium, which means that the strangle would only "cost"  $4441

*do you watch your positions intraday or do you generally ensure your covering futures stops are placed pre-market and review end of day ? I think my plan will include using my futures trading method to 'trade into' the strike price area where possible, rather than strike price + premium, although I want to minimise the time spent in front of the computer during market hours (for psychological reasons : too tempting to fiddle with positions).

I watch until I am in, because the futures stop levels are calculated using actual option fill prices. After that I leave it alone.

* do you think it is statistically and also 'real world' sound to write a larger quantity of higher probability options, say 90% or greater, than much higher value 70%+ ones (to achieve a significantly higher winning %) ?

Some of the subscribers only look at odds over a certain percentage, say 80%. My new system uses a more sophisticated calculation whereby the odds are incorporated into the return to give an "expected return". The new system is coming "real-soon-now" - Yes I know I've been saying that for more than 4 years. The guts of it is a pattern analysis (of underlying markets) algorithm that produces results in terms of probabilities of hitting certain targets within specified times. These would then replace the implied volatility as a basis of calculating the "odds" and the odds will in turn be used to calculate expected return using the resulting normal distribution curves. You could do a crude expected return on the current listing. For example, on the first line, you could say that the June crude 4950 call has a 72.9% probability of missing and therefore an expected return of 72.9% of $540 = $393.66

Here is the partial listing used in some of the above examples:




Here are some questions put to me by Robert Pisani 25th May 2005


1 - I don't understand your break-even statement about your strategy in System 1.  If (for a call) the underlying rises to the strike of the option quickly enough, the option premium will rise and you won't be able to break even.  If your stop gets hit and you hedge with the underlying, unless you use delta to ratio hedge you are exposed to movements in the underlying.   For example, suppose you sell the first call on today's list, a Jul Corn 2300 at the price you have listed, 46 (what is "Value" -- theoretical value?).  The futures is now 2212.  If the futures rises to 2346, depending on how many days are left to expiry the price of the call may be much more than 46, perhaps as much as 100 or even more.  At this point the delta of the call will be a little more than .5, and if you buy a full underlying, you will be net long half an underlying and exposed to a decline in the underlying.

 You need to understand that the reason for hedging with a futures contract is in order to deliver at expiry date when the option gets exercised. At or after expiry date, the price of the option (premium) is irrelevant. So what if it is 100, or 1000 for that matter, all the purchaser of my call option can do is exercise that option i.e. "call" on me to deliver a Jul Corn futures contract, and bingo, I will have that contract to deliver because I bought it when my stop got hit at 2346. The futures price at expiry is also irrelevant provided that it is higher than the strike price. What to do when the price goes back below the strike is another story, and is dealt with elsewhere in this FAQ - see “Trading Rules”.

 The answer to your question, what is "Value" - theoretical value? No it is the dollar value of the premium.  

2 - The only reason to buy the underlying is to protect your short option position against a further rise in the underlying, but you don't want to get into managing a frontspread, so it seems that rather than buying the underlying you should just buy back the option and take your loss.

That's up to you. My analysis shows that I will make more (or lose less) by managing the frontspread. 

3 - Your list identifies interesting writes, but the list might be more useful if it identified the probably profitable spreads that could be built around these interesting writes.  Perhaps the premiums are too thin to overcome the costs (commissions and bid/ask losses) of the other components of the hedge -- is that the case?

It's up to you to use my list as the raw data to build further. A strangle for example can be quickly identified by visual inspection. 


I understand that a loss occurs only if the underlying rises quickly enough to cause the option premium to increase faster than the option's time decay causes it to decline.   (For calls) which of the following do your probabilities signify?
a - the probability that the underlying will not exceed the strike on expiry
b - the probability that the underlying will not exceed the strike anytime between now and expiry, or
c - the probability that the underlying price will not rise quickly enough to overcome the time decay 

(Note: c - would be hard to compute analytically but could be done with Monte Carlo).

It would be interesting to see the expected profit (and also the sd of profit) using your strategy, which you could also compute using Monte Carlo.

In my opinion, what is really needed is a backtesting model using actual historical data for both futures and options. For reasons that I won't go into here, I don't think a Monte Carlo simulation would be appropriate.