Watch HowardCohodas Trade Index Options Credit Spreads

[HowardCohodas]

I have a question Howard.
Your income is steady at 7-11%, and given the nature of your trading, this is to be expected. You have however had a major loss (>30%).

I am intrigued as to how the month ended with similar results to the others with such a loss. Did you up risk, trade more, or something else?

The one credit spread that had a 30.3% loss happened in Nov. That month my account increased 11.2%. This trade was in pre-production (small money) mode. I was none-the-less disappointed because it was the result of a mistake at entry that I let ride and came back to bite me.

Check out the risk analysis post linked in my signature to see how the serious loss of one spread might impact my account.

 

[MrGecko]

OK - I'm interested. Can you explain? I know nothing of these things...

I've already asked this question on a thread before, here is the original thread that explains the answer.

Disclaimer: This example is taken from Paul Wilmott Introduces Quantitative Finance. Before it was explained to me, I thought the answer was $0.6 just like everyone else. "It's easy when you know how", as they say.

 

[HowardCohodas]

Through some thoughtful PM discussions and the thoughtful discussions in this thread, perhaps I have developed additional insights. Let's see if I can explain it.

First the easy part...
The mechanism that generates the return is the closing of the price gap between the legs of the credit spread. This gap will go to zero at expiration.

If the underlying price nears or overtakes the short strike, the gap will widen and I will suffer a loss. I do not wait and hope for a rebound. If my loss on capital at risk exceeds 20% I bail out at the market. I am at the mercy of the market and may not get good prices, so a larger loss than 20% is likely. Since I only trade options on indexes with large volume, I am unlikely to encounter a problem with getting out of my trade due to no takers. Although I trade the spread as a unit, it is unlikely that both legs came from the same entity, so no one is necessarily "taking the other side" of the spread on entry or exit.

Now for the hard part...
All the magic exists in entry selection and to a lesser extent management of the spread.

The selection has two criteria; probability of touching and credit received. Neither of these separately or in combination necessarily produce strikes distanced symmetrically from the underlying instrument price. The implications of this are profound. First, it implies that there is a built-in bias on direction. Not my bias, but a bias built in by traders in pricing these options.

Second, because of this bias, one side of an Iron Condor will be more appealing than the other in terms of meeting my criteria. In many cases, one side will be acceptable and the other will not. Thus, seldom am I able to put on the spread to complete the Iron Condor immediately. It may take, hours, days or even weeks before market conditions make it fit my criteria.

In spite of this, I am able to complete the Iron Condor in more than 90% of the spreads. The weeklies are the toughest because there may be insufficient time for the market to change enough to get both in. In fact, this week I have been unable to establish even one spread on the current NDX weekly that matches my criteria.

I think there is less magic in the way I manage the spreads. I do not manage them as Iron Condors, only as spreads. I roll when opportunities develop and I bail when jeopardy encroaches.

What about days like the flash crash? I'm going to be hit, and hit hard. The contingency orders may not limit my loss to 20% to 30%. Worse yet, all expirations will be affected, not just the ones nearest to the expiration date.

However, even a flash crash is mitigated by the construction of the account. Only half of the spreads will be effected. The ones on the opposite side of the underlying instrument price will hit their near maximum profit. So I'm very unlikely to be wiped out unless trading stops permanently. I'm wounded, but functional.

 

[oiltanker]

as someone who doesn't know a gamma from a delta i've learnt a lot from this thread.there is not enough on spreads imo.

some thoughts it raised in me were

Do all edges fail over time? It took woolworths 100 years to go bust. In the end they lost their edge. Who would have thought microsoft might lose their edge? But google is moving in with a better edge? So edges are time limited?

is money management an edge? If one only shorts ftse at a 'likely' looking spot and then cuts the loser quickly and lets the winner run then success depends upon what one one thinks is a 'likely looking spot'? which involves some aspect of discretion gained from long experience?

will all edges run into a black swan at some point that 'disproves' the edge? maybe one is lucky and avoids them but then that is luck not an edge?

is current price the result of collective wisdom or collective ignorance? what would one have to look at to judge that? Soros keeps saying markets are wrongly priced due to innate bias in the market. So one can trade the bias in expectation of a profit? Indeed if there was no mispricing who would ever make profit? So the current price is often a misprice?

any true edge is worth goldmine so no one who understands that would put it on a board [see all the current court cases of people stealing code] so any system posted the edge [if there is one] is between the lines. ie the x factor that makes it work. As demonstrated in the turtles. people given exactly the same mechanical rules [no discretion necessary] and capital and the majority failed.
so the edge in any system is not in the rules but its application? Can one 'teach' that?

 

[MrGecko]

The selection has two criteria; probability of touching and credit received.

Howard, can you explain what you look for in:

a) Probability of touching
b) Credit Recieved
c) the ratio of these two

 

[HowardCohodas]

Howard, can you explain what you look for in:

a) Probability of touching
b) Credit Recieved
c) the ratio of these two

For a and b, you should read some of the links in my signature. They all point to posts in this thread that discuss your questions and many others.

Regarding c, their ratio has no meaning in my methods.

If you still have questions, please ask again.

 

[DionysusToast]

Regarding c, their ratio has no meaning in my methods.

 

[MrGecko]

For a and b, you should read some of the links in my signature. They all point to posts in this thread that discuss your questions and many others.

Regarding c, their ratio has no meaning in my methods.

If you still have questions, please ask again.

I have read all of your posts Howard. All I can see is a reference to a less than 10% probability of touching, with no mention of what credit you are looking for.

 

[robster970]

I have read all of your posts Howard. All I can see is a reference to a less than 10% probability of touching, with no mention of what credit you are looking for.

If I understand it correctly Gecko, Howard just uses a threshold on probability of touching as his only constraint which then drives a net profit figure based upon the option prices. From about 10 pages ago, I thought his PoT threshold was 15%.

http://www.trade2win.com/boards/tra...ex-options-credit-spreads-39.html#post1357424

 

[HowardCohodas]

I have read all of your posts Howard. All I can see is a reference to a less than 10% probability of touching, with no mention of what credit you are looking for.

My bad. I have written about it in several venues and assumed without first checking that I had written about it here as well.

In back testing sensitivity analysis I found I was on a relatively flat plain with respect to credit received. It is largely a matter of available spreads within the probability of touching limits. I personally aim for a net of 10% for both sides of the Iron Condor which is approximately 5% for each side. By observation, that usually provides me with at least one side, although the current weekly NDX is an example where it has not.

Depending on market dynamics, the 10% goal can sometimes be met with just one side. That does not materially change my approach to the credit required for the second spread. My minimum is 3% because even though no additional quarantined funds are required, I am assuming some additional risk.

 

[MrGecko]

I personally aim for a net of 10% for both sides of the Iron Condor which is approximately 5% for each side... My minimum is 3% because even though no additional quarantined funds are required, I am assuming some additional risk.

So this is 10% target (5% each side, minimum 3%) is (Credit recieved / Maximum loss) * 100?

 

[HowardCohodas]

If I understand it correctly Gecko, Howard just uses a threshold on probability of touching as his only constraint which then drives a net profit figure based upon the option prices. From about 10 pages ago, I thought his PoT threshold was 15%.

http://www.trade2win.com/boards/tra...ex-options-credit-spreads-39.html#post1357424

As PoT is largely associated with risk tolerance, I started out at 10% and inched my way up until it increased stomach acid production and then I backed off a notch. I'm currently between 15% and 20%. It remains to be seen if my comfort level changes over time.

Post #172

 

[HowardCohodas]

So this is 10% target (5% each side, minimum 3%) is (Credit recieved / Maximum loss) * 100?

Not sure of the question.

The poor risk/reward ratio is compensated for by the probability of closing the trade profitably and limiting the loss to less than the maximum when not profitable.

Post #172

 

[MrGecko]

The highlighted portion illustrates the 1875/1850 spread where the short option has a probability of touching of 15% for a credit of $1.75 for a potential profit of 7.5% at expiration. I will treat the probability of touching as a proxy for the chances that the I will not let the option expire at zero and earn maximum return. Then I compute the mathematical expectation of this position as 85% chance of earning 7.5% return and a 15% chance of losing 15%. This yields a mathematical expectation of 4% return.

I would look into this probability of touching thins Howard, and to be honest I would either ditch it or publish how you calculated it yourself. It is misleading, and you probably have some circular reference in your calculations of expectancy.

 

[HowardCohodas]

I would look into this probability of touching thins Howard, and to be honest I would either ditch it or publish how you calculated it yourself. It is misleading, and you probably have some circular reference in your calculations of expectancy.

Why should I ditch a calculation that has proved quite accurate in 53 closed spreads. 48 profitable, 3 small losses, 1 limit loss, 0 max loss. Do you consider this statistically significant?

I though I had described my two approaches to calculating the "probability of reaching" in this thread.

Show me where I have a circular reference in my analysis. Risk Analysis I admittedly wrote it in somewhat of a hurry, so I may have overlooked some significant flaw in my analysis.

 

[MrGecko]

Howard, do me a favour and tell me the details of a particular spread that you are looking at trading right now - it only has to be close to being a go'er.

 

[robster970]

I think that would be incredibly useful - to go through step by step, the selection, creation, management and eventual close of a single IC. The whole life-cycle.

So I think people get that your risk management is based entirely around probability of touching. What's less clear is the selection of the spread itself. In the example in post 172 did you purely select this spread because it's PoT suited your risk and no other reason?

 

[HowardCohodas]

Howard, do me a favour and tell me the details of a particular spread that you are looking at trading right now - it only has to be close to being a go'er.

Great idea. I'm off to a lunch appointment, but shall address it on my return.

 

[MrGecko]

Why should I ditch a calculation that has proved quite accurate in 53 closed spreads. 48 profitable, 3 small losses, 1 limit loss, 0 max loss. Do you consider this statistically significant?

I though I had described my two approaches to calculating the "probability of reaching" in this thread.

Show me where I have a circular reference in my analysis. Risk Analysis I admittedly wrote it in somewhat of a hurry, so I may have overlooked some significant flaw in my analysis.

Howard, this kind of thinking is going to bite you in the ar$e.

Having gone over the ThinkOrSwim pdf about "Probability of Touching", I have learned that it is basically a confidence interval based on a normal distribution of stock returns using the front month Implied Volatility**. The Implied Volatility is derived from the last price of the option. There is where your circular reference is, you are looking at two functions of the same thing. In which case, it doesn't matter a scooby-doo which options you choose in your strategy, because the probability of touching and the option price will move in lock-step with one another. The laws of mathematics and arbitrage see to this.

So, what your strategy is saying is:

I'll look at spreads where, by virtue of the strike price, expiry and the Implied Volatility, the Probability of Touching is < 10% and the net Credit is > 5% (These figures are arbitrary, and it follows that the options you are spreading are also arbitrary). Then you are saying "well I want to sell time and collect theta, so I will sell the spread for a credit" without realising you are also short Volatility.

This makes sense if you think that the Implied Volatility used to derive the Probability of Touching and the option prices themselves is too high - if you think volatility is too high, then sell it. But you are selling Volatility systematically, without comparing Implied Volatility with expected realised Volatility.

If the Probability of Touching figure was derived using some volatility model, then you could be on to something - you would compare the probability of Touching produced by your model to the Probability of Touching implied by the option price, and take a position accordingly.

But you aren't doing that.

** Implied Volatility, Howard, is pretty much what everybody understands. Probability of Touching is a homogenous*** function of Implied Volatility and hardly anybody has even heard of it.

*** yet to prove this, but it's an educated guess.

 

[HowardCohodas]

28 DEC 2010 Trading Summary

28 DEC 2010 Trading Summary

Neither opportunity in today's trading plan panned out.

Iron Condor 26 was incomplete, lacking a call spread. At probability of touching of 15% to 20% (my current range) there were strike prices of 1380, 1385, and 1390. At a strike difference of $5.00 and seeking at least 5% credit, I'm looking for at least $.25. This was not available and there was no volume, so no trade.

Spread 56 - roll opportunity
91% of the potential profit had been reached so this was a roll opportunity. Sufficient credit was not available although volume was adequate.

Spread 36 - roll opportunity
The opportunity disappeared almost immediately after the market opened.

I increased my position on both sides of IC 22. The credits were slightly different than the original spread, so the average credit was adjusted to reflect that and make the derived results consistent.