CMLviz Trade Machine

[cuegis 683]

That is why I began by pointing out that I wasn't sure what he was trying to say!

I agree with you, and I don't buy into any of these catastrophy theories.

I was only referring to a few things that have already happened. The things that I actually saw myself. (except 1929-1935)

The things I pointed out really happened and it is only a matter of what semantics you want to use to describe it.

The stock market losing more than 50% of it's value, IS a collapse, no matter what words you use to define it.

I never bought into 1987 as being a real "crash". It was a 1 day event followed immediately by multi-decade bull market.

There was no damage on "deeper" levels.

1929 and 2008/2009 were both different animals (than 1987) and were anything but a 1 day event.

Everything was deeply damaged, on almost every level for a very extended period of time and almost everyone was affected. Just because the stock market is higher 9 years later does not minimize, or negate what happened.

The past never looks as enormous as it did when the event happened. But some doomsayers would have you believe that the future has "real" catastrophy in store.

It's only been 8 years, but just enough time for everyone to have forgotten about it enough to minimize the depth of what really happened.

Sadly, there will come a time in the future, when even larger things happen. But I doubt they will be the result, or play out, the way the doomsayers try to frieghten us into believing.

[Ophir Gottlieb 62]

1 hour ago, cuegis said:

That is why I began by pointing out that I wasn't sure what he was trying to say!

I agree with you, and I don't buy into any of these catastrophy theories.

I was only referring to a few things that have already happened. The things that I actually saw myself. (except 1929-1935)

The things I pointed out really happened and it is only a matter of what semantics you want to use to describe it.

The stock market losing more than 50% of it's value, IS a collapse, no matter what words you use to define it.

I never bought into 1987 as being a real "crash". It was a 1 day event followed immediately by multi-decade bull market.

There was no damage on "deeper" levels.

1929 and 2008/2009 were both different animals (than 1987) and were anything but a 1 day event.

Everything was deeply damaged, on almost every level for a very extended period of time and almost everyone was affected. Just because the stock market is higher 9 years later does not minimize, or negate what happened.

The past never looks as enormous as it did when the event happened. But some doomsayers would have you believe that the future has "real" catastrophy in store.

It's only been 8 years, but just enough time for everyone to have forgotten about it enough to minimize the depth of what really happened.

Sadly, there will come a time in the future, when even larger things happen. But I doubt they will be the result, or play out, the way the doomsayers try to frieghten us into believing.

Generally agree with you.

[Noah Katz 90]

2 hours ago, cuegis said:

I'm not sure what you are saying but late 70's, culminating in 1980 we had hyper-inflation. Prime was 12.5%, inflation was 20%+.

 

I'm really not trying to be snarky, but was just pointing out that we haven't seen it all in this country in our lifetimes.

 

 

Also,in my mind hyperinflation is 100's or 1000's of %, not tens.

 

 

That said, the chances of anything like what happened in the Weimar Republic or various 3rd-world countries is remote.

 

 

Still, Ophir (whom I thank profusely for the very illuminating analyses) didn't address *sovereign* debt which some doomsayers point to.

 

Perhaps someone can disabuse me of the notion that we and other large economies are on the same path as Greece.

 

In the past we grew out of debt, but how can that happen with the declining (or headed that way) populations of the world's largest economies, who continue to amass even more debt?

 

Oh, one more thing, re margin debt:

 

I'm not grasping the relevance of margin debt as a % of a particular index; seems to me what's important is how many available $ (margined or otherwise) are left for people to bet on the bull.

Edited April 17, 2017 by Noah Katz

[4tach 15]

19 hours ago, Ophir Gottlieb said:

Also, an interesting set-up I see with Tesla

Tesla (TSLA): There's an opportunity, now, with this volatility

 

@Ophir Gottlieb

In the above quote there is mention of a TSLA link that outlines a possible strategy. Can we set this up? I am a newbie so here is my pitiful rendition of such: 
[BUY 40 DELTA CALL & PUT]
[SELL 10 DELTA CALL & PUT]
which I translate to mean (in today's environment with TSLA @ 300)
TSLA MAY 19 2017  280/295 355/370 IRON CONDOR
In other words, it seems to me that the CML TradeMachine has great value in winnowing down and outlining a strategy and then the trader community (SO) has to further refine and discuss the details to see if it (the stragegy) has good prospects.
Is this the line of thinking that you think?

[Ophir Gottlieb 62]

18 minutes ago, 4tach said:

@Ophir Gottlieb

In the above quote there is mention of a TSLA link that outlines a possible strategy. Can we set this up? I am a newbie so here is my pitiful rendition of such: 
[BUY 40 DELTA CALL & PUT]
[SELL 10 DELTA CALL & PUT]
which I translate to mean (in today's environment with TSLA @ 300)
TSLA MAY 19 2017  280/295 355/370 IRON CONDOR
In other words, it seems to me that the CML TradeMachine has great value in winnowing down and outlining a strategy and then the trader community (SO) has to further refine and discuss the details to see if it (the stragegy) has good prospects.
Is this the line of thinking that you think?

May 19 Expiry

40 Delta

Call: 310

Put: 295

10 Delta:

Call: 355

Put: 250

 

You can always find the details of a trade through history by clicking on a back test tile (it will open up the full trade list).

 

I can't say really if you want to discuss a trade with the community. I think it could be useful, but I know many ppl that see a trade, test it, and roll with it. I think either way is fine. Find your comfort zone, define your strategy, become an expert at it, and you should do just fine. Feedback from trusted sources is very nice to have.

Edited April 18, 2017 by Ophir Gottlieb

[Ophir Gottlieb 62]

On 4/17/2017 at 9:26 AM, Gary said:

@Ophir Gottlieb great product ... thanks for making it available to us at a discount

Question - have you considered adding "max drawdown" to the summary "tiles" when running backtests?

Really looking forward to some of the new features like the excel download. Thanks again!

Amount risked is actually the max draw down accumulating through time or the hypothetical max loss on the trade (whichever is larger). A max draw down data point though is very clever. We will add to the list. Thank you!

[NikTam 163]

So would it be a reasonable strategy to pick some favorite stocks based on liquidity, volume, overall trend, then find the most profitable option strategy based on back-test results offered through CML, and then place many small trades?  

[Ophir Gottlieb 62]

2 minutes ago, NikTam said:

So would it be a reasonable strategy to pick some favorite stocks based on liquidity, volume, overall trend, then find the most profitable option strategy based on back-test results offered through CML, and then place many small trades?  

 

So, I can't give advice, but this is how I feel:

 

Vol and stock dynamics tend to be sticky (that's why, for example, people use technical analysis). 

 

It's these vol and stock dynamics that the Trade Machine (back-tester) finds for us and are basically impossible to find without a tool like this.

 

But, there must be two underlying beliefs:

(1) The stock direction will be, at least, semi-similar to the back-test period (this is why I do not believe more data is better for an option back-tester. I usually look at the last two-years or one-year (or even six-months), and not further, even though the back-tester has more history. This is broadly called stock dynamics.

(2) The vol dynamics will be at least semi-similar. In the short-run, this is "usually" true. See how VIX or VXX move essentially in long trends (for VXX it has been an 8-year trend). The vol dynamics are usually the piece we can count on unless we know of some huge event -- like TSLA releasing the Model 3.

 

The companies that I love to do this with are large ones.  So, AAPL, GOOGL, BAC, JPM, IBM, and several others.

Here are a few examples:

Trading options in the banks in the era of Donald Trump

There is Edge in Bank of America Options

Apple: You can be a better option trader

Mastering IBM Vol Dynamics to Profit from Options and Avoid Earnings

[sokl 1]

in this article, "apple: you can be a better player" you say, Said simply, if we ever face a put spread that is down 100%, we close it right then, and wait for the next week to start a new position. 

can you explain to me when "put spread that is down 100%" happens? 

thanks,

 

[cuegis 683]

VXX is not a good choice for looking back at history. Unless you are looking at a constant.

Because , as markets go through different phases, VXX and VIX will alternate from "Backwardation" to "Contango" and you won't be seeing the full picture.

Markets change, and morph into new and different things over time.

[Ophir Gottlieb 62]

8 minutes ago, sokl said:

in this article, "apple: you can be a better player" you say, Said simply, if we ever face a put spread that is down 100%, we close it right then, and wait for the next week to start a new position. 

can you explain to me when "put spread that is down 100%" happens? 

thanks,

 

Sure, if you a sell a put spread @ $2 (for example), if it goes to $4, you have a 100% loss. The back-test then buys the spread back for $4, takes the $2 loss (100%), and waits for the next period to put the spread on again. This prevents the spread from turning into a massive loser, like 200%, 300%, even 400%. It's a double risk adjustment:

 

1. use a put spread rather than a naked put

2. even with the spread,use a stop loss.

 

This turned a 10% losing strategy over two-years, into a 68% winning strategy. It uses the back-tester to examine the vol and stock dynamics of Apple to create a trading plan before placing the trade. 

 

This is why the back-tester exists. Plan. Execute. Never (ever) guess.

Edited April 18, 2017 by Ophir Gottlieb

[Ophir Gottlieb 62]

4 minutes ago, cuegis said:

VXX is not a good choice for looking back at history. Unless you are looking at a constant.

Because , as markets go through different phases, VXX and VIX will alternate from "Backwardation" to "Contango" and you won't be seeing the full picture.

Markets change, and morph into new and different things over time.

 

VXX is a great choice as it uncovers the vol dynamics of VIX. 

 

Here is the irrefutable evidence:

 

The Marvelous Implications of a Stop Loss and VIX Dynamics - The Trade That has Won for 8 Straight Years

 

 

The Marvelous Implications of a Stop Loss on iPath S&P 500 VIX Short Term Futures TM ETN (NYSEARCA:VXX)

Date Published 2-15-2017 
The Trade That Keeps Working 
The iPath S&P 500 VIX Short Term Futures TM ETN (NYSEARCA:VXX) is easily one of the most interesting instruments I have ever encountered as an option market maker physically on the NYSE ARCA floor and CBOE, remotely. It's a bet on contango (or backwardation depending on your position) in the VIX and it has been just unimaginably consistent. 

But, while the best keeps on working, the nature of the brief periods where it reverses means large losses can be sustained in a short-period, unless you're prepared. 
 

Understanding the risk profile of VXX has meant massively higher returns.

For the most part, the VXX just goes down all the time... Here is an all-time chart for VXX: 
 

But here is that chart over the last two-years: 
 

For obvious reasons, it has become one of the most crowded trades in the market -- short and short and short again. But, we can also examine this instrument with options. 

While buying puts seems like the obvious move, there is a slightly less risky approach -- selling call spreads. With a spread, we hedge one option with another as opposed to a naked long (or short) position. 

Here is what selling out of the money call spreads has done over the last 2-years, rolling every week (so trading weekly options) 
 

 
(Source: CMLviz Trade Machine)

That's a 73% return with 83 winning trades and just 21 losing trades for a 79.8% win rate. Since delta is actually a proxy for probability, owning a 40 delta option should be in the money about 40% of time. And that means that selling it should work about 60% of the time. 

What we have discovered here is 79.8% win rate, but the times when the VXX rises are abrupt and can cause portfolio draw downs, even if over the long-term the short call spreads were winners. 

It turns that as traders we have a tool for this behavior, and it's called a stop loss. You see, a short call spread can only make 100% in a single trade, but it can lose far more than that. It does not have a symmetric profit loss profile. 

But, if we cut those losses off at the knees and totally reconstruct our profit loss profile, we can actually erase a huge amount of risk and increase returns. Here's what happens when we put a stop loss at 50% for the call spreads every week. 

Here's how we do it: 
 

And now the results: 
 

 
(Source: CMLviz Trade Machine)

In English, if in any given week a call spread turns against us for a 50% loss, we buy it back and wait for the next week -- not allowing that trade to get worse. 

We see a 73% winner turn into a 126% winner and we took massively less risk by using a stop loss. While the win-rate has gone down, the returns have jumped higher. 

In in the case of the VXX and its rare but abrupt sporadic up turns, this is a highly sophisticated, albeit very easy approach, way of dealing with this phenomenon. 

THE KEY 
The key here is to find edge, optimize it -- in this case by putting in a tight stop -- and then to see if it's been sustained through time. For VXX it has, and that makes for a powerful result. 

We've just seen an explicit demonstration of the fact that there's a lot less 'luck' and a lot more planning in successful option trading than many people realize. Here is a quick 4-minute demonstration video that will change your option trading life forever: Tap here to see the Trade Machine in action

 

 

 

 

 

 

Edited April 18, 2017 by Ophir Gottlieb

[cuegis 683]

Stop losses are a funny thing. They can only be used if the market is open. So there is too much uncertainty involved (i.e. the market has a large gap from close to open)

Using option as a hedge against other options is a permanent stop loss because there cannot be any market gaps.

[cuegis 683]

1 minute ago, Ophir Gottlieb said:

 

VXX is a great choice as it uncovers the vol dynamics of VIX. 

 

Here is the irrefutable evidence:

 

The Marvelous Implications of a Stop Loss and VIX Dynamics - The Trade That has Won for 8 Straight Years

 

I hope you didn't choose the past 8 years, to the exclusion of any other 8 year period. Now that wouldn't be fair, would it?

[Ophir Gottlieb 62]

Just now, cuegis said:

Stop losses are a funny thing. They can only be used if the market is open. So there is too much uncertainty involved (i.e. the market has a large gap from close to open)

Using option as a hedge against other options is a permanent stop loss because there cannot be any market gaps.

 

But adding a stop limit is yet another layer of risk protection, and the back-tester does account for gaps. Try to think of trading as a continuum of knowledge growth.  We are all somewhere that is not a the 100% mark -- a growth mindset will allow us to see opportunities that exist, even if we did not think they existed before. 

[Ophir Gottlieb 62]

1 minute ago, cuegis said:

I hope you didn't choose the past 8 years, to the exclusion of any other 8 year period. Now that wouldn't be fair, would it?

 

VXX has only existed for 8 years and for each of those years it has worked.. Your tone is a little negative, which is fine by me, but it will eliminate opportunities for your trading. A closed mindset is disastrous for traders.

[Yowster 9,180]

10 minutes ago, Ophir Gottlieb said:

Sure, if you a sell a put spread @ $2 (for example), if it goes to $4, you have a 100% loss. The back-test then buys the spread back for $4, takes the $2 loss (100%), and waits for the next period to put the spread on again. This prevents the spread from turning into a massive loser, like 200%, 300%, even 400%. It's a double risk adjustment:

 

1. use a put spread rather than a naked put

2. even with the spread,use a stop loss.

 

This turned a 10% losing strategy over two-years, into a 68% winning strategy. It uses the back-tester to examine the vol and stock dynamics of Apple to create a trading plan before placing the trade. 

 

This is why the back-tester exists. Plan. Execute. Never (ever) guess.

@Ophir Gottlieb- I get what you are saying, but the terminology will confuse SO members.  When dealing with credit spreads, Kim (correctly IMO) bases his gain/loss percentage on the margin required and not the credit received.   For example if you sell a spread of width 10 for 2.00, then your margin requirement is $800 per spread and that is what gains and losses should be based on.   For what you are talking about, you mean get out of the credit spread when its value is 2x the credit you received for opening the trade.  

It's just a terminology thing - but it does bring an important question to mind.  When you are using the tool for selling a spread - is the gain/loss percentage calculated by the tool based on the up-front credit received???   Or on the margin requirement (width of spread minus credit received)???

[cuegis 683]

1 minute ago, Ophir Gottlieb said:

 

VXX has only existed for 8 years and for each of those years it has worked.. Your tone is a little negative, which is fine by me, but it will eliminate opportunities for your trading. A closed mindset is disastrous for traders.

It's not negative. I'm just pointing out that you would be better off looking at the spot VIX because it always measures "today".

The VXX and a variety of other products have downward and upward drifts to them, depending on different phases.

You HAVE TO have an open mindset because everything is always changing, and, the more participants plus the growth of technology and knowledge are always going to cause change to happen at a faster rate.

The last thing in the world I have is closed mindset.

I have seen SO many things in my lifetime that have behaved a certain way for a period of time. As they go from being unknown to known by everybody, they cease to exist.

There was no "Skew" in options prior to the 1987 crash. The entire way options are priced changed after that.

Everything is always changing, faster and faster. There are very few, if any, absolutes.

[Ophir Gottlieb 62]

2 minutes ago, Yowster said:

@Ophir Gottlieb- I get what you are saying, but the terminology will confuse SO members.  When dealing with credit spreads, Kim (correctly IMO) bases his gain/loss percentage on the margin required and not the credit received.   For example if you sell a spread of width 10 for 2.00, then your margin requirement is $800 per spread and that is what gains and losses should be based on.   For what you are talking about, you mean get out of the credit spread when its value is 2x the credit you received for opening the trade.  

It's just a terminology thing - but it does bring an important question to mind.  When you are using the tool for selling a spread - is the gain/loss percentage calculated by the tool based on the up-front credit received???   Or on the margin requirement (width of spread minus credit received)???

 

We use the rules that govern Reg-T for "amount risked." 

 

Here you go:

 

CALCULATING 'AMOUNT RISKED' FOR OPTION STRATEGIES 
If you are ever in doubt about the margin requirements, The Options Clearing Corporation publishes a very handy options margin calculator here. 
Options Margin Calculator 

In an effort to help save time, here are the results: 

Naked Short Put 
Initial margin requirement: 
* 100% of option proceeds, plus 20% of underlying security value less out-of-the-money amount, if any 
* minimum requirement is option proceeds plus 10% of the put’s aggregate exercise price (number of contracts x exercise price x $100) 
* proceeds received from sale of puts(s) may be applied to the initial margin requirement after position is established, ongoing maintenance margin requirement applies, and an increase (or decrease) in the margin required is possible 

What Does the CMLviz Back-tester Use? 
We use the most conservative value: 
100% of option proceeds, plus 20% of underlying security value less out-of-the-money amount, if any 

Here's an Example 
Position 
Short 1 Mar 100 puts(s) at $5.00 
Underlying stock at $100.00 
Put is at-the-money 

Initial Margin 
Margin Requirement (Amount Risked): $2,500.00 
Proceeds from sale of short put(s): $500.00 
Margin call (SMA debit): $2,000.00 

Naked Short Call 
Initial margin requirement: 
* 100% of option proceeds, plus 20% of underlying security value less out-of-the-money amount, if any 
* minimum requirement is option proceeds plus 10% of the underlying security value 
* proceeds received from sale of call(s) may be applied to the initial margin requirement after position is established, ongoing maintenance margin requirement applies, and an increase (or decrease) in the margin required is possible 

What Does the CMLviz Back-tester Use? 
Again, we use the most conservative value: 
100% of option proceeds, plus 20% of underlying security value less out-of-the-money amount, if any 

Here's an Example 
Position Short 1 Mar 100 call(s) at $5.00 
Underlying stock at $100.00 
Call is at-the-money 

Initial Margin 
Margin Requirement (Amount Risked): $2,500.00 
Proceeds from sale of short call(s): $500.00 
Margin call (SMA debit): $2,000.00 

Selling Spreads 
We take the minimum of either the naked short option risked or the max loss of the spread. 

Buying Options or Option Spreads 
The amount risked is the cost of the option or the spread. 

Can you explain more about short straddle and short strangle risk? 
In the most basic sense, the straddle and strangle risk calc is simply taking the put risk and adding it to the call risk. Where it gets more invovled, is that what the TradeMachine ultimately displays is the "max risked". This figure tracks how risk changes over the course of the entire backtest, and it takes into account the running cash balance, commissions, etc. 

If, for example, a certain out of the money (say 25 delta) short call were being sold over and over, and the stock never moved, then the cash balance of the backtest would be slowly building. So, if this option required $1000 of margin (aka $1000 of risk) the first time, and the system collected $100 for selling it, for a net risk of $900, each subsequent rollover the risk would go down by $100 as the trader's net cash balance builds. 

When there is both a short call and a short put, in most circumstances (other than trading "guts" in the money strangles) at least one of these options expires worthless and has a positive effect on the net cash. This, in turn, affects the risk, by increasing the cash, and reducing the risk from this point forward. 

The reverse is also true. If the trader loses $5000 on the first trade, the risk for the overall backtest is now $5000 higher, and so the max risked calculation at the end will reflect this much higher risk. So, even if the end result is very positive, the risk along the way may have been substantial. 

[Ophir Gottlieb 62]

1 minute ago, cuegis said:

It's not negative. I'm just pointing out that you would be better off looking at the spot VIX because it always measures "today".

The VXX and a variety of other products have downward and upward drifts to them, depending on different phases.

You HAVE TO have an open mindset because everything is always changing, and, the more participants plus the growth of technology and knowledge are always going to cause change to happen at a faster rate.

The last thing in the world I have is closed mindset.

I have seen SO many things in my lifetime that have behaved a certain way for a period of time. As they go from being unknown to known by everybody, they cease to exist.

There was no "Skew" in options prior to the 1987 crash. The entire way options are priced changed after that.

Everything is always changing, faster and faster. There are very few, if any, absolutes.

 

Agree.

[Yowster 9,180]

@Ophir Gottlieb- Thank you for the explanation.  So I assume the gain/loss% is based on the "amount risked" ??  

 

However, I did just notice something to the contrary.  I used your VXX example a few posts back of selling 40/15 delta call credit spread with a rollover of 7 days and a close trade when losses above 50%.   I ran it through the tool and drilled into the trade details, for the most recent "close of losses" occurrence I observed the following:

• Trade open on 4/7 - sell VXX 17/18.5 call credit spread for 0.19 (so margin is 1.50 width - 0.19 = 1.31).
• Trade closed due to losses on 4/10 for 0.34, for a loss of 0.15 (0.34 - 0.19).   While this 0.34 is close to 2x the opening credit received, it is nowhere close to a 50% loss based on the margin requirement of the credit spread (which would be around a 0.65 loss at a closing price of around 0.84).

So, it would appear that "Close Trade When" loss percentage may not be what the user thinks it is when dealing with credit spreads.

[Ophir Gottlieb 62]

26 minutes ago, Yowster said:

@Ophir Gottlieb- Thank you for the explanation.  So I assume the gain/loss% is based on the "amount risked" ??  

 

However, I did just notice something to the contrary.  I used your VXX example a few posts back of selling 40/15 delta call credit spread with a rollover of 7 days and a close trade when losses above 50%.   I ran it through the tool and drilled into the trade details, for the most recent "close of losses" occurrence I observed the following:

• Trade open on 4/7 - sell VXX 17/18.5 call credit spread for 0.19 (so margin is 1.50 width - 0.19 = 1.31).
• Trade closed due to losses on 4/10 for 0.34, for a loss of 0.15 (0.34 - 0.19).   While this 0.34 is close to 2x the opening credit received, it is nowhere close to a 50% loss based on the margin requirement of the credit spread (which would be around a 0.65 loss at a closing price of around 0.84).

So, it would appear that "Close Trade When" loss percentage may not be what the user thinks it is when dealing with credit spreads.

 

That's right. Amount lost is based on the credit, amount risked is based on Reg T. 

 

Since we use closing prices (for now), the stop was triggered once the 50% loss was hit, and the closing price was worse than the 50% loss -- thus the print.

 

Measuring amount lost based on the amount of the credit and the amount risked based on Reg T is generally considered the most accurate, reliable and representative form of measuring option spread PnL.

 

As an example:

Set up

Sell a credit spread @ $1

Width: $5

 

How CML (and most of trading world, including broker-dealers and CBOE) does it:

Amount risked: $5 (width)  - $1 (credit) = $4

Max Loss in %: (Amount risked)/(Credit)  = (4)/(1) = 400%

 

Hopefully that makes sense. This is also how margin is calculated per Reg T.

[cuegis 683]

Are you going to add some kind of filter to set limits on the amount of credit you would/would not be willing to accept based os a % of the distance between strikes?

Obviously you are never going to sell a 1.50 spread for .19 cents. I would hope!

 

There should be a way of filtering the minimum credit you will accept as a % of the spread width. If you run a backtest that has many trades included in the results that are like this, then it is not a fair and reliable backtest....because you would never put on a trade like this. Or, to take it to it's extreme, you would never sell a $1.50 spread for .01 cent........and then close it out for a 100% loss of .02 cents?

Edited April 18, 2017 by cuegis

[Yowster 9,180]

2 minutes ago, Ophir Gottlieb said:

 

That's right. Amount lost is based on the credit, amount risked is based on Reg T. 

 

Since we use closing prices (for now), the stop was triggered once the 50% loss was hit, and the closing price was worse than the 50% loss -- thus the print.

 

Measuring amount lost based on the amount of the credit and the amount risked based on Reg T is generally considered the most accurate, reliable and representative form of measuring option spread PnL.

 

As an example:

Set up

Sell a credit spread @ $1

Width: $5

 

How CML (and most of trading world, including broker-dealers and CBOE) does it:

Amount risked: $5 (width)  - $1 (credit) = $4

Max Loss in %: (Amount risked)/(Credit)  = (4)/(1) = 400%

 

Hopefully that makes sense. This is also how margin is calculated per Reg T.

I see our discrepancy.  I am in total agreement with the margin calculation, but I think we have always used a max loss of 100% for credit or debit spreads - I would always consider max loss to be 100% and that value for a credit spread is the margin requirement (which is equal to width - credit received). @Kim- please chime in with how you view this calculation.

So for your example of selling a $5 width spread for $1.  I have always considered the max gain percentage to be the credit received ($1) divided by the margin requirement ($4) = 25%, and the max loss to be 100% if the entire margin requirement of $4 was lost.    This terminology difference is good to know, as I think many of the SO members view the gain/loss percentages the same way that I do - which means that the "close trade when" percentage corresponds to a dollar figure which is much lower than what people may think it is.

[Ophir Gottlieb 62]

1 minute ago, cuegis said:

Are you going to add some kind of filter to set limits on the amount of credit you would/would not be willing to accept based os a % of the distance between strikes?

Obviously you are never going to sell a 1.50 spread for .19 cents. I would hope!

 

There should be a way of filtering the minimum credit you will accept as a % of the spread width. If you run a backtest that has many trades included in the results that are like this, then it is not a fair and reliable backtest....because you would never put on a trade like this. Or, to take it to it's extreme, you would never sell a $1.50 spread for .01 cent........and then close it out for a 100% loss of .02 cents?

 

We will be allowing for dollar denominated spreads by strike, rather than deltas, and that should make it much clearer to people that prefer to think of spreads in dollar widths to see what is going on, and which spreads to back-test. That fix is coming within a couple of weeks.

[Kim 7,943]

24 minutes ago, Yowster said:

I see our discrepancy.  I am in total agreement with the margin calculation, but I think we have always used a max loss of 100% for credit or debit spreads - I would always consider max loss to be 100% and that value for a credit spread is the margin requirement (which is equal to width - credit received). @Kim- please chime in with how you view this calculation.

So for your example of selling a $5 width spread for $1.  I have always considered the max gain percentage to be the credit received ($1) divided by the margin requirement ($4) = 25%, and the max loss to be 100% if the entire margin requirement of $4 was lost.    This terminology difference is good to know, as I think many of the SO members view the gain/loss percentages the same way that I do - which means that the "close trade when" percentage corresponds to a dollar figure which is much lower than what people may think it is.

Yes, 100% loss is based on margin, not credit received. This is something that is worth clarification in the software otherwise it could be confusing.

[Ophir Gottlieb 62]

50 minutes ago, Kim said:

Yes, 100% loss is based on margin, not credit received. This is something that is worth clarification in the software otherwise it could be confusing.

 

#1. This is exactly what we do. 

 

If you sell a 5 wide spread, and collect $1, the amount risked is $4.

 

If it expires worthless, you made $1 on a $4 risk, you make 25% (assuming margin requirements aren't less than the $5 spread).

 

If it expires at max loss ($4), then you lost 100%.

 

The only thing that is confusing ppl is the stop losses. 

 

#2. When calculating a stop loss, we do look at differently

 

This time let's do a short put @$1 for a $100 stock on the 100 strike (so, at-the-money)

 

Reg T says this requires at least 20% of the stock value ($20) minus the credit ($1), so this has a margin requirement of $19.

 

OK, then you go to the stop loss and think:

 

"alright, I sold this put @ $1, if it goes to $2, I'm out of this trade."

 

To us, that reads as:

"alright, I sold this put @ $1, if I lose 100% on my credit, I'm out of this trade."

 

So, we have the stops (and limits) set as a percent of credit or debit.

 

Hopefully that clears it all up -- the real takeaway is that we do use the margin amount as the max risk for the back-test calculations (what I presented above as #1)

 

 

Edited April 18, 2017 by Ophir Gottlieb
correction

[Yowster 9,180]

1 hour ago, Ophir Gottlieb said:

Hopefully that clears it all up -- the real takeaway is that we do use the margin amount as the max risk for the back-test calculations (what I presented above as #1)

@Ophir GottliebI have run some tests using credit spreads and this does NOT appear to be the case despite the Risked hyperlink saying it does use the margin amount. Run a few tests and see for yourself - it appears to be using the sum of the credits received as the risk amount. This results in the return percentage being way over-inflated for credit spreads, like 3x or 4x too high.  Just do an eyeball check on your own post with the selling VXX 40/15 delta call spread - that trade summary has a total risk of $5823 over 104 trades, which averages around $56 of risk per credit spread. That $56 is obviously wrong and way too low - selling OTM credit spreads for widths between 1.5 to 2.5 should have a risk of above $100 per spread (probably $150 or more). That $56 per spread figure is likely quite close to the average credit collected.  This has a huge impact on your tool as credit spreads will appear to be much more profitable than they really are, and make them seem like they are better choices than debit spreads when in some case they very well may not be.

Edited April 18, 2017 by Yowster

[Kim 7,943]

2 hours ago, Ophir Gottlieb said:

 

#1. This is exactly what we do. 

 

If you sell a 5 wide spread, and collect $1, the amount risked is $4.

 

If it expires worthless, you made $1 on a $4 risk, you make 25% (assuming margin requirements aren't less than the $5 spread).

 

If it expires at max loss ($4), then you lost 100%.

 

The only thing that is confusing ppl is the stop losses. 

 

#2. When calculating a stop loss, we do look at differently

 

This time let's do a short put @$1 for a $100 stock on the 100 strike (so, at-the-money)

 

Reg T says this requires at least 20% of the stock value ($20) minus the credit ($1), so this has a margin requirement of $19.

 

OK, then you go to the stop loss and think:

 

"alright, I sold this put @ $1, if it goes to $2, I'm out of this trade."

 

To us, that reads as:

"alright, I sold this put @ $1, if I lose 100% on my credit, I'm out of this trade."

 

So, we have the stops (and limits) set as a percent of credit or debit.

 

Hopefully that clears it all up -- the real takeaway is that we do use the margin amount as the max risk for the back-test calculations (what I presented above as #1)

 

 

That makes sense. As I said, maybe it's worth to mention it somehow in the software when using stop losses (this would apply to credit spreads only where margin is required). But P/L is calculated correctly, this is the most important thing.

One thing worth considering is adding average P/L per trade. When you see 200% cumulative return, it is a big difference if it's based on 10 trades or 50 trades. I understand that number of trades is part of the statistics, so average return can be easily calculated, but still it would be helpful to see it along with the cumulative return.

[Ophir Gottlieb 62]

1 minute ago, Kim said:

That makes sense. As I said, maybe it's worth to mention it somehow in the software when using stop losses (this would apply to credit spreads only where margin is required). But P/L is calculated correctly, this is the most important thing.

One thing worth considering is adding average P/L per trade. When you see 200% cumulative return, it is a big difference if it's based on 10 trades or 50 trades. I understand that number of trades is part of the statistics, so average return can be easily calculated, but still it would be helpful to see it along with the cumulative return.

 

Thanks, Kim -- great feedback.

We actually have avg win, avg loss, #wins, #losses, etc, if you click on a tile --that's how you get the details:

 

 

All there!

 

[cuegis 683]

 

"If you sell a 5 wide spread, and collect $1, the amount risked is $4.

If it expires worthless, you made $1 on a $4 risk, you make 25% (assuming margin requirements aren't less than the $5 spread).

If it expires at max loss ($4), then you lost 100%."

 

 

$4.00 = 100% of $1.00? Really?

 

[Ophir Gottlieb 62]

11 minutes ago, cuegis said:

 

"If you sell a 5 wide spread, and collect $1, the amount risked is $4.

If it expires worthless, you made $1 on a $4 risk, you make 25% (assuming margin requirements aren't less than the $5 spread).

If it expires at max loss ($4), then you lost 100%."

 

 

$4.00 = 100% of $1.00? Really?

 

This is a copy and paste of what I wrote above, but with boding emphasis added:

 

If you sell a 5 wide spread, and collect $1, the amount risked is $4.

 

If it expires worthless, you made $1 on a $4 risk, you make 25% (assuming margin requirements aren't less than the $5 spread).

 

If it expires at max loss ($4), then you lost 100%.

--

So, again, if your max risk is $4 and you lose $4, then your loss is 100%.

 

 

Edited April 19, 2017 by Ophir Gottlieb

[Kim 7,943]

18 minutes ago, Ophir Gottlieb said:

 

Thanks, Kim -- great feedback.

We actually have avg win, avg loss, #wins, #losses, etc, if you click on a tile --that's how you get the details:

 

 

All there!

 

Yes, those numbers are very helpful. But to me, the most important number in any strategy is "average % return". So yes, you can easily calculate it by dividing 40.9% by 26 (# of trades), but having it presented would definitely help to save time. Also, Avg win and Avg. loss would be more relevant to present as % rather than $. $467 return on 9k risk is not the same as $467 return on 5k risk. To me, what really matters is percentages, not dollar amounts.

It's like when someone telling you that he is making 10k per month in trading. Does it mean anything? Is it 10k on 100k account? Amazing! 10k on 10M account? Not so much..

[Ophir Gottlieb 62]

1 minute ago, Kim said:

Yes, those numbers are very helpful. But to me, the most important number in any strategy is "average % return". So yes, you can easily calculate it by dividing 40.9% by 26 (# of trades), but having it presented would definitely help to save time. Also, Avg win and Avg. loss would be more relevant to present as % rather than $. $467 return on 9k risk is not the same as $467 return on 5k risk. To me, what really matters is percentages, not dollar amounts.

It's like when someone telling you that he is making 10k per month in trading. Does it mean anything? Is it 10k on 100k account? Amazing! 10k on 10M account? Not so much..

 

OK, will add. Thanks!

[cuegis 683]

17 minutes ago, Ophir Gottlieb said:

This is a copy and paste of what I wrote above, but with boding emphasis added:

 

If you sell a 5 wide spread, and collect $1, the amount risked is $4.

 

If it expires worthless, you made $1 on a $4 risk, you make 25% (assuming margin requirements aren't less than the $5 spread).

 

If it expires at max loss ($4), then you lost 100%.

--

So, again, if your max risk is $4 and you lose $4, then your loss is 100%.

 

 

OK. I see the problem here. From all of the conversations I'm reading here, the one problem that seems to be prevalent is that many of us are speaking in different languages, with different definitions.

So, you are correct and so am I. We just are attaching different meanings to the same terms.Or vice versa!

[Kim 7,943]

2 minutes ago, cuegis said:

OK. I see the problem here. From all of the conversations I'm reading here, the one problem that seems to be prevalent is that many of us are speaking in different languages, with different definitions.

So, you are correct and so am I. We just are attaching different meanings to the same terms.Or vice versa!

I don't see any problem. Calculating returns based on maximum risk (or margin) is the only correct way to do it.

 

Many options gurus abuse the ignorance of the general public to inflate their returns. In the above example, if they get $1 credit on $5 spread and it expires worthless, they present it as 100% gain. In fact, it is 25% gain, not 100%. At wthe same time, if they lose $4 they would present it as 100% loss. Which is correct, just not consistent.

 

I believe the software presents the P/L correctly - the only issue is as discussed with stop losses.

[Yowster 9,180]

2 hours ago, cuegis said:

 

"If you sell a 5 wide spread, and collect $1, the amount risked is $4.

If it expires worthless, you made $1 on a $4 risk, you make 25% (assuming margin requirements aren't less than the $5 spread).

If it expires at max loss ($4), then you lost 100%."

 

 

$4.00 = 100% of $1.00? Really?

 

@cuegisI think of it this way.  With a credit spread, you collect your max gain upfront and if the trade works out then you get to keep all of it.  Your max risk is the margin requirement.

 

So making $1 on $4 risk is 25% gain. but losing $4 on $4 risk is 100% loss.   So, IMO, the last line of your post should be $4.00 = 100% of $4.00 (not $1.00 as you wrote because your are risking $4, not $1)

Edited April 19, 2017 by Yowster

[cuegis 683]

10 hours ago, Yowster said:

@cuegisI think of it this way.  With a credit spread, you collect your max gain upfront and if the trade works out then you get to keep all of it.  Your max risk is the margin requirement.

 

So making $1 on $4 risk is 25% gain. but losing $4 on $4 risk is 100% loss.   So, IMO, the last line of your post should be $4.00 = 100% of $4.00 (not $1.00 as you wrote because your are risking $4, not $1)

I think you are right.

There has been some confusion about the way to view this kind of thing because there are several ways, and different terminology.

But, in all cases, you are right about this one.

I guess I just never looked at it from this angle.

[Yowster 9,180]

@Ophir GottliebFirst off, thanks for bearing with me on the calculation of the Risked value (I think I am finally in synch there).   However, in looking at the detailed trade stats, I noticed something else that could skew results, and that is position sizing.   The only tuning knob available to the user in the tool settings is number of contracts, and this can result in significant differences in the amount invested in each iteration of the backtest trades.  This is caused by changes in underlying stock price as well as IV changes.   For example, in the VXX sell call 40/15 delta credit spread using a 6 month timeframe with 7 day rollover, the minimum trade had a width of 1, while the max trade had a width of 8 strikes (with many other different widths used between these values for other trade iterations).   So, in the case of min vs max using the same number of contracts there was a huge difference in the amount at risk when the width was 1 compared to when it was 8 - and the time when it was 8 counted a lot more to the overall return which could certainly skew overall performance results.  I believe the tools needs to provide a mechanism to have approximately the same amount at risk  for each iteration of the trade in the backtest - perhaps let the user specify a risk amount and have the tool adjust the number of contracts in each iteration to come closest to that value?

[Kim 7,943]

8 minutes ago, Yowster said:

@Ophir GottliebFirst off, thanks for bearing with me on the calculation of the Risked value (I think I am finally in synch there).   However, in looking at the detailed trade stats, I noticed something else that could skew results, and that is position sizing.   The only tuning knob available to the user in the tool settings is number of contracts, and this can result in significant differences in the amount invested in each iteration of the backtest trades.  This is caused by changes in underlying stock price as well as IV changes.   For example, in the VXX sell call 40/15 delta credit spread using a 6 month timeframe with 7 day rollover, the minimum trade had a width of 1, while the max trade had a width of 8 strikes (with many other different widths used between these values for other trade iterations).   So, in the case of min vs max using the same number of contracts there was a huge difference in the amount at risk when the width was 1 compared to when it was 8 - and the time when it was 8 counted a lot more to the overall return which could certainly skew overall performance results.  I believe the tools needs to provide a mechanism to have approximately the same amount at risk  for each iteration of the trade in the backtest - perhaps let the user specify a risk amount and have the tool adjust the number of contracts in each iteration to come closest to that value?

Very good point. It doesn't happen with many underlyings (we all know that VXX is a very special animal) but some stocks do move over period of 2-3 years.

[cuegis 683]

13 minutes ago, Yowster said:

@Ophir GottliebFirst off, thanks for bearing with me on the calculation of the Risked value (I think I am finally in synch there).   However, in looking at the detailed trade stats, I noticed something else that could skew results, and that is position sizing.   The only tuning knob available to the user in the tool settings is number of contracts, and this can result in significant differences in the amount invested in each iteration of the backtest trades.  This is caused by changes in underlying stock price as well as IV changes.   For example, in the VXX sell call 40/15 delta credit spread using a 6 month timeframe with 7 day rollover, the minimum trade had a width of 1, while the max trade had a width of 8 strikes (with many other different widths used between these values for other trade iterations).   So, in the case of min vs max using the same number of contracts there was a huge difference in the amount at risk when the width was 1 compared to when it was 8 - and the time when it was 8 counted a lot more to the overall return which could certainly skew overall performance results.  I believe the tools needs to provide a mechanism to have approximately the same amount at risk  for each iteration of the trade in the backtest - perhaps let the user specify a risk amount and have the tool adjust the number of contracts in each iteration to come closest to that value?

This is just a thought. I have not really thought it through. Plus I don't even know if it would be feasible but.

Would there be a way of having some larger "Master" number that one could input, which would signify "account size"?

Then, have all of your trades based off the assumption that , in any given test, you would be allocating 10%, for example, of the total account size to that trade?

I'm sure this is more difficult than it sounds.

Edited April 19, 2017 by cuegis

[Yowster 9,180]

24 minutes ago, Kim said:

Very good point. It doesn't happen with many underlyings (we all know that VXX is a very special animal) but some stocks do move over period of 2-3 years.

@KimI did the same test with AMZN using 3 year timeframe and simply buying 50 delta call and monthly rollover, and even in this case there were big differences in the amount invested each iteration.   Stock price rose a lot which had an effect, but even clicking on the never trade earnings button there were significant price increases as earnings dates approached.    So, I think for all stocks that have IV increases heading into earnings the tool will invest more in those months because the IV rise will cause the option prices to go up (even if we avoid holding over the earnings announcement).   Here is the open trade detail data for buying AMZN 50 delta calls with 30 day rollover over a 3 year timeframe and clicking the never trade earnings button - even in this relatively simple case there are significant differences in the amount invested each iteration.

 

Date             Open Price

21-Apr-14         $14.53

28-Apr-14         $11.33

30-May-14        $8.6

27-Jun-14         $12.1

28-Jul-14          $8.93

29-Aug-14        $7.05

26-Sep-14        $11.98

27-Oct-14         $8.6

28-Nov-14        $8.5

26-Dec-14        $9.4

23-Jan-15         $13.35

2-Feb-15           $12.18

6-Mar-15           $9.93

2-Apr-15           $15.27

27-Apr-15         $11.4

29-May-15        $10.75

26-Jun-15         $17.48

27-Jul-15          $17.15

28-Aug-15        $18.33

25-Sep-15        $25.2

26-Oct-15         $15.6

27-Nov-15        $19.98

24-Dec-15        $17.8

22-Jan-16         $29.05

1-Feb-16           $20.05

4-Mar-16           $16.85

1-Apr-16           $27.55

2-May-16          $19.23

3-Jun-16           $16.45

1-Jul-16            $31.33

1-Aug-16          $16.27

2-Sep-16          $13.15

30-Sep-16        $29.4

31-Oct-16         $19.08

2-Dec-16          $18

30-Dec-16        $23.1

27-Jan-17         $24.73

6-Feb-17           $15.7

10-Mar-17         $13.53

7-Apr-17           $27.55

 

 

[Ophir Gottlieb 62]

1 hour ago, Yowster said:

@Ophir GottliebFirst off, thanks for bearing with me on the calculation of the Risked value (I think I am finally in synch there).   However, in looking at the detailed trade stats, I noticed something else that could skew results, and that is position sizing.   The only tuning knob available to the user in the tool settings is number of contracts, and this can result in significant differences in the amount invested in each iteration of the backtest trades.  This is caused by changes in underlying stock price as well as IV changes.   For example, in the VXX sell call 40/15 delta credit spread using a 6 month timeframe with 7 day rollover, the minimum trade had a width of 1, while the max trade had a width of 8 strikes (with many other different widths used between these values for other trade iterations).   So, in the case of min vs max using the same number of contracts there was a huge difference in the amount at risk when the width was 1 compared to when it was 8 - and the time when it was 8 counted a lot more to the overall return which could certainly skew overall performance results.  I believe the tools needs to provide a mechanism to have approximately the same amount at risk  for each iteration of the trade in the backtest - perhaps let the user specify a risk amount and have the tool adjust the number of contracts in each iteration to come closest to that value?

 

We are adding functionality so you can specify spread widths by dollars (rather than deltas). So, you can back-test a $5 wide call spread that is always $3 out of the money (or whatever you decide). That way the risk per trade remains similar. You can expect that by next week.

[Yowster 9,180]

3 minutes ago, Ophir Gottlieb said:

 

We are adding functionality so you can specify spread widths by dollars (rather than deltas). So, you can back-test a $5 wide call spread that is always $3 out of the money (or whatever you decide). That way the risk per trade remains similar. You can expect that by next week.

@Ophir GottliebThat should make things better, although not as good as being able to specify an amount to risk each iteration and then have the tool compute the number of contracts/spreads to use for each iteration that comes closest to the value you specify.   How about my simple AMZN example in my prior post - buying 50 delta calls which still showed wide variations in the amount invested each iteration (min of $8.50, max of $31.33 per contract), what can be done to even out the invested amount each iteration for this case???   IMO, for a valid backtest, the tool should give equal weight (or as close as possible) to each iteration.  But, as things stand now that is simply not the case, as some test iterations will count much more to the overall return than others will.

[Ophir Gottlieb 62]

2 minutes ago, Yowster said:

@Ophir GottliebThat should make things better, although not as good as being able to specify an amount to risk each iteration and then have the tool compute the number of contracts/spreads to use for each iteration that comes closest to the value you specify.   How about my simple AMZN example in my prior post - buying 50 delta calls which still showed wide variations in the amount invested each iteration (min of $8.50, max of $31.33 per contract), what can be done to even out the invested amount each iteration for this case???   IMO, for a valid backtest, the tool should give equal weight (or as close as possible) to each iteration.  But, as things stand now that is simply not the case, as some test iterations will count much more to the overall return than others will.

 

Hi Yowster,

 

I think everything you are saying is valid -- and very shrewd, it's just not what our members want right now. We do polls every month and develop quite quickly. For example, today Excel downloads, Copy downloads, immediate versus wait timing on stops and 5 years of history will be added to the platform. In a matter of days later, we will provide dollar width spreads (not just delta). In a few weeks after that we will have custom strategies (build any custom strategy up to 4 legs) and build any custom timing around earnings (beyond our current functionality of two-days before and two-days after). 

Once that is done, we will ask again, and get another development list to be completed in days, weeks and a month. If an equal weighted back-test is in the top three, that's what will come next -- if vol charting, vol scanning, technicals, etc are more desired, then we will build that. We are community driven and so far that has helped us build extremely quickly with great purpose.

 

I hope this clarifies our production progress and feature list and i do appreciate yuor feedback -- it's very valuable and helpful.

 

My very best,

Ophir

[Yowster 9,180]

@Ophir GottliebThank you for the explanation, you can add equal-weighted backtest to your wishlist.     Although, I should note that if your upcoming Excel download function provides a gain/loss% along with each closing transaction, then I could easily compute the equal-weight return quite quickly and with little effort.   As things stand now, copy & paste of the trade details into excel still requires me to do quite a bit of data manipulation to compute this.

Edited April 19, 2017 by Yowster

[Ophir Gottlieb 62]

1 minute ago, Yowster said:

@Ophir GottliebThank you for the explanation, you can add equal-weighted backtest to your wishlist.     Although, I should note that if your upcoming Excel download function provides a gain/loss% along with each closing transaction, then I could easily compute the equal-weight return quite quickly and with little effort.   As things stand now, copy & paste of the trade details into excel still requires me to do quite a bit of data manipulation to compute this.

 

Yeah, we are actually giving two solutions:

 

1. Download into Excel ('nuff said)

2. Copy to clipboard

This is actually my preferred way. It is (i) faster than downloading a file and (ii) works in Google sheets faster.

 

But in any case -- both will be available in a matter of hours. Here is a snapshot of beta:

 

 

 

[Yowster 9,180]

@Ophir Gottlieb I'm sure the download & copy functions will work fine.  I am just saying that if the data includes a column for gain/loss% on each trade iteration that I could compute an equal-weight return percentage quite quickly and easily.   This kind of goes along with what @Kim suggested about preferring to see PNL results in percentages instead of dollars (or shown in addition to the dollars).  

Edited April 19, 2017 by Yowster

[Ophir Gottlieb 62]

 

 

The Surprising Reality of Option Trading in Morgan Stanley (NYSE:MS)

Date Published: 2017-04-19 
Written by Ophir Gottlieb 

LEDE 
In a bull market, selling puts or put spreads is generally a good bet -- but with Morgan Stanley (NYSE:MS), it turns that owning options, rather than selling them, has been monster winner over 3-years, 2-years, 1-year and even the last 6-months. 

PREFACE 
Morgan Stanley stock has actually been more volatile than you might have expected. For a company that isn't really a takeover candidate, doesn't really face a black swan downside risk and is generally considered a leader in its industry, the stock has not been settled. 

Here is the three-year stock chart via Yahoo! Finance: 
 

Getting long Morgan Stanley options can take many forms, but perhaps the simplest is a long, out-of-the-money strangle. 

STRANGLES 
Buying a strangle means buying both a call and a put. Here is how that strategy has done over the last 3-years, trading every month, and always avoiding earnings. 
 

We see a 446% return while the stock was up 40%. But, owning options is also a risky investment in that it is an investment that "something" will happen to the stock. Here is how that option strategy (in red) has done relative to the stock (in gray) over those 3-years. 
 

A chart is wonderful in that it really gets at the point -- while this strategy was a gigantic winner, it was filled with risk. We can see it bobbing up and down through time. 

But, as risky as this has been, it has, surprisingly, worked over all time periods. Here are the results over the last 2-years, again, trading monthly and always avoiding earnings. 
 

That's a 173% return when the stock rose just 16%. Again, here is the option strategy (in red) compared to the stock (in gray). 
 

Yet again, we can see the risk inherent in the options -- but when it comes down to it, it has been a big winner, returning nearly 10-fold the stock. Next we look at the last year: 
 

That's a 59.6% return while the stock has risen 59%. For the shorter time frame, the stock and options actually performed very similarly. 

WHAT JUST HAPPENED 
In a time when option portfolios may be leaning short puts and put spreads, it's nice to find an alternative to diversify that overarching short volatility bet in a name that has consistently shown more realized volatility than the options were pricing in. 

What an interesting result and an interesting hedge to a short vol option portfolio. 

This is how people profit from the option market - it's preparation, not luck. 

To see how to do this for any stock, index or ETF and for any strategy, with just the click of a few buttons, we welcome you to watch this 4-minute demonstration video: 

Tap Here to See the Tools at Work 

Thanks for reading. 

Risk Disclosure 
You should read the Characteristics and Risks of Standardized Options. 

Past performance is not an indication of future results. 

Trading futures and options involves the risk of loss. Please consider carefully whether futures or options are appropriate to your financial situation. Only risk capital should be used when trading futures or options. Investors could lose more than their initial investment. 

Past results are not necessarily indicative of future results. The risk of loss in trading can be substantial, carefully consider the inherent risks of such an investment in light of your financial condition. 

The author has no position in Alibaba Group Holding Ltd (NYSE:BABA) as of this writing. 

Back-test link.

 

 

 

 

 

 

[Yowster 9,180]

quick observation for anyone else plugging scenarios like the MS post above into the tool and getting different results.   Note that by default the setting for Execution Fill Type is halfway between mid and market (and I think that is what is used in these samples), but I had changed my setting to mid market because the vast majority of SO trades I am able to open/close within a few cents of mid-point pricing.   This had a huge effect on the %return when compared to what was in the article (mine was much higher).   I would expect a bit higher but the result was much higher.   This got me thinking Why such the big difference???  I came up with 2 thoughts:

1. The tool is currently using end-of-day data, and I have noticed that bid/ask spreads at market close tend to be a little wider than what they are during market hours.
2. When closing winning trades, the options with bigger gains are ITM by quite a bit and the bid/ask spreads for deeper ITM options tend to be quite a bit wider than those near ATM (especially at market close).

 

What setting is best to use???   That is up to the user.   But after many years of options trading I can safely say that the vast majority of my trades are filled within a few cents of the bid/ask mid-point (if I have to chase too far from the mid-point to open a trade, I typically don't open it).