Red Option "newsletter" (part of Thinkorswim/TD Ameritrade)

[cwerdna]

Anyone here subscribe to www.redoption.com? I only heard about them recently via a marketing email since I have TD AM/TOS accounts. I think they've pitched the service to me before when I wasn't paying attention and knew nothing options.

My email had a deal where I could get 2 strategies free for 60 days. I later got an update email that had another code for 2 more strats free for 60 days. They're normally $20/month per strat. I'm only paying for 1 strat right now...but will be paying for all the ones I still want when my trial's over.

I'm currently subscribed to iron condor, index calendar, vertical spread, collar and weeklies and seeing how they are.

One great thing about them is that they to have autotrade but only in TD AM/TOS accounts. You just set how much you want to allocate per trade. They sound out emails discussing the trade, risk, rationale, etc. They seem to be pretty active in taking profits, reducing risk, etc. by making adjustments/closing part of the position. That could suck if you're having them do small trades and on TD AM's ripoff default commissions ($9.99 + $0.75/contract) but I'm not .

I asked them how it's done since I once didn't get a fill on an advisory that went out. Turns out I had insufficient buying power at the time and they don't want to take people into margin call. As they explained "These orders fill at the same time the advisory is sent. All contracts are filled under a house account then allocated to individual accounts shortly after. " That sorta explains why it's TD AM/TOS only for autotrade. That also explains how they can get the pricing on fills for everyone.

Too bad they can't autotrade on IB.

So far, they seem ok but I haven't been on them for long. IIRC, I haven't hit any losses yet and have had some small to decent gainers. Some positions have been partly closed (to take some gains). For others they've been rolling.

I did see one notification go out about a loss but I didn't have any position in that. The trade began before I subscribed.

[cwerdna]

I went thru compiling my P/L so far w/the closed trades. I'm not 100% sure this is complete so I'll have to go back and double check my emails vs. TOS.

They seem to have done pretty well for me, so far w/o a single loss yet. In some notification emails of trades that don't involve me (opened before I joined), I have seen some losses.

Here's what my profit so far (after commissions) by trades. I joined on Nov 6th.

(weekly strat:)

SPY calendar: 36.3% gain

CRM dbl calendar: 13.6% gain

(index calendar strat:)

DIA calendar: 25.1% gain

(vertical strat:)

SPY vertical: 65.3% gain

RIMM vertical: 24.8% gain (they opened this on 11/26 and closed it on 11/27)

(collar strat):

DD collar: 1.0% gain

(Iron condor strat:)

RUT IC: 6.7% gain

Obviously, this is not a long track record, but so far, paying for the collar strat doesn't seem worth it as each strat is $20/month (once my trial is over). The profit was only $44 on that DD collar (on an investment of $4252, including commissions) which they opened on 11/15/12 and closed on 12/10/12.

On the SPY calendar 36.3% winner, I had a debit of $2025.98, credit of $1546.92 from a roll and $1215 credit to close, making for a $735.94 profit.

All of this stuff was autotraded for me. Crossing my fingers that they can keep it up with lots of winners and few losers (hopefully small ones).

.

Edited December 14, 2012 by cwerdna

[RobertB]

cwerdna,

I once accepted TOS's trial of Red Options, and I was very unimpressed with their methodology. This was about three years ago. Perhaps it's different today. But they seem not to understand the nature of probabilities and options market pricing and risk. They made a lot of statements that were foolish and ignorant. (again, this was three years ago). For example, they would say things such as, "This trade has an 80% chance of being profitable" without accounting for the trade's risk and reward properly. Over time one would expect in aggregate either a wash or a loss if i recall from my tests. When I called them up to discuss any of this, they didn't really seem interested. If they are still making their picks in the same way, I would attribute your recent success with them as pure luck.

[cwerdna]

Interesting. Well, given their history I see at https://www.redoption.com/redop/about and the analysis tools that the Thinkorswim platform provides, I'm not surprised by their methodology. I have been thru some of the Investools training that well (although, admittedly, I need to go back through some of it again, in more detail). What I see in their emails/advisories seems more/less inline what's taught in Investools training and what you can get out of the Analyze tab.

The sample trades at https://www.redoptio...edop/strategies are similar to the stuff they currently send out.

Here's excerpt of an email for an IWM trade they opened earlier this month:

Probability: There is a 64% probability that our short $79 strike will be out of the money at Jan expiration. This fits in our guideline of over 60% for short credit vertical positions. Time Decay (Theta) also works in our favor as our short vertical loses value over the life of the trade.

Risk: We are risking $1.49 to make a potential $0.51 on this short vertical position. We profit in three out of four scenarios: if IWM trades higher, remains at current levels or trades lower but remains above our break-even level at Jan expiration...

I guess we will see. It looks like today I'll have a loss on one of their plays, a SPY double calendar. There was an opening debit of $1.22, $0.56 credit on a roll, $0.15 credit to close a call calendar w/a remaining risk of $0.51. That remaining risk was some December (week 2) 133 puts that expired worthless today.

When you tried out their trial, did you turn on autotrade? Which strategies did you choose? Did you make money?

Edited December 14, 2012 by cwerdna

[RobertB]

Thanks for the further information.

Looking at the trade excerpt you posted, I don't see that as a good trade at all. Perhaps there are other aspects to the trade that you didn't post, but assuming the probability analysis is based around expiration time, the numbers don't add up in your favor.

Assume you did that trade 100 times. 64% of the time your short position expires worthless and you profit. 36% of the time it doesn't. That means that out of 100 trades your profit would be (64 * 0.51) = 32.64, while your losses would (36 * 1.49) = 53.64.

The probabilities are definitely not in your favor. Not to mention the huge risk of a drawdown occurring.

To answer your question, I never traded any of their recommendations after analyzing the trades they offered. It simply appeared that they find trades that have a high probability of success, while ignoring risk.

For example (and I'm just making this up on the fly): Suppose I were to toss a fair coin six times. My offer to you is that I will pay you $1000 if at least one tail appears during the six tosses. But you have to pay me $75,000 if the six tosses all come up heads. That would be a bad bet for you take because if you were to take this bet 64 times (which is the number of combinations of heads and tails possible), you would expect to gain $63,000, but lose $75,000. Eventually, probabilites do catch up with you. And my experience three years ago with Red Option was that the kinds of trades they provide are really no different from that coin toss example. In the coin toss example, you have a 63/64 chance of winning the bet, but when you lose 1/64 of the time, you lose all your profits and more.

[cwerdna]

I see... at the risk of getting into trouble (?), I'll post the entire email (minus the boilerplate stuff at the bottom about risks, who to call for support.) (Sorry about highlighting I'm too lazy to remove it all while retaining the rest of the formatting.)

Trade Advisory

December 4, 2012

RED Option Strategy: Vertical Advisory

Underlying: iShares Russell 2000 ETF (IWM)

Status: Opening Trade

Trade: Sell the Jan 79 puts and buy the Jan 77 puts

Trade Price: $0.51 credit

Underlying Price: $81.72

Trade Risk: $1.49

Trade Duration: Short Term

Buying Power Reduction: $1.49

Trade Explanation: For the Vertical Advisory in IWM, we are selling the Jan 79/77 put vertical for a credit of $0.51 to open.

Price Action: IWM is the ETF that reflects the small cap Index. Shares have fallen today after hitting monthly highs just yesterday morning. IWM is now firmly above it 50-day moving average, which becomes support to the downside and would work well for this bullish vertical. We expect IWM shares to remain in a tight range and perhaps retrace higher into the end of the year. We have a cushion of 3% to the downside before IWM reaches our break-even of $78.49.

Volatility: Volatility has begun to expand slightly in IWM over the last few sessions which increases the premium collected on short out-of-the-money verticals. We went out to the Jan monthly cycle to collect more premium in the vertical.

Probability: There is a 64% probability that our short $79 strike will be out of the money at Jan expiration. This fits in our guideline of over 60% for short credit vertical positions. Time Decay (Theta) also works in our favor as our short vertical loses value over the life of the trade.

Risk: We are risking $1.49 to make a potential $0.51 on this short vertical position. We profit in three out of four scenarios: if IWMtrades higher, remains at current levels or trades lower but remains above our break-even level at Jan expiration.

Trade Duration: We have 45 days to adjust or close this trade at Jan expiration. Our ideal scenario is for IWM to remain above our short $79 strike for the duration of the trade.

Logic: IWM has seen resurgence recently and the overall market looks to be trending higher. We want to take advantage of the recent pull-back this afternoon and initiate the bullish directional vertical. Although the ETF could continue lower, we are giving ourselves a decent 3% cushion to the downside. We have good probabilities for the amount of credit collected vs. the risk in the trade.

We will continuously monitor our positions to determine if adjustments need to be made or when to close out of the trade.

RED Option

Original Trade Price: $0.51 credit

Closing/Adjustment Price:

I have to head out right now but I'll comment more later. I've observed that they do seem to be pretty good about making adjustments and taking profits (sometimes by closing 1/2 the position) and reducing amount of risk left. But hey, I'm just an options amateur.

I picked this example because it was one I quickly dug up that had a % probability listed.

Edited December 15, 2012 by cwerdna

[ammarmalhas]

cwerdna, you are an "options amateur"?

Hehehe, I wish I were an amateur like you!

[Kim]

I checked the service few years ago and I agree with Robert, but maybe things have changed.

As a side note, a service that is affiliated with a brokerage has a HUGE conflict of interests. My impression was that they are more concerned about generating commissions for TOS than profits for their subscribers. In the calendar service, they had a lot of cheap calendars in the 0.50 range. If you are paying 1.25 per contract, the commissions will eat a whopping 10% of the trade value. So you need 10% gain just to break even.

Why do you think they are so cheap and they auto-trade with TOS only? The subscription fee is a very small part compared to commissions that they generate for TOS.

[cwerdna]

As a side note, a service that is affiliated with a brokerage is a HUGE conflict of interests. My impression was that they are more concerned about generating commissions for TOS than profits for their subscribers.

...

Why do you think they are so cheap and they auto-trade with TOS only? The subscription fee is a very small part compared to commissions that they generate for TOS.

On your first point, that had also crossed my mind and more so as I've been examining how commissions heavy the trades tend to be. I agree that it is possibly a vehicle for them to generate commissions.

I figure that if the service sucks (results in net losses), that they won't be able to have a sustainable business and will lose subscribers. However, if there enough incoming new people to replace them... But yes, the subscription fee per strat is tiny compared to the commissions.

I wonder how other newsletters handle auto-trades across multiple brokerages? Maybe I should start another thread for that? To me, there'd be the same issue that we have here, where a trade alert goes out, adjustment, roll, closure, etc. goes out, they could automatically place a limit order but there's no guarantee it'll be filled. And, how far away should the limit be set?

With Red Option's use of a house account and divvying up afterward, at least everyone w/autotrade on gets a fill at the price mentioned in the alert.

[Kim]

I went thru compiling my P/L so far w/the closed trades. I'm not 100% sure this is complete so I'll have to go back and double check my emails vs. TOS.

They seem to have done pretty well for me, so far w/o a single loss yet. In some notification emails of trades that don't involve me (opened before I joined), I have seen some losses.

Here's what my profit so far (after commissions) by trades. I joined on Nov 6th.

(weekly strat:)

SPY calendar: 36.3% gain

CRM dbl calendar: 13.6% gain

(index calendar strat:)

DIA calendar: 25.1% gain

(vertical strat:)

SPY vertical: 65.3% gain

RIMM vertical: 24.8% gain (they opened this on 11/26 and closed it on 11/27)

(collar strat):

DD collar: 1.0% gain

(Iron condor strat:)

RUT IC: 6.7% gain

Obviously, this is not a long track record, but so far, paying for the collar strat doesn't seem worth it as each strat is $20/month (once my trial is over). The profit was only $44 on that DD collar (on an investment of $4252, including commissions) which they opened on 11/15/12 and closed on 12/10/12.

On the SPY calendar 36.3% winner, I had a debit of $2025.98, credit of $1546.92 from a roll and $1215 credit to close, making for a $735.94 profit.

All of this stuff was autotraded for me. Crossing my fingers that they can keep it up with lots of winners and few losers (hopefully small ones).

.

Did you ask to see their complete track record before starting trading? I did. The results are not pretty. (all results are overall yearly returns before commissions, meaning average return per trade).

Collar strategy shows -10.61% loss for 2012. Out of 10 trades, 3 100% losers. Your DD collar shows 36.37% gain.

Equity calendar shows -10.28% loss, including 2 100% losers.

Index calendar shows -8.30% loss, including 2 100% losers. Your DIA calendar shows 32.00% gain.

Vertical shows -0.75% loss, including 4 100% losers.

Weekly and double double are the only strategies showing a gain (around 5-8%). However, as you can see from the examples, your returns will be typically 5-7% less due to commissions, so even those strategies would lose money after commissions. What is even more troubling is the number of 100% losers.

[Kim]

cwerdna,

I once accepted TOS's trial of Red Options, and I was very unimpressed with their methodology. This was about three years ago. Perhaps it's different today. But they seem not to understand the nature of probabilities and options market pricing and risk. They made a lot of statements that were foolish and ignorant. (again, this was three years ago). For example, they would say things such as, "This trade has an 80% chance of being profitable" without accounting for the trade's risk and reward properly. Over time one would expect in aggregate either a wash or a loss if i recall from my tests. When I called them up to discuss any of this, they didn't really seem interested. If they are still making their picks in the same way, I would attribute your recent success with them as pure luck.

This is an excellent point Robert, and one that many people miss.

Probability of success is directly related to the risk/reward. When you have a 80% probability of success. your risk/reward is typically not good - you would risk around $4 to make $1, but you will win 80% of the time. The question is how much you lose when you lose.

Red Option Iron Condor strategy is a good example. They did 21 trades in 2012, 16 winners and only 5 losers, but the overall results is -0.57% loss. Why? Because the typical winner was around 15-20%, but out of 5 losers, one was 100% and two 70-80%.

As a side note, most of the IC trades are $2 spreads, so with 1.25 commissions, the -.57% loss will become -10% loss. This confirms once again the concern that their main goal is generating commissions for TOS and not profits for you (otherwise why trade $2 spreads on SPY and IWM when you can trade $10 spreads on RUT?)

[cwelsh]

I would also like to point out that the "80% probability of success" or "67.4% chance the short strike will get hit" is a load of complete BS.

Sure, that's what the deltas tell you, but any option trader that has ever looked at it and tracked it knows (or hopefully knows) that those are "theoretical" probabilities, and not actual probabilities.

I tried spreads and ICs all of the time. In trading IWM spreads this year, I ONLY opened a trade if it was outside of a 2SD probable move (so 95.45% of the move). And while I used my calculations for STD, every time, the "probability of success" according to TOS was well above 90%. However, out of 41 trades on IWM this year so far, 7 of them had to be adjusted or closed early. That means "theoretically" 90-95% of the trades "should" have been successes, but only 83% were.

This is not unique to any one instrument, it applies to everything. Heck, look at AAPL the last two months, whether you are on a 1 day, 5 day, 7 day, or 30 day window, we have had 2+SD daily moves over 10 times in the last thirty days according to the TOS "probability of success."

NEVER base your expected returns off of "probability of success." Do REAL backtesting, using REAL data -- actual chance of success is virtually always lower. Of course that does not mean you can't still succeed at it -- just have to plan accordingly.

[Kim]

Chris,

While you might be right in general, the fact that 7 trades had to be adjusted is not necessarily an indication that the probability of success is not correct. The probability of success refers to expiration - 80% success means that in 80% of the cases, the underlying will be between the strikes at expiration, but during the life of the trade, this probability is lower. Many times the underlying will go outside the range but retreat back by expiration. That doesn't mean you should not adjust - in fact, usually you should adjust much before the underlying gets to the sold strike.

[cwelsh]

My example obviously includes adjustments, but in any historical analysis, you'll find the probabilities are mistaken, often grossly so. The problem is those probabilities, for the most part, assume normal distribution (some MM use logarithmic distribution), but that does not match reality.

There is not one stock in existence that, over any significant period, has any type of normal distribution.

Take for instance AAPL, how many times in the past few months has an option that had a 30% or less chance of finishing in the money, finished in the money? --

The Oct1 655 put only had a 17% chance of finishing in the money on the day it was issued -- and it did;

The Oct2 630 put only had an 11% chance of finishing in the money on the day it was issued -- and it did

The Oct finishing had no options with a 30% or less finishing ITM

The Oct 4 605 put had a 27% chance of finishing in the money -- and it did

The Nov 1 580 put had a 27% chance of finishing in the money -- and it did

The Nov2 550 put had a 5% chance of finishing in the money -- and it did

The Nov finish had no options with a 30% or less finishing

The Nov3 570 Call had a 14% chance of finishing in the money -- and it did

The Nov4 had no options with a 30% or less finishing

So in a single two month period, SIX out of nine options, that had LESS than a thirty percent chance of finishing in the money -- did. (This of course means all those options with a 35% chance, 42% chance, etc., also finished in the money).

The disparity gets even more pronounced the closer you get to 50%. Yes, this is an extreme example over a defined time period -- but it holds true on every single option there is over time. The theoretical probabilities simply don't reflect reality, and anyone who trades options needs to wrap their head around that or they won't understand the risk they're taking.

[Kim]

So you are basically saying that over time, options underprice the risk? I'm not sure I accept that.

One of the edges of IC sellers is the fact that over time, IV tends to be slightly higher than HV. AAPL is definitely behaving differently lately, but if you go to any index, you will see that on average, IV is higher than HV.

[RobertB]

So many people get very confused with probabilities. Those with no formal education in the subject can get definitely confused. But even those with a formal education in the subject easily can get off track.

One area of confusion I see often is people thinking that processes, such as stock movement, follow probabilites, instead of probabilities describing processes. Probabilities are just mathematical descriptions based on past history and other considerations. The processes themselves know nothing about the probabilities. They don't know when you had started counting your sequence, for instance. They have no memory.

How about the so-called gambler's paradox. A gambler is watching a roulette wheel and notices that it has landed on black ten times in a row. He knows that the odds are 50/50 (ignoring "0" and "00" for the moment) between landing on a black or a red for each individual spin. He also know that the probability of getting ten blacks in a row is 1/1024 (2 to the tenth power is 1024); that is, virtually zero. So he thinks that FOR SURE the next spin has to be a red since the chances of NOT getting a red in 11 spins is virtually zero.

What the gambler has forgotten (or, more likely simply ignorant of) is that the roulette wheel has no memory. For each individual spin the chances are 50/50 between red and black regardless of the recent past.

So even though it is true that BEFORE THE 11 SPINS there is an exact probability of 1/2048 that in 11 spins of a roulette wheel you would get 11 blacks in a row, once the first ten have occurred and you notice all blacks, you cannot say anything about the next spin other than that there's a 50/50 chance for either black or red--just like all the other spins. Again, the roulette wheel has no memory of what has occurred prior.

The same is true with stock market probabilities. Even if they were exact like the roulette wheel (and stock market probabilites are surely not exact), you can use them only as guides into the future for a particular time range. But once the stock movement has started to occur, you need to at the very least recalculate the probabilities for whatever range of time is still of interest to you. If the stock hasn't moved as much as it was "supposed to" have during the initial part of the time range, you cannot assume it will "make it up" in the latter part. If you think this, you are falling victim to a version of the gambler's paradox.

One more thing: Along the lines of what I just said, I've seen traders completely misuse probabilities and standard deviations when attempting to make correct market decisions. I once read a pseudo-scientific article written by someone whom I believe was sincere. But ... oh so how misguided! His system went something like this. Throughout the trading day he constructs a bell-shaped (normal) curve of a stock based on its intraday movements. If a majority of the bell curve seems to be being built on either the left or the right side of the bell, he says there is a great chance that the rest of the bell "has to be" built on the opposite side. Thus, he would use this "knowledge" to day trade the latter half of the trading day.

The errors in his thinking are numerous. One of the errors, of course, is that the stock movement doesn't "know" its on someone's left- or right-side of a bell. Indeed, if he had instead constructed a bell in real time with a total time range of only the first half of the day (instead of the full day), he would have ended up with a completely different bell that is is "filled in" on both sides. Everything about his method is arbitrary including the time range. There are an infinite number of time ranges and an infinite number of possible histogram bin possibilities when constructing your histogram. Yet he swore by this "scientific" system. Lots of ignorance abounds.

[cwerdna]

cwerdna, you are an "options amateur"?

Hehehe, I wish I were an amateur like you!

Yes, I am a total amateur compared to Kim and some of the other folks here. :/

Did you ask to see their complete track record before starting trading? I did. The results are not pretty. (all results are overall yearly returns before commissions, meaning average return per trade).

Collar strategy shows -10.61% loss for 2012. Out of 10 trades, 3 100% losers. Your DD collar shows 36.37% gain.

Equity calendar shows -10.28% loss, including 2 100% losers.

Index calendar shows -8.30% loss, including 2 100% losers. Your DIA calendar shows 32.00% gain.

Vertical shows -0.75% loss, including 4 100% losers.

Weekly and double double are the only strategies showing a gain (around 5-8%). However, as you can see from the examples, your returns will be typically 5-7% less due to commissions, so even those strategies would lose money after commissions. What is even more troubling is the number of 100% losers.

I'm embarrassed to say I did not (long story). I did request it earlier today and you're right on those. It is not pretty at all. I've discontinued autotrade on all my subscribed strategies for now and well let the remainders get closed per their recommendations.

It's rather dismaying and am surprised (or maybe I shouldn't be?) that they're continuing to offer all these strats.

They calculated collars differently than I did but I see where they're coming from. The way they calculated is a little odd (to me).

I also got their 2011 track record. Double double last year made money but was only 3.49% gain, which is less than what it's gotten so for YTD. That strat seems somewhat commission heavy as well due to the # of legs. Weeklys last year lost 17.62% (vs. making money this year).

The only strat that made $ both years was covered calls. At least they seem to be decent stock pickers. My only reservations about it are that it ties up a lot of capital for potentially a long time and something very bad could happen to a stock where the premium from the call isn't anywhere near sufficient to make up for the loss.

Thanks guys. This has been a useful and interesting discussion. I think I'm going to go w/covered call and drop all the rest once the positions are closed.

FYI, if anyone wants to subscribe or at least wants to try covered call, the coupon codes MARKETBLOG (I used this before and it's in their emails) and TICKER (just found it on the net) should still work for 2 strats free for 2 months. SMACKDOWN was another code I'd received (in marketing email) which I did use. I'm not sure if that still works.

You don't have to autotrade.

Edited December 18, 2012 by cwerdna

[Kim]

Covered calls worked well, but don't forget that we are in the up trending market most of 2012. Major indexes are up 15-20%. Simply doing CC on SPY would probably produce better results with less work.

Double double worked not bad, but like you mentioned, look at the commissions. They had a lot of trades under $1. Those are 4 leg trades, so assuming 1.25 per contract, that's $10 commissions. With $1 spread, the gain is reduced by 10%. Sp 8.86% gain becomes a loss.

[cwelsh]

Robert:

Great post, and I think there are two important points to be taken:

1. In the example you gave, the article writer was assuming the market HAS to revert to the mean, which is just idiotic, if that was the case then, over time, the markets would have no movement. Your roulette example was good. Just because something has happened 20 times in a row, does not mean its any more than a 50/50 shot it will happen the next time. BUT, going forward the math would be good -- so the odds of getting black ten times in a row AGAIN would be 1/1024.

2. Stock market probabilities DO change every tick. If you entered into a vertical call spread that only has a 10% chance of hitting the short strike, but the market spikes up, the next day, the chance of hitting the short strike is going to be higher -- NOT the original 10%.

As to Kim:

I am 100% certain the option market does not efficiently price -- it gets pretty darn close, but again, pick any index, or any stock, over any time period, and just graph out the daily price movements (or go read Taleb's books). They do NOT follow a normal or logarithmic distribution -- the number of outliers is always "improbable," but yet it happens in almost any market conditions.

I got into quite a heated argument with an MBA finance professor here in town over the inefficiencies in option pricing. I challenged him to find any stock, or any index, that matched the theory. We went through all the major indexes and probably 50 stocks -- and NONE of them did.

I spent quite a lot of time over a number of years tracking "option" probabilities -- and the daily probability values simply don't match up to what actually occurs. Recently they have been greatly understating risk (which tends to happen in low volatility markets). In 2008 they were greatly overstating risk (as generally happens in times of high volatility). I'm not saying options always underprice risk -- sometimes they overprice it. What I am saying is that the assumed probabilities are NOT the real probabilities.

Take Robert's roulette example. Sure the gambler just witnessed 10 blacks in a row, but the next time the chance is only 50% that a black will show up. Over time, its been experimentally proven, that the odds balance out. So, over 100,000 spins, you're going to have pretty close to 50,000 reds and 50,000 blacks -- but anywhere in those 100,000 spins, you might have a span of 10 blacks in a row -- or perfectly alternating red/black -- or 21 reds in a row.

That's NOT how option probabilities work. On Monday the 100/110 vertical call of stock XYZ (currently at 90) might have a "probability" of 10% of having the short strike hit. On Tuesday the price has jumped to 103. The "probability" of the 100 strike finishing ITM is probably now around 52 or 53%. On Wednesday let's say the price has jumped to 110. Now the "probability" is in the 70% range. Yet then on Thursday, the price plummets back to 94. On Friday it goes to 99. Well, your Monday probability was "correct" and your Tuesday and Wednesday were "wrong." If you keep a chart, over an extended period of time -- these probabilities DO NOT balance out, as they do with red/black on the wheel.

This is true for the simple fact that option pricing assumes either (a) normal distribution of price movement or ( logarithmic pricing of stock movement AND prices in a measurement of volatility. Due to the fact that assumptions (a) and ( are FALSE assumptions and that volatility is fluid -- you end up with inaccurate option probabilities.

If this wasn't true, I'd have retired filthy stinking rich by now. Why? Simple -- I would run a scan every single day across all options, finding the "misspriced" ones -- e.g. the ones that "pay" more than their probabilities. By way of example -- if a $10 vertical call spread had a ten percent probability of finishing ITM and I could receive $1.20 on the spread, over time, I would make 20%. Yet, over any significant time period, this does not hold true.

MOST people refuse to accept this, academics in particular. But that's simply a refusal to accept reality. I occasionally needle the professor I know, offering to pay a large sum of money, when he can produce any stock or index that behaves how the option probabilities predict -- in several years he has yet to find one real world example that matches the option probabilities.

And again -- sometimes the probabilities understate risk and sometimes overstate -- I've yet to find out a reliable way to make that determination, but my spread models try too -- I'm constantly tweaking them.

Taleb does a much better job explaining this than I do -- but simply put, probabilities in trading do not match real life outcomes.

[Marco]

I'm not sure about stocks but index vol is rich on the long run (look at the performance of indices like PUT and BXM). You can maybe argue whether market prices tail risk correctly (too cheap?)

As you say yourself probabilities change every day so the probability of any option to end up in the money today is just an estimate and therefore very likely to be wrong (hence the market reassesses that probability every day) that doesn't mean for me that the option model that the market is using is wrong big time (again maybe with the exception of tail risk)

Even if it is wrong - every decission I make about the other major price impacts (mainly spot and imlied vol) will have a much bigger impact on my trading. Unless you really have a market neutral (delta, gamma, vega) portfolio which buys and sells options with 'wrong' probabilties it matter much more where the market is going (spot and IV)

[cwelsh]

I also want to apologize for completely hijacking this thread .

If anyone wants to really track the probabilities (and what I did), pick a stock and/or index and every day, sometime after the open, write down the strike for the 33%, and 10% probabilities. So for instance, if you wanted to figure out the accuracies of the AAPL weeklies, on AAPL right now I would enter in my log the following:

535 call 36%

555 call 8%

520 put 32%

500 put 10%.

Then, at expiration, mark the expiring price and which ones finished ITM. Tomorrow the above will change of course, so my table will be different. Over a period of a year, you'll end up with 502 entries per "percentage" (251 trading days in the year). That's a fairly large sample. If the percentages are accurate, then you SHOULD have 167 of the 33%, and 25 of the 10% that finish ITM. However, that simply doesn't happen.

[Marco]

I also want to apologize for completely hijacking this thread .

If anyone wants to really track the probabilities (and what I did), pick a stock and/or index and every day, sometime after the open, write down the strike for the 33%, and 10% probabilities. So for instance, if you wanted to figure out the accuracies of the AAPL weeklies, on AAPL right now I would enter in my log the following:

535 call 36%

555 call 8%

520 put 32%

500 put 10%.

Then, at expiration, mark the expiring price and which ones finished ITM. Tomorrow the above will change of course, so my table will be different. Over a period of a year, you'll end up with 502 entries per "percentage" (251 trading days in the year). That's a fairly large sample. If the percentages are accurate, then you SHOULD have 167 of the 33%, and 25 of the 10% that finish ITM. However, that simply doesn't happen.

you say you have done this? (over a year?) do you mind sharing the no's?

thx

[cwerdna]

I also want to apologize for completely hijacking this thread .

No worries. This has been an interesting discussion.

Maybe RobertB is right... it doesn't seem like Red Option's strategies of using these theoretical probabilities (at least at they levels they choose) is a very good idea given their not so hot returns.

[Kim]

The problem is not necessarily theoretical probabilities. The Iron Condor strategy is a good example. Like I mentioned, they did 21 trades in 2012, 16 winners and only 5 losers, but the overall results is -0.57% loss. Why? Because the typical winner was around 15-20%, but out of 5 losers, one was 100% and two 70-80%.

The winning ratio was 76%, which is actually pretty close to theoretical probabilities, if you look at the credit they get. Here is the problem: when your typical profit is around 15-20% on the winning trades, you cannot afford to lose more than 25-30% on the losing trades.

[cwelsh]

ASSUMING probabilities are correct (obviously I think they are not and others do, but lets take that out of the discussion), Kim's point is a good one, and one I really hope all option traders understand.

If you have a theoretical probability of success of 90% per trade and you are only getting a theoretical maximum of 10% return per trade, over the long run you'll lose. This might seem like a "break even" strategy -- but it's not. You NEVER get the theoretical return, if from commission loss only. You also typically have slippage, and if you close out a trade instead of letting it expire, you'll have even less of a gain. If you are REALLY good (e.g. lucky), and can close your losers out quickly before facing a big loss, and let all of your winners ride completely, and are never wrong, this MIGHT be a winning strategy.

What do I do? I take 2 SD trades (95.45% probability of success) that only pay 10% or more. I "hope" this gives me a big enough error margin, closing early, and slippage. If we toss out these past three months, I've been quite successful at it. (This week in particular has not been fun).

That said, the problem with the Red Option newsletter is as Kim noted. If you're trading 80% probability credit spreads, for a 15% return (after commissions and slippage), your odds of winning in the long term are basically zero. (If you don't understand this last sentence please either post a message or send KIm or I a note).

[Marco]

What do I do? I take 2 SD trades (95.45% probability of success) that only pay 10% or more. I "hope" this gives me a big enough error margin, closing early, and slippage. If we toss out these past three months, I've been quite successful at it. (This week in particular has not been fun).

if you say you take 2 SD trades do you look back at historical prices/returns or do you take the implied SD/probability in the options?

Both can be totally off (too high or too low) if you compare them to reality and with the benefit of hindsight once the trade has expired or is closed.

Which brings me back to my point of: I don't believe in a massive misprice of options and even if there is one - will the impact on option prices be bigger then the impact of my timing in buying/selling at the right time (in terms of picking spot and IV with a good timing)?

[Kim]

One of the reasons why I don't offer auto-trading:

 

http://www.ripoffreport.com/r/TD-Ameritrade-IncRed-Options/Omaha-Nebraska-68127/TD-Ameritrade-Inc-Red-Options-Misled-me-to-believe-investing-in-Options-CallsPuts-and-1053998#comment_3

[David007]

I have used Red Option for the past 3 months.  They have made numerous trades with very few wins.  My advice is to learn to do it yourself before you allow them to trade for you. I just do not understand if you a using a computer system how can you have so many wrong picks for options. Red Option should be able to get at least 55% but they are doing about 45% and let negative picks grow to outrageous size.