Insane Emergency RE strategy

[ERD50]

Quote:

Originally Posted by dixonge After a quick count it looks like I'm overdue! So far I've placed 37 of the recommended trades, 2-4 contracts each, none have gone ITM or required a buy-back.

and a bit later...

Quote:

Originally Posted by dixonge Reminds me of the roulette table gain. I was always surprised how many times black or red could be hit in a row.

I think you answered your own question.



Now, one for haha and FIRE'd@51:

haha - what do you think of FIRE'd@51's analysis of an ~2.8% annual return on option selling? It makes sense to me, the option seller is putting up cash and has to make some money to do that, so some return should be expected. But "excess" returns would get arbitraged away, I would think.

FIRE'd@51 - well, two questions now.

1) Do you think that the calculation is very sensitive to slight changes in those numbers - would you expect an average of 10 'snapshots' to be pretty close to the 2.8% number?

2) After typing my note to haha, it occurred to me that if you reverse-engineered the Black-Scholes model, would everything wash out, and that 2.8% (or thereabouts), just represents the expected return on a "risk free" investment? Maybe with a bit of "alpha" thrown in for good measure?

Not that selling options is risk-free of course, but the long term average return should start to approximate something closer to that (plus "alpha"?), I would think. No free lunch and all.

-ERD50

[dixonge]

Quote:

Originally Posted by ERD50 I think you answered your own question.

If roulette had red or black strings of 37 in a row, I'd think the casino would replace the wheel if not the whole table. I've never heard of anything like that happening, although there are some interesting roulette wheel tales out there:

Gambling's Greatest Wins, Runs, Records and Legends - Part 2

[ERD50]

Quote:

Originally Posted by dixonge If roulette had red or black strings of 37 in a row, I'd think the casino would replace the wheel if not the whole table. I've never heard of anything like that happening, although there are some interesting roulette wheel tales out there: Gambling's Greatest Wins, Runs, Records and Legends - Part 2

You are still not "getting" the odds versus payout thing, are you?

Getting 37 reds/blacks in a row is not comparable at ALL to your getting 37 winning spread trades in a row. Your "odds" of winning are ~ .88. The odds of winning red/black are about .47.

Lets see, .88^37, is about comparable to a string of six red or black in a row, and closer to five in a row if you didn't preselect red or black.

Now six red or black in a row is out there a bit, but it sure isn't a rarity. What is rare is the person who can make money off any slight imbalances in the system. What isn't rare is how the house, or the "market", adjusts accordingly.

When you hit 219 of these trades successfully in a row, call the Guinness people - we can document them right here - good luck!

-ERD50

[dixonge]

Quote:

Originally Posted by ERD50 What isn't rare is how the house, or the "market", adjusts accordingly.

Can you explain how the 'market' would adjust if I, an individual trader, or a group of traders even, managed to successfully make gains at a rate slightly higher than the market mean? How would they know, why would they care? Who is 'they' any way? Market makers? Brokers? Exchange seat owners? Do they have meetings where they tweak rules and rates to make sure no one beats the house? Does the market self-adjust?

[haha]

Quote:

Originally Posted by dixonge Can you explain how the 'market' would adjust if I, an individual trader, or a group of traders even, managed to successfully make gains at a rate slightly higher than the market mean? How would they know, why would they care? Who is 'they' any way? Market makers? Brokers? Exchange seat owners? Do they have meetings where they tweak rules and rates to make sure no one beats the house? Does the market self-adjust?

He doesn't mean that. In the aggregate and over time, practioners of your craft will eke out a small return, subject however to wide deviations.

You may be the winner, balanced by other losers. But best not to count on that.

BTW, this really is deep water for a non-swimmer.

Ha

[ERD50]

Quote:

Originally Posted by dixonge Does the market self-adjust?

Yes.

It isn't people sitting around "figuring this stuff out". It just happens.

Let's say the cost of raising beef goes up - the retail price must go up. Shoppers on average, will probably buy a bit more alternatives, chicken, pork, veggies, etc. It might be almost unconscious to them, sometimes they look at the price of beef/chicken, and choose chicken - and this gets swayed ever so slightly with the price increase. Things adjust further as demand for beef declines, and demand for alternatives rise - it hits an equilibrium.

The shoppers didn't band together to decide this, no one "tweaked the rules or set the rates", it just happens. That is what free markets are all about.

If credit spreads paid better risk adjusted returns than other investments, the market would jump on credit spreads and that would drive the premiums down, and we would be right back at equilibrium. It just happens. No conspiracy theory or smoke-filled rooms required.

As far as finding some advantage to take in a world of freely traded options - well, I like haha's perpetual motion machine analogy. Entertaining to think about, the proponents claim that others just are being "closed minded". But I think the chances of violating the laws of supply and demand are only slightly better than violating the laws of physics. And as close to zero to be impractical to profit from in any meaningful way. You can say I'm closed minded, but I'd say I have a basic understanding of physics and markets.

But it makes no difference what I think, if you are going to post your trades, we will see.

Suggested reading: Taleb's - The Black Swan good discussion of risk and luck, Taleb supposedly makes money by buying suspected miss-priced options. But he has banks of computers and PhDs and backing, and I understand he can go many months with nothing but losses, and then the occasional big win. But he is playing the opposite game from you - he takes lots of losing bets, and the occasional big win.

Also, Against the Gods: The Remarkable Story of Risk by Peter L. Bernstein - great history of risk, there was a time when the payouts on die (bones at the time) didn't match the odds. But that was a long time ago.

-ERD50

[FIRE'd@51]

Quote:

Originally Posted by ERD50 FIRE'd@51 - well, two questions now. 1) Do you think that the calculation is very sensitive to slight changes in those numbers - would you expect an average of 10 'snapshots' to be pretty close to the 2.8% number?

So long as the strikes are chosen such that they are the same number of sigmas in the standard normal distribution, the probabilities A,B, and C will be the same. The actual put spread premium will be less if the implied volatility is lower, but the distance between the strikes will be also, so the maximum loss will be reduced as well. I haven't worked it out, but my guess is that the expected return wouldn't change that much, if at all, although there may be an overall scaling factor. Of course, fixed costs (commissions, bid-ask spreads, etc) will matter more if put spread premiums are lower.

If I flip a coin 10 times, the expected result is 5 heads and 5 tails, but there is a significant probability of other outcomes with only 10 flips. For 100 flips I would get much closer to one-half heads. The same principle holds here. IIRC, the standard deviation of the expected result is reduced by the square root of the number of independent trials. So I would expect you would need a lot more than 10 trials to actually realize close to the 2.8% expected return with any degree of certainty.

Quote:

Originally Posted by ERD50 2) After typing my note to haha, it occurred to me that if you reverse-engineered the Black-Scholes model, would everything wash out, and that 2.8% (or thereabouts), just represents the expected return on a "risk free" investment? Maybe with a bit of "alpha" thrown in for good measure? Not that selling options is risk-free of course, but the long term average return should start to approximate something closer to that (plus "alpha"?), I would think. No free lunch and all.

The 2.8% is an expected "excess return" (not an expected alpha), since you are taking on extra risk to earn it if the options are efficiently priced. The margin set aside (which can be securities) will earn whatever it is invested in. If you post T-bills as the margin set aside, you would expect to earn the risk-free rate plus 2.8%. If it's a stock portfolio (e.g. shares of SPY), the 2.8% is added to the expected return of the portfolio. The only source of "alpha" here would be if the realized volatility turns out to be less than the implied volatility (i.e. the VIX), since this would mean you sold an overpriced put spread.

A true Black-Scholes hedge (with a dynamically adjusting stock position) theoretically would earn the risk-free rate, unless you sold an over-priced option (or bought an under-priced one). This would only be alpha if you could predict future realized volatility with any degree of success.

[dixonge]

Quote:

Originally Posted by haha BTW, this really is deep water for a non-swimmer. Ha

well at least you didn't lower yourself to a personal jab......

[dixonge]

Quote:

Originally Posted by ERD50 As far as finding some advantage to take in a world of freely traded options - well, I like haha's perpetual motion machine analogy. Entertaining to think about, the proponents claim that others just are being "closed minded". But I think the chances of violating the laws of supply and demand are only slightly better than violating the laws of physics.

Well I will leave the great debate RE: efficiency/randomness of markets to others. And yes, I'll update this thread as my trades progress.

[ERD50]

Quote:

Originally Posted by FIRE'd@51 . If it's a stock portfolio (e.g. shares of SPY), the 2.8% is added to the expected return of the portfolio.

Thanks FIRE'd@51, that is what I was really trying to get to, but I mangled my phrasing a bit (or a lot). A couple points boost to a portfolio over the long run would mean a lot to a retiree looking to draw 3-4% a year.

Quote:

well at least you didn't lower yourself to a personal jab......

You've been a good sport - thanks for that. I think that is the best way to learn. We are grown-ups, we can save the kid gloves for tea with the vicar.

But you did say you didn't have much knowledge in statistics, I think haha was just addressing that. And no, you don't need to know probability and stats to invest, but it can help get a base for expected outcomes.

So good luck, keep us posted, I suspect you will do well, just not as well as you are hoping for. Me, I'm happy with a 2% boost (or whatever).

-ERD50

[Hamlet]

I think this might be a good time to post Mike Caro's roulette strategy, which reduces the house edge to zero.

Poker1.com - Home of Mike Caro University of Poker, Gambling, and Life Strategy

[ERD50]

Quote:

Originally Posted by Hamlet I think this might be a good time to post Mike Caro's roulette strategy, which reduces the house edge to zero. Poker1.com - Home of Mike Caro University of Poker, Gambling, and Life Strategy

very good, thanks - ERD50

[FIRE'd@51]

Quote:

Originally Posted by ERD50 Thanks FIRE'd@51, that is what I was really trying to get to, but I mangled my phrasing a bit (or a lot). A couple points boost to a portfolio over the long run would mean a lot to a retiree looking to draw 3-4% a year.

While it sure would be nice, if you mean a point or two annually, it is probably not achievable on the whole portfolio without taking undue risk. Given the fact that there is approximately a 1 in 10 chance of losing the entire margin set aside, I find it very unlikely that any retiree would commit a large portion of his portfolio to this strategy. More likely, the exposure would be limited to something like 5% of the total portfolio, which would cut the annual expected excess return on the whole portfolio to 0.14% This is not insignificant, but nothing like the numbers dixonge is talking about achieving (but he is a still-working "young dreamer" who can afford to take on more risk than most of us). For someone withdrawing 80K from a 2 million dollar portfolio (a 4% withdrawal rate), the expected return enhancement by exposing 5% of his/her retirement portfolio to this strategy is 2.8K or a 3.5% enhancement to the withdrawal $ (essentially the inflation adjustment).

My guess is that investors who use this (and similar) option-writing strategies will implement them opportunistically, when they "feel" that they have a statistical edge, which increases the probability of landing in area A above the 88% I have calculated. Usually this is a result of a volatility bet (they think options are overvalued), or as dixonge believes, it is possible to identify support and resistance areas which aids in selecting the strike prices. As one who spent his professional career focusing on ways to control portfolio risk, I find options to be an extremely useful tool. With options one can tailor the risk profile of a portfolio in almost any way one wants. The non-linear payoff pattern of options allows one to eliminate whole regions of the outcome probaility distribution. But I do believe there is no free lunch. I have watched too many traders blow themselves up selling what they thought were overpriced options. As we have discussed above, when option sellers are wrong, they are really wrong. Perhaps a more prudent way to implement selling credit spreads is to sell call spreads (although this reduces the long-term expected return of a buy-and-hold equity portfolio). At least if you lose the maximum on the call spread, you are probably making large amounts of money elsewhere in the portfolio. The problem with put spreads is that the times you lose the maximum, your entire equity portfolio is probably tanking.

Another possible advantage of selling call spreads is that for equity options in general, the higher the strike price, the lower the implied volatility. So if you sell a call spread you are paying less for the call you buy (in terms of implied volatility) than the one you sell. This may actually provide some alpha over the long term. With puts it's just the opposite. The put you are buying is more expensive (again in terms of implied volatility) than the one you are selling.

[ERD50]

Quote:

Originally Posted by FIRE'd@51 While it sure would be nice, if you mean a point or two annually, it is probably not achievable on the whole portfolio without taking undue risk.

Sorry, FIRE'd@51, I didn't include enough information in that statement that I made.

I agree with you, doing it with credit spreads would run too much risk of blow-up since a loss is a big loss. A bad run can hurt too much.

What I was referring to was something more along the lines of the BMX and BMY strategies from an earlier post. Since I'm holding SPY in my portfolio anyway, can I earn another point or two a year be selling CCs? I made the leap (not the LEAP) that that ~ 2.8% number you calculated would roughly apply to most option selling constructs. In the case of BXM (selling ATM or just barely OTM CCs), they trade all the added gain for reduced volatility (at least that was the result in that time period). For BXY (2% OTM CCs) they took a little added gain and a bit less volatility reduction than BXM (more total return yet lower volatility than SPY alone). I'm shooting more towards adding gain w/o affecting volatility much at all (going a bit further out with the OTM calls).

Of course I will get capped from time to time ( I did last April ), and this "opportunity cost" can wipe out many months of gains (but left me still slightly ahead), and I've already crawled further ahead in May/June and (so far) July. And when I say "gains" I mean "gains above/beyond what I get from holding the index itself". That is another difference with the credit spread - the seller of those spreads is not partaking in the market, which can be a good or bad thing, but the market has been good this past Q.

I should try your same calculations on my OTM calls - I'm generally shooting for whatever will get me ~ 0.75%/month premium, so my OTM% automatically moves in/out with VIX. My not-so-well-constructed back test of this seems to come up with giving roughly half of that 8% annual premium back to capped gains. But I'm really going more on my gut feel that the option seller is taking some risk (the cap, and has to hold the underlying in this case), he has to get rewarded for that by those hoping to profit with nothing more at risk than their "bet". In effect, I am "the house". In this case, I don't expect the house to have a big advantage, but I'll take a small one.

-ERD50

[FIRE'd@51]

ERD50

The expected return on a long call is positive (because SPY has a positive expected return), so if you follow a covered-write strategy over the long-term I would expect it would underperform a B&H theoretically. I think we had a thread on the BXM a couple of years ago, where we discussed a paper which showed a backtest where BXM actually outperformed the S&P 500. IIRC, I think the implied volatlity of the calls sold actually turned out to be 1-2 percentage points above the subsequent realized volatility over the backtest period, and this pretty much accounted for the outperformance (i.e., the calls sold were overpriced), but I don't remember the details and will have to go back and reread the paper.

I just plugged some numbers into my Black-Scholes spreadsheet, and it looks like a 30-day call 9% OTM at the current VIX level of 30% would generate a premium of 0.75%. A quick "back-of-the-envelope" calculation using the methodology in your post #89 gives a 30-day volatility of 8.66%, so having a one-month 9% OTM call exercised against you would be about a one sigma event (16% probability).

[ziggy29]

Quote:

Originally Posted by FIRE'd@51 The expected return on a long call is positive (because SPY has a positive expected return)

Seems to me that the positive expected return has to exceed the cost of the call option PLUS the transaction costs in order to generate a positive ROI.

Once you add in the price of the option and the commission, I'm not sure it's a sure thing that the average-case expected return is positive.

[FIRE'd@51]

Quote:

Originally Posted by ziggy29 Seems to me that the positive expected return has to exceed the cost of the call option PLUS the transaction costs in order to generate a positive ROI. Once you add in the price of the option and the commission, I'm not sure it's a sure thing that the average-case expected return is positive.

Unless you overpay for the option, the expected return will be positive.

In theory, you can calculate the expected price of the call at expiration by integrating the call payoff pattern over a lognormal stock price distribution with a mean equal to the expected price of the stock at expiration. If you have access to a Black-Scholes calculator, you can "trick" it into doing this calculation by setting the interest rate equal to zero and the stock price equal to its expected value at expiration. The call price it returns will be the expected price at expiration.

Remember, most of the time you will lose money on the call, but the effect of the tail is very significant in this calculation, and is what gives the call the positive expected return.

BTW, no one would buy a call if the expected return weren't positive.

[ERD50]

Quote:

Originally Posted by FIRE'd@51 ERD50 The expected return on a long call is positive (because SPY has a positive expected return),

I disagree with that, or I am misunderstanding what you are saying (or both ).

I would expect that the return on buying calls would be negative because of the time decay of the premium (which is why I sell them). Sure, the underlying is expected to have a positive bias, but the seller of the call is not going to give away that positive bias w/o some reward over and above just holding the underlying. What would motivate anyone to sell a call against their position, if they did not expect (on average) to gain from it?

I've always looked at call sellers as "the house" - the sellers have their capital invested, and have the downside risk, so they need some reward for selling that call. And I look at the buyers of calls as speculators/gamblers - they put up their bet and hope to win, but they have nothing more at stake than their bet. The chance of winning "big" is their implied reward.

I look at puts as insurance buyers/sellers - same story, sort of inverse.

from your next post:

Quote:

BTW, no one would buy a call if the expected return weren't positive.

People buy lottery tickets everyday, and the expected return is terrible. I have insurance, and I don't expect a positive return, I expect a reduction in risk. And I accept that the insurance cos can pool that risk and profit from it (they better, or they won't be there to pay my claim if/when I need it).

Or, another way to look at it, since options themselves are a zero-sum gain, both the buyer and seller can't win. Seems to me the seller needs to profit or they just would not play. There would be nothing in it for them. The buyers get the chance for a big win, or insurance - and they pay to play.

Is there something wrong in my thinking?


-ERD50

[FIRE'd@51]

I can only urge you to do the calculation to see for yourself (or use the Black-Scholes trick I posted above).

I believe you are mathematically inclined

<C> = integral from K to infinity of ( S - K) x F(S) x dS

where

<C> = expected price of call at expiration
S = stock price
K = strike price
F(S) = lognormal probability distribution

With regard to the market-maker, he is running a large delta book with lots of long and short option positions (many of them matched), and a small residual stock position (may be long or short) to keep the hedge ratio of the book near zero. He is not worried about any one single position, but of the overall book.

I am taking a break to go watch the ball game - I will check back later

[ERD50]

Quote:

Originally Posted by FIRE'd@51 I can only urge you to do the calculation to see for yourself (or use the Black-Scholes trick I posted above).

I may take a stab at that later ( I gotta run also).

But I always feel most comfortable when several different views of a situation re-enforce each other. And I'm having trouble reconciling what the math might say with my analysis that the buyer would be motivated to accept an expected negative return, and the seller a positive return. And that closes the "zero sum game" loop of option sales.

OK a bit more.... sure, the combination of the underlying and the options can be a net positive rather than "zero-sum". But again, that implies the seller is willing to give up some of his gains - why would he do that w/o the expectation of a positive return? Math does not explain why people buy lottery tickets, we need to include human motivations into this.

-ERD50