01-11-2002, 08:05 PM
All math heads try to hedge changes in the underlying stock price, and in expected volatility - whatever that means. But what if we wanted to hedge against some sort of economic disaster?
The problem with buying stock-index puts, is that you lose as the market rises, and you lose as time passes. And, if the market does drop, you have already paid for fat tails, so you have to pick the exact bottom to exit to recoup this cost.
Moreover, near-the-money puts are mostly delta, and deep-in-the-money puts are mostly vega. So what you're holding changes as the market moves. So, if we buy near-the-money puts and prices rise, we gain vega as we move into the fat tails, but we get wrecked in delta. So if we start deep, we've already paid for the vega, and if we start shallow, to earn vega, we get delta, more or less.
The stock market rises as the positive-sum consensus for the economy rises, and falls as the positive-sum consensus falls. So is there a way we can bet on this, without simply selling short futures?
My hypothesis is that you could sell short futures, and hedge them by buying deep-in-the-money calls that are trading at intrinsic. Say the S&P is at 1200, you short 300 futures, buy 150 1000 calls, 100 950 calls, and 50 900 calls. If the market rises, or falls slowly, you break even. If the market crashes, you take off the position as each strike is crossed.
As your delta declines, and your futures become unhedged, your vega explodes. Suppose the market blasts right past your strikes, so that you're stuck unhedged. Unlike the long puts, the harder it is to time and unwind this strategy, the more money you make! The whole thing, so far as I can tell, is contingent on in-the-money calls not pricing in fat tails, more or less.
In other words, we are short stock, by being short futures. And we are long stock by being long in-the-money calls. So far as delta, we are hedged. But so far as vega, which is correlated to the delta in the stocks - because both are inversely proprtional to P or the positive sum of transactions - we end up hedging or betting on a liquidity crisis twice, both by being short stock and long volatility.
As stocks go down, calls do down but, unlike puts, they go up more! We have bought rights to lean on people and hedged against a glut of people to lean on at the same time. The long calls win whether there are too many people to absorb risk and stocks rise, are there are too few. And then we use the short futures to hedge the calls' delta, which we cannot do by buying puts and buying stock, because, well figure it out for yourself!
So, what have I missed, and why will or will not this strategy make money?
eLROY
P.S. Dr. Bill, that was a nice trade in an otherwise choppy S&P the other day. Also, the reason I wrote such abtsract posts is because I WAS very busy at the time, and did not have time to come up with something focused, relevant, or novel, that's just the junk that bounces around in the back of my brain effortlessly at all times - and sometimes comes out.
But so far as your feel trades, you stated the problem perfectly. You saw a great trade, and yet talked yourself out of it. Why? Because you could not quantify it for yourself in black and white. I see this a lot in poker where, because a situation is a tough call, and the multiple variables are hard to quantify and weigh in the time alloted, you end up leaning towards doing what you feel like doing.
Rather than calling or folding an eqaul number of times, more often than not you use it as an excuse to call. The risk you avoid his how angry you will get at yourself, and how much confidence you will lose, if you fold incorrectly and lose the whole pot, rather than call incorrectly and lose a single bet. So you harness the gray-ness of the situation to become what you called a "hope trader." Maybe not you, but people.
In the market, you use your doubt as an excuse to stay out if you are out, and in if you are in. Then you always 1) think you were right, and 2) get frustrated. If you think about buying, but don't and then it rises, you say, well I was right to think it would rise. Or if instead it goes down, you say see, I was smart to smell trouble, and pass that one up.
And what allows most people to play these games with themselves is that they cannot quantify what they felt in retrospect, they are free to say they saw or felt whatever they want and, as such, they aren't really testing their assumptions! The never learn their strategy was wrong from a trade they didn't enter, instead, they learn to second-guess their feelings if it goes good, or either or, their ego is tied up in it too much, they elude themselves and refuse to be pinned down, to be wrong.
In other words, if you have a feel to start with - and some people do (I do but only when I start out broke) - you're fine. But if your feel stops working, or you start losing, and you can't even quantify what you've been doing, most poeple are screwed.
And most people don't have time to start out on the floor and load years of abstract observations and experiences into ther brain. All thay have is a chart - a fairly new and strange oddity to work with - and they can't afford to lose until they "get the hang of it."
So another beauty of a quantifiable strategy is that it can be discussed, and debated, and picked apart, and taught, like a religion. It doesn't just force people to sink or swim on their own instinct. And I really feel that one purpose of the two-plus-two sites is to take losers and turn them into winners fast, we don't get too many naturals here!
Also, this brings us to some issues between what is known as purely mechanical or "systems" trading, and what is known as "discretionary" trading. Let's look at them in the context of Jesse Livermore's two key factors in trading success, 1) sensitiveness to mob psychology, and 2) willingness to take a loss.
So far as sensitiveness to mob psychology, many traders will tell you that there is nothing they can add to a predefined computer algorithm, by overriding it with their own discretion. But even though they could, in theory, follow the same rules the computer follows without using a computer, what the rigid computer program offers that they cannot, is willingness to take a loss.
But I think this is just a matter of repetitions. When someone trading 10 instruments makes an average of four trades a year in each, for 10 years, he has what, 200 chances to learn to take a loss? You can get that many repetitions playing poker in one night.
With profitability dependent purely on how disinterestedly you muck losing hands no matter how much you have already invested in them, and on how fearlessly you raise winning hands even though you could just sit tight and still win, poker is often nothing more than a test of the ability to subvert your own natural tendencies. I don't see it as a real trading issue for poker players, but for traders, many need structure.
The next issue with mechanical systems is back-testing and confidence. The mob acts like the mob because they are free to act on hunches, which can steer you in different directions from moment to moment. To dampen your hunches, you freeze out action altogether.
Moreover, even if you have had steady success with a certain trading style over a period of years, it is hard to say that you wouldn't have had more success if you did something slightly different. Or you start to think, well, maybe things have changed. Without hard numbers, and a history, there really is no telling what you might do!
The only way many people are able to override their own instinct, and break this cycle, is when they know exactly how much money someone can expect to make or lose over the long run, doing one thing versus another. And you cannot have this much confidence in your judgment, not unless you are able to reduce it to an algorithm, program it into a computer, and back-test it over a period of years.
I think it is just silly, to look at a chart, or a string of profits, and say "I know what works" if you can't back-test it. It may work - but you won't - not unless you have cut-and-dry numbers to weigh, with a cool head. Or at least somebody might, but I wouldn't hope to be able to teach it.
The other main problem with computerzied - or at least rigidly structured - trading, comes about as a result of shortcomings in attempts to predefine and program what the eye can see. Given the limited ability of computers to associate as many different super-groupings of points as abstractly as the human eye, and given that larger "complete" or big-picture collections give you smaller statistical samples anyway, you are almost forced to use "discretion" in many situations.
But please, don't stop telling us about your trades! Rather, help us out by describing what you see, if you can!
The problem with buying stock-index puts, is that you lose as the market rises, and you lose as time passes. And, if the market does drop, you have already paid for fat tails, so you have to pick the exact bottom to exit to recoup this cost.
Moreover, near-the-money puts are mostly delta, and deep-in-the-money puts are mostly vega. So what you're holding changes as the market moves. So, if we buy near-the-money puts and prices rise, we gain vega as we move into the fat tails, but we get wrecked in delta. So if we start deep, we've already paid for the vega, and if we start shallow, to earn vega, we get delta, more or less.
The stock market rises as the positive-sum consensus for the economy rises, and falls as the positive-sum consensus falls. So is there a way we can bet on this, without simply selling short futures?
My hypothesis is that you could sell short futures, and hedge them by buying deep-in-the-money calls that are trading at intrinsic. Say the S&P is at 1200, you short 300 futures, buy 150 1000 calls, 100 950 calls, and 50 900 calls. If the market rises, or falls slowly, you break even. If the market crashes, you take off the position as each strike is crossed.
As your delta declines, and your futures become unhedged, your vega explodes. Suppose the market blasts right past your strikes, so that you're stuck unhedged. Unlike the long puts, the harder it is to time and unwind this strategy, the more money you make! The whole thing, so far as I can tell, is contingent on in-the-money calls not pricing in fat tails, more or less.
In other words, we are short stock, by being short futures. And we are long stock by being long in-the-money calls. So far as delta, we are hedged. But so far as vega, which is correlated to the delta in the stocks - because both are inversely proprtional to P or the positive sum of transactions - we end up hedging or betting on a liquidity crisis twice, both by being short stock and long volatility.
As stocks go down, calls do down but, unlike puts, they go up more! We have bought rights to lean on people and hedged against a glut of people to lean on at the same time. The long calls win whether there are too many people to absorb risk and stocks rise, are there are too few. And then we use the short futures to hedge the calls' delta, which we cannot do by buying puts and buying stock, because, well figure it out for yourself!
So, what have I missed, and why will or will not this strategy make money?
eLROY
P.S. Dr. Bill, that was a nice trade in an otherwise choppy S&P the other day. Also, the reason I wrote such abtsract posts is because I WAS very busy at the time, and did not have time to come up with something focused, relevant, or novel, that's just the junk that bounces around in the back of my brain effortlessly at all times - and sometimes comes out.
But so far as your feel trades, you stated the problem perfectly. You saw a great trade, and yet talked yourself out of it. Why? Because you could not quantify it for yourself in black and white. I see this a lot in poker where, because a situation is a tough call, and the multiple variables are hard to quantify and weigh in the time alloted, you end up leaning towards doing what you feel like doing.
Rather than calling or folding an eqaul number of times, more often than not you use it as an excuse to call. The risk you avoid his how angry you will get at yourself, and how much confidence you will lose, if you fold incorrectly and lose the whole pot, rather than call incorrectly and lose a single bet. So you harness the gray-ness of the situation to become what you called a "hope trader." Maybe not you, but people.
In the market, you use your doubt as an excuse to stay out if you are out, and in if you are in. Then you always 1) think you were right, and 2) get frustrated. If you think about buying, but don't and then it rises, you say, well I was right to think it would rise. Or if instead it goes down, you say see, I was smart to smell trouble, and pass that one up.
And what allows most people to play these games with themselves is that they cannot quantify what they felt in retrospect, they are free to say they saw or felt whatever they want and, as such, they aren't really testing their assumptions! The never learn their strategy was wrong from a trade they didn't enter, instead, they learn to second-guess their feelings if it goes good, or either or, their ego is tied up in it too much, they elude themselves and refuse to be pinned down, to be wrong.
In other words, if you have a feel to start with - and some people do (I do but only when I start out broke) - you're fine. But if your feel stops working, or you start losing, and you can't even quantify what you've been doing, most poeple are screwed.
And most people don't have time to start out on the floor and load years of abstract observations and experiences into ther brain. All thay have is a chart - a fairly new and strange oddity to work with - and they can't afford to lose until they "get the hang of it."
So another beauty of a quantifiable strategy is that it can be discussed, and debated, and picked apart, and taught, like a religion. It doesn't just force people to sink or swim on their own instinct. And I really feel that one purpose of the two-plus-two sites is to take losers and turn them into winners fast, we don't get too many naturals here!
Also, this brings us to some issues between what is known as purely mechanical or "systems" trading, and what is known as "discretionary" trading. Let's look at them in the context of Jesse Livermore's two key factors in trading success, 1) sensitiveness to mob psychology, and 2) willingness to take a loss.
So far as sensitiveness to mob psychology, many traders will tell you that there is nothing they can add to a predefined computer algorithm, by overriding it with their own discretion. But even though they could, in theory, follow the same rules the computer follows without using a computer, what the rigid computer program offers that they cannot, is willingness to take a loss.
But I think this is just a matter of repetitions. When someone trading 10 instruments makes an average of four trades a year in each, for 10 years, he has what, 200 chances to learn to take a loss? You can get that many repetitions playing poker in one night.
With profitability dependent purely on how disinterestedly you muck losing hands no matter how much you have already invested in them, and on how fearlessly you raise winning hands even though you could just sit tight and still win, poker is often nothing more than a test of the ability to subvert your own natural tendencies. I don't see it as a real trading issue for poker players, but for traders, many need structure.
The next issue with mechanical systems is back-testing and confidence. The mob acts like the mob because they are free to act on hunches, which can steer you in different directions from moment to moment. To dampen your hunches, you freeze out action altogether.
Moreover, even if you have had steady success with a certain trading style over a period of years, it is hard to say that you wouldn't have had more success if you did something slightly different. Or you start to think, well, maybe things have changed. Without hard numbers, and a history, there really is no telling what you might do!
The only way many people are able to override their own instinct, and break this cycle, is when they know exactly how much money someone can expect to make or lose over the long run, doing one thing versus another. And you cannot have this much confidence in your judgment, not unless you are able to reduce it to an algorithm, program it into a computer, and back-test it over a period of years.
I think it is just silly, to look at a chart, or a string of profits, and say "I know what works" if you can't back-test it. It may work - but you won't - not unless you have cut-and-dry numbers to weigh, with a cool head. Or at least somebody might, but I wouldn't hope to be able to teach it.
The other main problem with computerzied - or at least rigidly structured - trading, comes about as a result of shortcomings in attempts to predefine and program what the eye can see. Given the limited ability of computers to associate as many different super-groupings of points as abstractly as the human eye, and given that larger "complete" or big-picture collections give you smaller statistical samples anyway, you are almost forced to use "discretion" in many situations.
But please, don't stop telling us about your trades! Rather, help us out by describing what you see, if you can!