Market vs. limit orders

Once you find one paper, it's easy to find several more. Here are a few links:


http://www.sec.gov/rules/other/3437302.shtml


http://www.nyse.com/about/NT0001875A.html


http://www.olin.wustl.edu/faculty/liuh/Papers/Limit_Market.pdf


The rough consensus of the papers I've found is that on average, you get better execution with limit orders than with market orders. However, there are exceptions (in particular, if you are trading on very short horizon information), and there are possible weaknesses to each study, which I'm guessing eLROY can elaborate on.

Having literally spent years looking at every new piece of garbage that has come out of academia, or out of the NYSE (who fired their chief economist after decimalization), I will not even look at these studies. I would rather read the ingredients on a cereal box.


It is a guarantee that they are wrong, that they have missed some critical detail. Every paper is wrong. They're all wrong. Can I say that again? The people who write these papers are not even good economists, and many of them have never actually placed a market or a limit order. They begin out of college with the assumption "the stock market is efficient and can't be beat" so why would they even risk a dime?


There is, actually, a predominant common error they all share, but I am trying to remember what it is, rather than having to re-read them to remind myself. Don't be surprised if it is so simple that they limit their studies to orders which get filled. I'm serious!


The truth is, I doubt they can actually even do a "study." Are they really going to get some guy at Fidelity Magellan, or anywhere on down, to roll some dice before he decides whether to use a market or a limit, and where to place the limit relative to the last tick, or when to fire the market? I don't even think they could raise the money to do this experiment with just 1,000-lots, which would be irrelevant to their customers anyway. And how would they construct a control group?


That is not to completely rule out the possibility that, in some market microstructures, with some size orders, limit orders will be used more often in a profitable strategy which chooses which to use based on various factors in a given situation. It is also certain that the stock market will either go up or down today, and that the planet is getting either warmer or cooler. But even with 50% odds, academics, and the global warming/cooling crowd, seem to find a way to be wrong.


leroy


P.S. Okay, I caved and glanced at the unfamiliar Austin/Olin study, and it is gibberish. They're not even talking about the stock market, they're talking about some abstract situation in some abtstract model of the stock market. The first thing they need to do is prove to me how their notion of the stock market has any relavance to the actual stock market, as they don't suggest the appearance of having done much trading.


Okay, I waded through some more and found this quotation (I couldn't help myself):


"As a result, when a specialist changes her quotes it might be hard for other market participants to infer whether the quotes changed due to the arrival of new information or to inventory control considerations on her own behalf."


Oh, look! I hadn't even noticed how arrogant these pricks are to specify the specialist is a girl by using "her" rather than using "his" which does not suggest the specialist was either a man or woman any more than a flock of geese is presumed not to contain any ganders. Oh, but they're just better than the rest of us. And, moreover, not only when we used "he" were we implying the specialist was a man, we were also wrong. Thank Heaven we have these asses here to correct us.


But the real problem with this quotation is their original assumption of a symmetric fair value, or a "fair" price, which would be uniform across all participants if they had identical information. In fact, asymmetric utility, and asymmetric information costs, is the whole reason stocks, or anything else, exist. What they're saying, I think, is did the specialist raise his bid because earnings prospects for the company had increased, or because he found himself short an uncomfortable amount, or because there were buyers?


The reason the specialist raised his bid was to reflect the presence of nearby buyers, whether they included him or somebody else, and regardless of why they were buying. People don't buy because a stock is below fair value, rather fair value can only be arrived at after seeing who wants to buy - which can never be discovered perfectly.


But this type of thinking is so mushy-headed I do not even wish to continue. The entire academic model of the stock market is, understandably, wrong. I am apparently the world's greatest financial economist, or at least that is what I am beginning to think. So if you want to know anything, and I can't give you an answer, you're better off assuming the answer is unknown.

I have been trying to drag through the incredibly-boring "Atlas Shrugged" by Ayn Rand these last few weeks. It is Alan Greenspan's favorite book.


And yet, while The Fountainhead may have at least had an interesting plot, the popular economics - which is what Rand's books are really celebrated for - are all wrong.


I love religion, and I love Jesus. I hate the Taliban, and believe they should suffer the realities of Darwinism, and yet at the same time I fully appreciate their necessity for annihiliating those giant Hindu(?) idols. It was one of the few "right" things they did - with all rightness being measured in terms of whether it increased their survival odds as part of a realistically-textured culture or strategy.


But this belief that we must choose between "reason" and the unknown, or between what is "rational" and things that don't work, is utterly preposterous. Human knowledge and behavior is not the product of reason but of evolution, and history is the only laboratory that can prove any complex phenomenon.


The point being, not only are these economists wrong, but whether they are "right" or "wrong" has no meaning in any context. Their research is utterly worthless. They CANNOT know and, yet, nobody needs for them to.


leroy

BUILT TO BE WRONG


Looking at all these dot-com blowouts, and watching the stock market slide each day - all these silly, puffed-up ideas and complete idiot business models -you might start to wonder, is every single person on Earth a total moron, are they ever going to get it right? But then you remember that the human brain was never designed to figure these things out - to get them right in the first place - but merely to be a vessel for religion and imititation.


For example, suppose in the earliest stages of human upbringing, you tell a child to wash his hands before dinner. Are you going to explain to him the growth and survival patterns of tiny microscopic organisms called bacteria - which somehow the whole human race survived without knowing for thousands of years - and how our daily demographic migration patterns cause bacteria to become distributed on various surfaces you're statistically likley to pick up a bad trace of now and then, and how when transferred to food and left to culture under certain chance temperature and other conditions, this bacteria can then be transferred back to your body where certain tissues are susceptible to it? Is the idiot kid going to figure this out for himself?? No!! The reason the child washes his hands before coming to dinner is because if he doesn't, his mother is going to smack him!


So you get all these idiot ideas. Why does the Federal Reserve lower interest rates? Does it have something to do with evolved inventory decisions to denominate excess personal claims on spoiling assets in cash creating a self-perpetuating price signal masking true investment demand?? Huh??? No, of course everyone knows the reason the Fed eases rates is "to stimulate the economy." What is the number one criterion dictating the potential success or failure of a new B2B e-commerce exchange on the Internet? Oh wait, I know this one... it's... "Critical Mass!" Why shalt thou not steal? Does it have something to do with inducing catallactic cooperation incentives and creating a spontaneous price-implicit need-transmission network between total strangers dissipating geographic entropy in an extended order of non-tribal human interaction? Perhaps... But, barring that, the main reason we don't steal is because we'll go to jail first and hell second.


So, of course, given that it would be totally unfeasible to run a society - whether made up of men or dogs - on actual reasons, the human brain is designed to seize on all these what you might call substitute reasons, these simplifications which - whether you label them "religion" or "nonsense" - do the job. All people need to know is that if you act a certain way you'll get a certain result or, even better, if you do this the outcome will be good, if you do that the outcome will be bad, never mind why, or even what the specific outcome will be!


But of course some of the ideas will be right, and some of the ideas - no matter how popular or reasonable-sounding to the broad masses of society - will be wrong. And the only way to separate out the rigth from the wrong - given that even the most brilliant scientists and economists are too stupid to figure it out for themselves - is by evolution. Over a period of years natural selection will inexorably kill off the people who carry the wrong ideas, and slowly replace them with the people and businesses that, completely by chance, hold the right ones - without any central figure in Washington or elsewhere needing to look out across the landscape and declare he knows what works or why.


You say how can a complex economy function without anybody, not even the smartest among us, having even the slightest clue what is going on? How can the New York Times be filled from cover to cover each day with absolute bilge, the basest of nonsense, and none of the silliness which people latch onto like brain chewing gum having even the slightest basis in reality? (And remember, isolated facts, like the President's age, are merely molecules, the magic only happens when people piece together a random pieces of undeniable plastic and aluminim to hallucinate a car where there may be none.) And yet out of this fantasy land an actual paper does arrive at our door each day - you can touch it - and we don't all starve on hallucinated nonsense that merely fulfills our grossly simplified notion of what food is supposed to look and taste like?


Well of course it cannot be a requirement of complex ecosystems, whether composed of single-celled organisms, forest animals, or human beings, that the participants have any idea what is going on for them to function. Of course since animals can't have ideas, the functioning of the systems they operate in has to be not just idea-resistant, but completely idea-independent - which leaves people free to think whatever they want. Individually we may think and eat garbage, and individually we may perish, but the system is designed to work through all that, to pick us up and then throw us away like spent fuel rods, or like unwitting slaves in some sci-fi-nightmare/brave-new-world idea hatchery. Experiencing no cost for individual lives on the total number, the system wants us to think nonsense, experiment, and die as a sacrifice in the name of progress.


For the price system to function across kingdoms and oceans in medieveal Europe, for example, the last thing that was needed was a bunch of pointy-heads like me who understand the intricacies of non-redundant capacity and consumption signals in the supply chain. Quite the opposite, all that was required was a bunch of loudmouth, simpleminded merchants who traveled a lot, talked to a lot of people wherever they traveled, remembered what all the people in the different places told them, empathized with their needs and haves, and tallied them against one another - and had the political instincts to move between cultures. The trade routes were a primitive fiber-optic network, only better, because of a number of esoteric non-redundant inter-coordination problems, satisfied only by geographic friction, but which nobody will ever understand or - need to understand!


You might say the problem with a price and production system run entirely by people who, by necessity, haven't the faintest clue how it works is that, without knowing what they are doing, your average smart person will be inclined to shoot the middleman when prices go up, attempt to abolish the price system and replace it with a more rational set of incentives - created by good, thinking people coming together and that we can all agree on -or even try to cook up some new sort of marketplace or B2B supply chain completely from scratch on the Internet! But the beauty of the non-reality system we have in place is that people can be that dumb and, thanks to evolution, the problem will take care of itself. No one need ever get a clue!


By this rule, success in business, for instance, does not come from figuring anything out or from building a better mousetrap, as the saying goes. Rather, it comes from how well you are tuned into the nonsense-network. If two different companies are producing two different widgets, it is impossible for anyone to have any idea which one will work better or why. What is possible, and what is relevant, is how quickly you can learn what is the right kind of widget to build, and whether you have enough contacts at whatever turns out to be the good-widget factory, to get a job doing whatever they tell you has been discovered to work, before the fad climaxes and dies out. At no time is it necessary, or even possible, to figure anything out, not unless you want to start a new wave of lab rats, or even risk living and dying as a lab-rat yourself.


So, given that the real living thing is the collective culture, and the system into which this culture arranges people, whatever they may think, how is it that individuals allow themselves to be made into suckers, or fall prey as rats in the lab of untested ideas? At the heart of it is a culture of left-win arrogance - the idea that man can think and can shape his own world - which you find in abundance in university communities like Cambridge and Palo Alto. Ideas are incubated in self-insulating intellectual communities like the academic peer-review circle, among people who, having resigned themselves to a certain lifestyle and secured it, pay no price for being wrong. (Idle-rich housewives gossiping in salons enjoy a similarly rare freedom to venture far and wide from the moorings of reality in their musings, as well as people all over who believe something is true just because the right people think it is, and who safely voted for Bill Clinton because the imperceptible feedback to the cost in their indiviudal lives is born collectively, and is too remote and abstract to shape individual behaviors in a single country or generation).


The first layer of erecting the castle in the air that is the new idea is the existing set of substitute or nonsense ideas - ideas such as wash your hands or thou shalt not steal - that were never meant to be dissected and recombined and used as building blocks in newer, more complex notions, by unbridled "smart" people. Or at least the seemingly logical building blocks (where the logic is merely an illusion or a substitute idea) were never tested for efficacy in these new combinations or applications, but that is what is endeavored to be done. We all know 5 + 5 = 3, the Fed will stimulate the economy, and non-farmers who do no "real" phsyical labor will exploit us, so it's a foregone conclusion what will happen if we add a third five!


Since the original belief system is not what is real but, by necessity, what is plausible in order to get people to behave a certain way, the first test the new idea faces is plausibility. In fact, at this stage the idea is just like a virus whose only necessary trait is plausibility, given the existing set of nonsense assumptions that do work, and it need not have any basis in reality. All the idea needs is an arrogant Ivy-League professor, a straight-A student capable of absorbing it without question - even understanding it's abstract "logic" - and then an established or highly-evolved, successful business for the student to go infect when he gets a good job out of college!


So the idea-virus spreads and becomes more popular depending on how plausible it is and how well it resonates with the existing set of nonsense-beliefs, which may themelves be too new to be tested by anything but plausibility! It is at this point where the lab rats - the people who actually pay a price for being wrong in the real world - get swept up in it, to learn if it is a good or bad virus-idea. And as large groups of people start behaving differently, of course teh texture of society, in which pre-existing behaviors must continue to operate, is thrown into a constant flux and turmoil, so that yet another third behavior may be put on top, but only so long as some fourth behavior, whihc is doomed, persists. So this may self-perpetuate for a time, to some sort of a social climax or revolution, followed by a decay, a collapse, and a return to previous norms.


It is important, getting back to the beginning, to remember that people have no idea what is going on, rather their brains are built primarily for the adoption, transmission, and even defense by fire against competing idea carriers, of bad ideas to which they are slaves.


When you bring home a new electrical appliance, you can probably figure out how to put the batteries in without needing to assimilate a copy of the instructions - which assimilable instruction-teachings someone else figured out by trial and error, and then made a million copies of for a million different people in potentially identical situtaions. It is even quite possible that, if you had never even experienced a battery or an electronic appliance before in your life, you might discard the two into a chimpanzee's cage for lack of any idea how they were connected or what they could be combined to produce. Then that dumb, fidgeting ape might even get the positives and negatives lined up right in a few days, then other chimps could copy him and get the same result. Lord only knows what idiot ideas they might seize on for what the thing is or why it does what it does, only when it started makign noises it would probably play their instincts for a sucker by convinving them there must be a baby chimp inside the box.


But again, the point of the brain is not to figure things out, but merely to transmit them, through the written and spoken word once they have already been figured out, and then to remember them. Because with a million idiots plodding along in their own little fantasy lands, stumbling on things that work is inevitable (and then you do more of what you're doing, have lots of kids, and write lots of magazine articles). A single human brain, in a single lifetime, no matter if it were as big as a house, could never hope to compete with a thousand imitating zombies experimenting at random over a thousand years, in terms of its ability to figure things out and, fortunately, it doesn't need to.


It is a very broad misconception that, even in every day life, people are thinking. At most, they are comparing past experiences, or series of events in time translated into a loose network of summarizing or representative symbols, to the present situation, and choosing which past experience or combination the present experience is most like. Or they are trying to fit a situation to a certain label, so they can know which rule, learned from their own expereince or that of others taught them, to apply, or to communicate the perception of the situation with others and elicit a coordinated pattern of pre-conditioned behaviors. Beyond that, a person is just an evolved stimulus-response catalog of varying sorting precision and complexity. Moreover, since a person cannot shape the events that go into his catalog, he cannot, as is popularly believed, think and reason outside this catalog.


Of course there is some exception to this rule, based on the geometric increase in complexity between the very local environment and the very macro environment in which the individual must operate simultaneously. And calling this difference in complexity geometric is a geometric understatement. In very small systems, a sufficient number of catalog experiences may present a discernible pattern of rules, such that whatever visual or logical events may occur, they may be foreseen by the individual despite having never been experienced by anyone who could have possibly learned a response behavior to teach him! So sure, somebody might be able to get something right right under his nose, and so long as it involves only one or two people, whom he knows very well, interacting, and not a lot of math. There is an equlibrium in the complexity and size of the moral code, and how specific it needs to get, Deuteronomy vs. the New Covenant and so forth - but nowadays we just teach everyoenb how to read and let them figure that otu for themselves. Things that occur many times in a lifetime, and don't cost you your life, can be left to be relearned by each subsequent generation.


Why somebody would want to listen to his college professor in fields other than sales, medicine and engineering, and learn some experimental new-fangled nonsense that will probably doom him to a life of poverty, when he could go right down the street to his neighborhood church or synangogue and learn something that's worked - and made people wealthier - for a thousand years, is beyond me. But I would imagine the level of suckers in a society, or the level of experimentation, operates at some magical, incomprehensible equilibrium too, there is an order to the madness. Certainly, in the larger catallaxy, 99% of what 99% of people think and do for a living is utterly useless. But who can claim to sort it out apart from history itself?


So, in conclusion regarding this whole so-called Internet Bubble, of course a bunch of smart college guys though up a new, non-evolved business topology, or a creature that had never been put to the test. The creature is only viable if sub-evolution within the creature is viable. The problem is that businesses which, as minds, cannot evolve time-series models in the new supply chain topology because there is no friction and the associations are not tree-shaped and friction-asymmetric. Huh? The keys to the maintenance of the micrtostructure of empty space (which I cut out of this already overlong essay), the keys to the stimulus-response in the human brain, the geographic friction within the brain itself, dividing the continuous gray matter into different super and sub-regions with distinct topology


enough...


loudmouth

In reality, whether limit or market orders work better is ultimately contingent on the behavior of other market participants. And everyone's game has leaks. (Also, notice that if everyone thought one were correct, the other would become correct.)


If everyone didn't have leaks, it would simply be a matter of cycling, where market orders worked only after the pool of people using them had been sufficiently weeded down, and the pool of limiters benefiting had been beefed up at the marketers' expense. In that simple model, the only real error would be your poor timing of the cycle, or of other people's timing of the timing of the evolutionary cycle.


But, as in poker, we know that the stock market is not purely Darwinian because it is not a closed system. The total number of players can rise over time, and there will always be new fish to replace the spent ones. And the existing fish have jobs, so they can resist evolution, and make the same mistakes over and over without dying.


So the question, given that whether limit or market orders work is a result of the stock market not being a closed, zero-sum game, is where does the new money come from. The new money, more often than not, comes from people using limit orders.


You just want to do the oppoiste of what the people making errors are likely to do. People using market orders, with nothing else to go on, are less likley to originate from the group making the most errors.


There will probably always be a hair too many people using limits.


leroy

Okay, so now you really know how to get under my skin!


Let's look a little deeper at this use of the word "hers" with regard to things pertaining to the NYSE specialist, as it may offer a shortcut into the flawed thinking under-cutting the whole paper. I have already pointed out several flaws in their use of the word "hers":


1. It assumes that, when other people throughout history have used the word his, they have been referring exclusively to a man - as opposed to the Latin neutral generalization. This shows little apart from the fact that they are suckers, and poor studies, but an easy sell.


2. It shows that they like to invent things, and that they have confidence in their ability to do so, because they are better than the rest of us. So they offer us new interpretations of the words "his" and "hers" that show how much more wisdom they have.


3. It assumes people who used "his" when referring to a NYSE specialist - and implying a male - were wrong, even though I have never seen or heard of a female NYSE specialist.


But there is an even more important underlying flaw than any of these! It is something like:


4. They believe that there is no economic basis for a specialist to be either male or female, and it is rather some groundless, self-fulfilling human bias which causes most specialists to be men. Some ancient, meaningless idea actually manifests in the physical world, and the idea itself perists like a perpetual-motion machine, depsite the fact that no idea can survive for a thousand years without having some reality to reinforce it. Therefore, by simply electing to use the word "her" when referring to an NYSE specialist, the free-floating idea, and therefore the physical conditions will be altered, causing there to be more female NYSE specialists in the future. In other words, there is no physical reality, except that which is true and manifests based on popular consensus (or "delusion").


This view of the world goes hand in hand with the idea that there is a universal "fair price" for a stock that can be symmetric across all participants. It is reflected in Burton Malkiel's "Firm Foundation" vs. "Castle In The Air" dichotomy (in that "Random Walk" book).


Rather than there being a flexible "Castle In The Air" which is symmetric across all participants, there is a single firm foundation which is nevertheless asymmetric(!) across all participants. In other words, people are not free to invent and interpret reality as they choose, and yet for many people there can be many different realities. Their idea is not unlike saying well, if special relativity says we all have different clocks, can't I pick my own clock? Isn't the only way a clock can exist is if we all agree on it? Isn't time, aren't fair prices socially constructed?


Aren't we free to think what we want, and the yardstick of accuracy is peer review?


No, no, no, no, no, no, and NO.


El-Roi

In that offensive Austin/Olin paper, they are bent on annoying me with this notion of a "true" price for a stock. But the true price is presumably the price that would be selected if "all" information were known - that is to say if the entire universe were compressed to a single point.


A game of craps, for instance, is intentionally random. If NASA built a table to exacting specifications, and a robotic arm to throw the dice, they could know how they would land. So we introduce random elements, or elements that are not known in advance, on purpose. Even if NASA did build the non-random dice thrower, some unseen factor, like a meteorite, could conceivably hit the system from the outside, and create an unexpected event. So, the only way to "know" how the dice will fall is to "know" the entire state of the universe. But the only object big enough to carry all that information is the universe itself (ignoring light cones and quantum trapdoors into nothing).


So, to "know" the true price of a stock is not unlike to "know" the true position of a rock. Not only is the rock at different positions relative to witnesses in different inertial frames of reference, but you can still only know it in a certain timeframe. You can say, well, I know all the rocks that will likely collide with it in the next five minutes, which will result in it being at location X. Or, if I figure in all the rocks I can expect to collide with it in the next day, I can say that will result in it being at position Y. But I can never know ALL the rocks that will collide with it between now and the end of time, so I have nothing to discount it from.


Except perhaps I can assume that EVERY rock will collide with it by the end of time, so it will have impacted and thereby changed the trajectory of the entire universe - as if it had infinite weight. Different stocks are worth different amounts to different people depending on their economic frame of reference, their asymmetric circumstances. If I were to really attempt to assume what would happen if the stock were made available to EVERY potential buyer by the end of time, it seems it would be infinite.


If there were no geographic friction in the universe, we could access deposits of precious metals on all planets at once without lifting a finger. All things would have infinite value, because of their infinte ability to react with one another, and the entire universe would dissipate in a single flash. Every thing would be adjacent to every other thing. Poof!


At the very least, the entropy value of all a corporation's assets and adjacencies are equal to the speed of light squared. Forgive my math, but the way I understand it, f = ma, and therefore the accumulated energy of a moving body can be described as e = ma-squared. Since there is no fixed ether, and the potential velocity relative to the remainder of the universe or some object therein is c, or the speed of light, therefore its energy = mc-squared, or something like that.


In other words, the "true" price is the price in the absence of geographic friction, or in the absence of reality. Therefore, the only reason a stock has some mundane value is because of geographic friction. Or something. There may even be people on other planets who have a value for buying Microsoft stock. But the actual price is the product of geographic friction.


But you begin to see the kinds of pitfalls you run into when you start making "assumptions" and building models of open systems - systems whose properties are a product of their very openness. The mathematics and the theory are beyond the intellect of anybody who is alive today or ever will be. So you can figure what would happen in an ideal model. But then you have to analyze the differences between the ideal model and reality to arrive at a conclusion.


You cannot construct an open model by fixing unknown variables.


I swear, I'll stop now.


loudmouth

"...there are possible weaknesses to each study, which I'm guessing eLROY can elaborate on.


Did you have to encourage him? /images/smile.gif


I haven't had time to do anything more than briefly scan through these particular studies, but I'm wondering how they figure the true cost of a missed fill. That to me is the key thing.


TRLS