Ok so what is the essence of trading ?
First, noob's right here so feel free to throw shit or give me some lessons. Can't tell of myself as having any real experience or knowledge in trading but here are what I know about types of trading:
- Arbitrage: Basically a mispricing of the same asset in different markets. I don't think this type of opportunity arises a lot in modern market and it probably requires lightning-speed technology to capture this profit which is unavailable to the laymen.
-
Gut-feeling speculation: You base your buy/sell on feeling and "analysis". However, in the world of finance, there's little to none objective truths (heard somebody said God himself don't know what a "true" P/E is). Say you're a fundie investor, a company is being traded at P/E = 200, all rational analysis points out that it is overpriced. Later the stock rises to P/E = 500 before collapsing but you're fucked already. On a smaller scale, I believe there's no way on Earth to tell whether a stock's P/E should be 26 or 27. Tech analysis is obscurant and probably half-true at best. Both camps are a little bit better than rolling the dice IMHO (though it's good if you have the privilege of gambling with others' money)
-
Insider trading: Based on tip by informants. Illegal so nuff said.
-
Market-making: Quote both a bid/ask price like the people at the money exchange counter. To do this, you must have a large stock of the asset and a steady flow of customer.
-
Stat arb, HFT ... : the vogue of trading these days. Done by computers at possibly very short intervals so no fundie analysis involved I believe. Relying on "regression to the mean" and "deviation from statistical history" but here's what is confusing me. I mean this is not even a true form of arbitrage, how do you rely on statistical and historical trend to consistently make money ? Also the issue of Black Swan (which is easily explained by human psychology IMHO, saying that human action is random and independent is like saying in a house fire, everyone follows a random and independent course of action, no, they're all rushing to the exit)
Ok guys, that's it. Let's pull up a seat and have a thoughtful discussion.
for me, essence of trading = making money.
I'm confused. are you asking what types of trading are out there, or which one makes most money? If former you've already listed them all more or less, and if latter it depends on what best fits you I guess
^ Okay so beside pure arbitrage and market making, all others seem to be arcane and frankly, a bit baseless and unreliable at least to me. And I'm confused at most at the stat arb/HFT route. Anybody willing to shed some light ?
Market makers don´t have a large stock of the asset. What happens basically is that they will have a short/long position after the trade. If it's a liquid enough asset they will do the opposite trade at a more favorable price and keep the difference. Ideally that would be the bid/ask spread, but in reality people won´t hit your bid and ask equally, so you move the spread accordingly with your view on the market and/or use derivatives. People who trade less liquid products hedge their risk using derivatives. But again, market makers do have positions and make money on them.
Thanks pal. I somewhat understand the use of derivatives to hedge risk. Would you mind explaining a bit more about "move the spread accordingly with your view on the market" ? Because as far as I know, any prediction on price movement is speculative at best. I know that traders could "read the tape" to predict very short-term movement but even this is not 100% reliable because people can jump in and out of the market at any points and there's a variety of dark pools available to mask their orders.
Imagine you think the market is going up. Therefore you want to be long, so you want people to hit your bid more than your offer. You can move your mid price up, so people can sell at a higher price to you, but buy more expensive, which will encourage people to just sell to you, therefore leaving you with a long position. You can also simply move your bid up, tightening the spread and having the same effect. You don´t necessarily need to be taking a view on the market to do this. Imagine you already have a big short position because of the trades you've done throughout the day, and you want to reduce your risk in that security, even if you think the market is going down. You'd do the same thing as above.
About the HFT taking over trading, it´s a discussion we've had many times. I think it will reduce the number of jobs in very liquid products, (say cash eq) but it will not make them disappear. No matter how good your model is, it can never trade on it´s own better than a good trader in difficult situations. You will always need human traders because weird completely unpredictable situations happen much more often than people would think. Ask John Meriwether and his friends at LTCM about what happens when you use statistical arbitrage without supervision. And it will never take over illiquid asset classes, which are responsible for a high % of trading jobs, simply because HFT is based on liquidity.
The true vogue of trading right now is HFT/ computer assisted trading but it isn't that statistical arbitrage. Stat arb is pretty vague (since clearly anyone buying or selling equity thinks they are involved in statistical arbitrage of some sort). For the true practitioners however, it is true that statistics are often thrown out of the window when human beings are involved on an individual level but as a group there are trends. Couldn't you predict what any particular clique in high school would be doing on a Friday night? Granted Jenny may be grounded or Bill might suddenly decide he wants to be the next Aristotle and choose to study but as a whole the group will continue with their process.
Data mining, neural network computing, and machine learning is where it's at in HFT/ computer assisted trading to me. Collect so much information it would crash a F500 server system and let the computer start to figure things out.
Thanks pal. 1. If HFT doesn't look like stat arb. What does it look like ? Scalping ? Market making ?
I love how thread is titled "essence of trading" and people are all over the place with their answers.
The true "essnece" of trading has nothing to do with finance or technology. Trading has existed for thousands of years before HFT began. Trading will continue to exist going forward, and it is likely that it will not have anything to do with finance as it does today.
Traders on the floor or in the pits today are no different than the traders in Venetian Ports or on the Silk Road.
The essence of trading is that it seeks to rectify the anomaly of the double coincidence of wants. For any tranasaction to occur, two parties must meet in the same location, at the same moment in time, and agree to the proper exchange of a particular set of goods or services. Of course, not every buyer or seller of a good is in the right place at the right time with the right terms. In fact, most trades are not.
Enter the trader. The trader seeks to rectify this anomaly by buying or selling goods at one point in time in one location, and then buying or selling them at another point in time or in another location. He risks his own capital to do so, but if he does it right, can either make a significant profit in the exchange, or lose a significant amount of money.
Up until the Industrial Revolution (lets say 1800), most trading was about taking advantage of spatial arbitrage opportunities. In other words, moving goods from one place to another. This brings me back to the traders I mentioned before. Venetian merchants, Silk Roadsmen in Central Asia, Slave Traders in Africa and the New World, the VOC and the British East India Company, the two largest coporations ever, who moved goods across the Earth to take advantage of these arbitrage opportunities.
From the Industrial Revolution onward, traders sought primarily chromatic arbitrage opportunities. These are differences in the price of an asset between two moments in time. They may have continued to make geographical trades, but they could make greater profits in time-denominated assets. This coincides with the founding of the NYSE. It was a time when men traded railroad bonds, equities, and eventually commodities and currencies, in real time. They sought to make money off the differences in prices between one point in time and another. The markets in financial derivatives are far larger than the markets in global trade, with interest rate derivatives reaching alomst 600 trillion in deals done.
The problem, of course, is that this age is coming to again. Innovation is no longer being driven by time based variables. Read this article for a nice primer. Extracting value through time isn't really profitable on a go forward basis.
In fact, the rise of HFT proves this. The idea behind HFT is that the only way to have any signficant edge in trading is to be able to do millions of small transactions for half a penny arbitrage opportunities. Once most commoditized markets, such as FX, bonds, equities, etc, move to this model, their will basically be no possible way to make any money in trading. Everything will be traded in real time for incredible small spreads, and almost all value that would typically be extracted by a trader will end up back in the hands of the actual producers.
Thus, the essence of trading going forward is to determine which variable (instead of space or time) we will base our economic system on, and then finding a way to take advantage of arbitrage opportunities in that.
I hope you aren't planning on a long-term career in trading, because almost all of it will be automated and their really won't be any money left to be made in it. The only areas of trading where their will be money to be made are in extremely sophisticated, structured transactions. This value of this type of trading, however, isn't related to a time-dependent variable, but instead comes from the superior knowledge required to create the trade, thus proving my point even further.
From the age of the Roman's to the beginning of the Industrial Revolution, trading was mostly done to
And by your own analysis how do you suppose there would be no money left in trading even stocks, the simplest of financial transactions. Just because people have gotten quicker doesn't change anything, these computers aren't voodoo machines they rely on human analysis. Assuming that stock trading will not be profitable is assuming that someone finds a way to model future human behavior or data mining techniques become so efficient that a computer can recognize only the news that would affect the market and also the magnitude of the news.You should try to learn to use the tools around you instead of saying that it is over because you refuse to learn.
First off dude, don't critize me like that. You obviously missed the point. My guess is that you don't really understand HFT or how it works.
HFT don't attempt in any way to model human behavior. They typically employ statistical arbitrage opportunities. For example, if a security is trading a certain deviation beyond its typical range, it will buy/sell really quickly in order to make a small profit. This is a time-dependent strategy. The program realizes that the security was trading outside of its typical range, and bought it at one time and sold it at another time in order to take advantage of this discrepancy.
Other strategies include traditional market-making. The HFT market maker of today is no different than the Silk Road trader of yesterday, as I mentioned in my previous response. Buy low or sell high in one place/time, sell high/buy low in another. Except HFT does this at a much faster pace, with a lower spread and higher volumes.
It isn't that HFT trading won't be profitable. It is that it will become commoditized, and thus won't provide the tremendous value opportunity that it does today. My father has worked in trading for 30 years, and he told me that his entire department will soon be replaced by a computer program.
In fact, nearly 70% of exchange trades today are HFT. As that continues to grow, the spreads will continue to narrow. Where an equity market-maker on the NYSE floor made a very large spread in the 20's, or a modest spread in the 80's, today the spread is basically non-existant. Remember, the spread is the difference in price over a period of time. My overall point is that the potential value of a unit of time is significantly decreasing. Where a unit of time would produce a very large value (because of a large spread) in the 20's, it hardly produces any value today.
Which brings me back to my original point, which is the same point made in the article I referenced in my previous post. Up until the industrial revolution, the greatest arbitrage opportunity was geographic. Moving goods from one part of the world to another allowed traders to make outsized profits. Note, this had very little to do with time. A trader could take 6 months or one year moving goods from China to England, or from Africa to US. It didn't matter to much about the duration of the trip. The fact that these regions were far enough away from each other allowed arbitrage opportunities to exist.
Then the industrial revolution came along, and time became a variable. Time became a potential source of value. Trains and steamships allowed goods to be moved fast. Traders now had to move goods quickly, as geographial arbitrage opportunities ceased to exist.
Move forward to today. Financial securities are basically time-dependent assets. A bond has a set duration, equities have quarterly dividends, futures and options contracts agree to certain settlement dates, etc. Thus, trading in any of these is a way to arbitrage the value of time.
HFT only exacerbates this. It basically arbitrages time instantly, to the point where the potential value of a unit of time is nothing. My point is that this proves that the value opportunity which existed in trading financial securities in the past will cease to exist, and very likely within the next few years. Trading won't go away, since it will be just as likely that whatever variable our economic system in the future is dependent on will also provide potential arbitrage opportunities. If you think this won't happen soon, you are absolutely wrong. You can say I am not willing to learn all you want, but the fact of the matter is I am looking at the big picture.
I thought this thread is going to evolve into another boring discussion of what-kind-of-trading & desk-is-the-best-for-frosh type of thing, but your input is quite interesting. Spatial vs. time arb is also something I've been thinking about for a while, although in a different context since I work in commodities. However your writing seems to end up rather abruptly. where's your last paragraph?
We're all over the place but you had to write a book to make your point. The guy wanted to know what trading is now and going forward. I don't think the point was to get an analysis on the history of trading, actually looks like the first time I've ever seen the site cut someone off mid rant.
Ok here is an example of what I thought could make statistical arbitrage through analysis of P/E ratios.
Argument: we can say that equities will be some distance from their true earnings and form an average value. We can then take an overall market analysis and say that other equities are some distance also from their average and recognize what kind of market situation we are in. In a period of recession logically you would expect that the prices would be closer to their earnings and of course the opposite in expansion. Check against other equities in the same sector and judge their distance from the mean to see if a bubble is forming. In my analysis recognizing that these equities have moved away from their average values would be positive and signal to buy but at some critical value recognize there will be mean reversion.
The explanation is simple but the issues that I'm sure you recognize are difficult to solve. Some look at how past behaviors of related businesses in markets that open before ours move relative to our markets and form an opinion. Regardless it always relies on recognizing something that's at least a little more than just looking at a straight up and down price chart but the idea is to have a simple idea that you can model and prove statistically. You want to be able to form a null hypothesis and have as few independent variables affecting the dependent as possible. It's probably not as complicated to model as you might think, the programming aspect to get to market might be though.
^ Thanks, bud. Sound like a sophisticated analysis of the buying/selling force to me. Maybe that's the basis of the system: that whatever buying/selling force is out there, they've been around for a while and will not change much esp. in short-term therefore you can expect some consistency and predictability in gaming the system. Trading these days seems to demand an insane amount of technical proficiency. Do you recommend any books or software tools to master ?
Depends on the route you want to go down. C++ is the normal language (but honestly you can learn python, C, Java, I saw in a post on here that Jane Street uses OCAML which is pretty random) that can build the software and stata is a powerful statistical analysis package. It's going to take a lot of determination to learn a programming language on your own. Taking one class in programming at a CC even helps you get comfortable with it though. The future though is data mining with AI and neural networks SQL is getting big and I plan on learning it.
There will always be trading between men with potential to profit as long as some once in history genius doesn't figure out how to predict human behavior. You can program in any language but SQL is probably the most universal for building data structures simple learn it. Learn to use your knowledge and all the tools around you and you'll eat liberal arts traders for breakfast. All systems automated, just means a level playing field; a more exciting chess match. Embrace it.
Okay bro, how about any book ? And somehow I hate the general tea-leaf reading tech analysis guide with a passion. I'm reading "Beat the market" by Ed Thorp but so far the guy was still rambling on and on about how he discovered a strategy similar to covered calls. I've got 101 course in C and SQL but I guess it's light years apart from how to put them into practice.
For the record, I know I'm taking this too literally. I was in the mood to organize some of my thoughts on the subject and I took the forum title literally. Obviously, OP is looking for good info on trading, but those questions get boring after a while and I wanted to stir the pot.
Also, a P/E of 200-500 is absurd. S&P trades around 12, maybe 18-20. 200 is double what some companies were valued at the top of the tech bubble.
Now I become an OP for the 1st time and see it myself how the whole topic morphed into a duel between 2 users (not necessary a bad thing)
I think I took that number from the valuation of Cisco at the height of the tech bubble. Might be inaccurate though.
No you aren't looking at the big picture, what you describe is one part of computer assisted trading. All that you are talking about is one aspect of a business model like Jane Street. It is the new model for market making. You can purchase an equity and sell it at basically the same price (only make .0001 cent profit) because the time you are exposed to the position is under nanoseconds.
Market Making is over of course in liquid securities as a real profit making tool and that is a good thing. I can assure you any time you are exposed to market forces whether it is for two weeks or a nanosecond you are never guaranteed things will work in your favor.
Maybe I didn't go into enough detail on that analysis to allow anyone to form a real argument but those times of unheard of values were the key to making money. You would realize a bubble was forming and ride in it until a critical value was reached at which point reward/risk
Essence of trading? Buy low and Sell high. BTW trading is not that all shining when comparing to some other business. Reason? Trading requires a seller and buyer, which means traders have to slick two asses ( one ass sometime ) and have to pay tax for protection. Now if you are interested in the more profitable business, I can tell you that No. 1 on that list is called "Revolution".
Est cupiditate veniam magnam eaque temporibus quia. Temporibus fugiat laborum facilis in amet animi. Sint at autem praesentium et explicabo. Est aliquam modi enim et.
Reiciendis est expedita sit incidunt ut molestias et. Maxime aliquid ut quasi ipsa inventore. Sint assumenda nesciunt molestiae omnis.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...
Voluptatem eos dolor quisquam quasi. Corporis laboriosam fugiat ut similique maxime hic omnis. Qui culpa minima eos in nam quia. Qui illum qui ut corporis aut et. Sed maxime cupiditate incidunt voluptatum provident ipsam quibusdam. Impedit nihil repellendus ipsam sunt ea omnis dolor.
Sed et aut aut odit eum. Quod aliquam rerum consectetur dolorem et sapiente dignissimos.
Corporis nihil autem est quos accusamus et. Eos culpa fugit molestias nobis totam eum rerum. Aspernatur mollitia qui laudantium quas eos rem aut. Quae nemo quisquam quidem illum ad. Impedit velit eaque qui tempore est.