I saw the 250% or so move in that the other day and thought about the time I owned that stock sometime back in the late 90's/early 2000's. Check out the chart back then. lol.
Well that article was good for a quick laugh. At the risk of sounding like a pompous ass, I'm just going to simply say whoever wrote it knows nothing about what "high frequency trading" entails. And for that I don't blame him either, because if he actually did know, he would be trading instead of writing. Another fluff piece about the "mysterious world of Wall st.", and those thieves stealing the average Joe's $. If you look back at my first post in this thread, the strategies that I named, MOC, OPG, Pairs, are exactly the "high frequency" trading that this article describes. This is what I have been doing for a living the last 2 years at Merrill, while these strategies have been in existence for almost 10 years. It makes me sick listening to the talking heads on CNBC talk about this like some new phenomenon, making it the hot new topic. I am upset because all this attention MAY spell the end to the strategies' viability. This is because these strategies rely on the EDGE that it provides you. However with increased attention, comes an increase in players, and thus a decrease in edge. A good example would be online Poker, a similar zero-sum game. When online Poker first broke on to the scene, making money was relatively simple. However, after it's popularity exploded in the last few years, what little edge you had before is now greatly diminished due to an exponential increase in players, reducing the donkey-factor. I find that whenever there is an edge, or imbalance in life, eventually it will be saturated as soon as it gets recognized. Likewise, for the "high frequency trading" strategies, what makes them work is the EDGE that they give you. Not some high speed computing mumbo jumbo. I admit that the higher the computing speed the better. This is because we are sending thousands of orders through black boxes- automated trading machines that will activate based on predetermined criteria. So the faster the computer, the faster our orders go through. However this advantage is minuscule compared to the real edge, which is the strategy itself. Without divulging any proprietary info, basically all these strategies take advantage of some sort of market imbalance. These imbalances are real, however often they are hidden. The simplest and most obvious example that I can give would be the Russell rebalance, which occurs once every year. This day is one of the most important, and most profitable because of the sheer simplicity behind the forces at work. In essence, every fund tracking the Russell 2000 small cap MUST buy/sell which ever stocks being added/deleted. The simplest strategy in the world would be to acquire this list of companies, and trade them accordingly. The reason this day is so profitable is because these funds have total disregard for what price they end up paying for said stocks. This is because their goal is never profitability, but merely to track the Russell 2000 perfectly. This creates a giant imbalance on this day, propelling the respective stocks in exaggerated moves- especially since they are small caps after all, with a small float. Personally, I spent an entire year back-testing and developing my own strategies, since no one with a working black box will give you their own. And in the end, I can tell you this stuff will be profitable even if it is running off a laptop, although with obvious disadvantages. Implying that Wall st. traders are basically stealing from the average Joe because they have faster computers is just patently false. Saying that traders are "manipulating markets in unfair ways" is also ridiculous. Maybe this occurs on isolated instances, however it is not what "high frequency trading" is all about. Traders have the same right to buy and sell stocks, just like "real" buyers and sellers. The only "arms race" going on is developing the next big viable strategy, not advancement in computing speed. And as a side note, I personally know some folks at Goldman and what happened that day when their algos was stolen. That was the same day that the NYSE delayed closing by 15 minutes. Goldman historically has been THE major player in the closing imbalance market. I would estimate that they cover 40% of all these imbalances. However on this day they were completely locked out. What this means is a huge opportunity for everyone else. And to add fuel to the fire, the NYSE closing was delayed due to quote malfunctioning. This meant that when I previously entered my positions at 3:40 and 3:50, thinking I was putting up with 10-20 minutes of market risk, in reality it turned out to be 25-35 minutes. This is a huge deal since this strategy is all about using an exaggerated amount of buying power to make a small %. To limit our risk, we try our best to endure the least amount of market risk as possible, meanwhile taking advantage the the final print. Nearly DOUBLING that time means more than doubling the risk. Now thinking back, I was probably risking my entire account at that point, which is a scary thought. But in many ways this worked to my advantage, since very few traders were willing to put up with such risk, meaning much of the imbalance was left untouched. In addition to knowing what we know now -that Goldman was not a player on that day- it worked out very well. Anyways, I just wanted to give a more accurate representation of what "high frequency trading" is. In case you guys were interested. And to debunk what this article purports.
That hot blonde guest on CNBC just made some great points debunking "front running" by high frequency trading. God I love it when she starts talking "algos".
She wasn't knockout beautiful, but there's something about a pretty good looking woman with brains .... hell, she had me at "colo". The term "colo" is a bit ambiguous, though. I worked for a company that had had colo(cated) servers at a major colo here in the Dallas area. Our clients' (could be an average joe trader) automated trades were exececuted through those servers. They didn't sit right next to NYSE servers in NY, for example - but they were at a major colo in the DFW area. It was an interesting debate, though that was about to come to blows.
1. IMHO she WAS a knockout, and yes the bigger the words she used the bigger my... nvm. 2. It was getting heated because she was exposing their other guest as a fraud. He really didn't understand what high frequency trading was about at all. Really the jist of my previous post was that it is a form of arbitrage. Which is not illegal, nor immoral. 3. What the hell is up with Cabrera. You are an anchor. You have a journalism degree. You know NOTHING about trading. It pains me to listen to you, or any other CNBC talking heads attempt to talk about, or rationalize, the market. The next time you want to act heated about a subject, make sure it is one you are actually familiar with. The whole argument those talking heads were jumping on was the supposed "fee" paid to the exchanges. Like I described in my previous post, speed is NOT why these strategies make money, it is the smart guys writing the algorithms to arbitrage opportunities created by market imbalances that make money. However, our friends at CNBC would never know this, because they don't even know what high frequency trading is to begin with.
Cabrera is annoying. She annoys me, but not more than Kudlow. That dude is a walking political editorial and they have him on for hours. We went from something like DOW 14000 to 9000 before he finally admitted there may be a recession. Charlie Gasparino is another one that gets on my nerves. There are only a handful of people on there that are really financially knowledgable, I think, which I don't necessarily mind as long as they can get the people they interview to debate one another and those people are knowledgable. The shows have taken on a more political tone the past couple of years, which I know can't be escaped, but it's annoying.
Wait what? The dude just gave a base, bear and bull case for the S&P spanning 1000 points. Why didn't he just say, "I don't know where the hell we are going." In depth analysis only here at CNBC folks.
I think no one really knows so it is a fair answer. I remember all those guys repeatedly calling a bottom.
A fair answer is exactly what you just said: "No one really knows. I don't know." Instead of: "I'll give you 3 cases spanning the S&P's range for the last 3 years. I know one will be right!" That's just like not saying anything at all.
Our black box and gray boxes are built from the ground up by programmers who are partners of the firm, who work with us on a daily basis. For order entry most firms use either Sterling Pro, Hammer/Anvil, or Goldman Sach's Redi Plus. The latter being the most reliable from what I hear. We use a proprietary platform based in Sterling Pro. However the software element is only perhaps 5% of the equation. Hardware being even less. I remember when I first started, I did everything manually, using an Excel sheet. I linked Excel to our servers using DDE, for the data feed. Did my needed calculations. Then copy and pasted everything into my order entry. I did this daily, and it was the most primitive way of trading possible. However it still worked. What is required for these strategies are two things. Neither being what that article, or what CNBC, claims to give professional traders the advantage- speed of execution, proximity to the exchanges. It is obviously not about the speed since all I needed was Excel, trading slower than molasses. First you need to acquire/develop a working strategy. This may be difficult for non-professionals because of a lack of input. Although no firm would ever hand out strategies to its traders, they can provide a framework to work from. I have access to data telling me when an imbalance is present. Although this data is nothing proprietary, it may not be available to everyone. Once you have the data, it is a matter of how you use it. How to properly structure and develop an algorithm to take advantage of it. More and more people are using these strategies. And seeing how CNBC is jumping on it, this will probably only get worse. This means this first requirement will become less and less of an advantage as the market becomes saturated. However, there is one more barrier of entry, probably the most important, and it is capital availability. The crux of this strategy is using massive buying power to take advantage of a very small opportunity. As much as I talk about imbalance this imbalance that, it probably only amounts to a 0.1% difference on an average day between imbalance and fair value. So to meaningfully profit from this, you need a lot of money. This is the biggest deterrent to your average trader, since they may not have the buying power available. To give an idea, on an average day I send thousands orders for 100-300 stocks in a 10 minute time frame, from 3:50 to 4:00. On options expiration days this doubles. On the Russell rebalance day, there were 1000+ stocks with imbalances. This means the bare minimum needed on that day was $2,000,000+, assuming you only get into 100 shares each. This is probably something many do not have access to, or simply aren't willing to risk. I think Rebate trading was another topic that was brought up. I fail to see how this strategy is taking advantage of anyone. Rebate traders simply offer their shares on exchanges, and when taken, are rewarded with a small cut. Something like $0.0003 per share. These people trade millions of shares a day, so they reap a nice profit. However this is deserving reward for a needed service- providing liquidity. Besides, they mainly profit from the exchanges, not from investors or average traders.
Yes, I noticed that demand for programmers - tech types still exists even after the layoffs over the past year in the Financial industry. Wilmott Jobs The ability of large firms to have specialists for various tasks is an advantage over the individual trader. I have a friend that uses Excel quite a bit for his research & analysis. His partner does data scraping after the Close and then sends it to him in spreadsheets for analysis. From time to time, he will send me a chart done in Excel and it looks <i>different</i> after looking at charts from Stock Charts and similar. The rest of the time, I think he uses Meta Stock. Finding an <i>Edge</i> is extremely important and extremely difficult to do.
There is demand for quantitative developers. For the group that I work in, 2 years ago when I first got there, there was a team of more than 30 tech guys. Today, there are 3. This means we are having a bunch of technical issues on a constant basis, ie server problems, software/hardware etc. However, the quant guys have only increased over this time. This is because more and more people are trading this way. I guess people are finding out it is profitable, and I can only imagine after how much CNBC talks about it now. All you need is an idea of how to trade it, then pay the quant guy to write it for you. Well I was describing a closing strategy in my previous post. There are other strategies for the open, as well as midday. There are even strategies where you can hold overnight- nothing like what I was trying to do in the TA thread. I wouldn't call that a "strategy". For the closing strategy, it required me to use Excel as a filter. Using the data feed from our servers, to calculate according to set parameters, and to go long/short a list of stocks, and by how many shares. Then to crop this info to our order entry platform, and send the orders. This all had to be done manually, between 3:50 and 4:00. Which can be very stressful. Nowadays, with the help of the quant guys, this is all automated in a black box. So I can sit at home and chat on the BBS instead.
I didn't know that they had cut the support staff like Network & Server Admins and similar. Probably since they are considered <i>overhead</i> and there is difficulty assigning <i>value</i> to what they do while the Quant developers are part of profit centers. I related the story of my friend to show that Excel is still used by people to do many things. Used for statistics, graphing and as a database even though it is a general purpose program rather than a specialized tool. Anything keyed off of the FTSE closing operations? Or the Direxion rebalance or Proshares rebalance?
